Location via proxy:   [ UP ]  
[Report a bug]   [Manage cookies]                

Networkx Reference

You are on page 1of 740

NetworkX Reference

Release 2.3rc1.dev20190222214247

Aric Hagberg, Dan Schult, Pieter Swart

Feb 22, 2019


CONTENTS

1 Introduction 1
1.1 NetworkX Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Graph Creation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.4 Graph Reporting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3
1.5 Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.6 Drawing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
1.7 Data Structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 Graph types 7
2.1 Which graph class should I use? . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Basic graph types . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.3 Graph Views . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 127

3 Algorithms 129
3.1 Approximations and Heuristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 129
3.2 Assortativity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 140
3.3 Bipartite . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 150
3.4 Boundary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
3.5 Bridges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
3.6 Centrality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
3.7 Chains . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
3.8 Chordal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 211
3.9 Clique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 213
3.10 Clustering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
3.11 Coloring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 222
3.12 Communicability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 226
3.13 Communities . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
3.14 Components . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 238
3.15 Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
3.16 Cores . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 282
3.17 Covering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 285
3.18 Cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 286
3.19 Cuts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 290
3.20 Directed Acyclic Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 294
3.21 Dispersion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
3.22 Distance Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
3.23 Distance-Regular Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
3.24 Dominance . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 307
3.25 Dominating Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308

i
3.26 Efficiency . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 310
3.27 Eulerian . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
3.28 Flows . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 313
3.29 Graphical degree sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 340
3.30 Hierarchy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
3.31 Hybrid . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 344
3.32 Isolates . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 346
3.33 Isomorphism . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 347
3.34 Link Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
3.35 Link Prediction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 367
3.36 Lowest Common Ancestor . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 373
3.37 Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
3.38 Minors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 377
3.39 Maximal independent set . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 383
3.40 Node Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
3.41 Operators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
3.42 Planarity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 395
3.43 Reciprocity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 399
3.44 Rich Club . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
3.45 Shortest Paths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
3.46 Similarity Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 437
3.47 Simple Paths . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 444
3.48 Similarity Measures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 448
3.49 s metric . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 451
3.50 Sparsifiers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 452
3.51 Structural holes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 453
3.52 Swap . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 455
3.53 Tournament . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 457
3.54 Traversal . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 460
3.55 Tree . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 472
3.56 Triads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 486
3.57 Vitality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 487
3.58 Voronoi cells . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 488
3.59 Wiener index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 489

4 Functions 491
4.1 Graph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 491
4.2 Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 497
4.3 Edges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 499
4.4 Self loops . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 500
4.5 Attributes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 501
4.6 Freezing graph structure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 506

5 Graph generators 509


5.1 Atlas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 509
5.2 Classic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 510
5.3 Expanders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 518
5.4 Lattice . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 519
5.5 Small . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 522
5.6 Random Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 526
5.7 Duplication Divergence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 537
5.8 Degree Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 539
5.9 Random Clustered . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 545
5.10 Directed . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 546

ii
5.11 Geometric . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 550
5.12 Line Graph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 557
5.13 Ego Graph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 559
5.14 Stochastic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 560
5.15 Intersection . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 561
5.16 Social Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 562
5.17 Community . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 563
5.18 Spectral . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 570
5.19 Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 571
5.20 Non Isomorphic Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 573
5.21 Triads . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574
5.22 Joint Degree Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 574
5.23 Mycielski . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 576

6 Linear algebra 579


6.1 Graph Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 579
6.2 Laplacian Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 580
6.3 Spectrum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 583
6.4 Algebraic Connectivity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 584
6.5 Attribute Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 587
6.6 Modularity Matrices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 591

7 Converting to and from other data formats 595


7.1 To NetworkX Graph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 595
7.2 Dictionaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 596
7.3 Lists . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 597
7.4 Numpy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 598
7.5 Scipy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 605
7.6 Pandas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 607

8 Relabeling nodes 613


8.1 Relabeling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 613

9 Reading and writing graphs 617


9.1 Adjacency List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 617
9.2 Multiline Adjacency List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 620
9.3 Edge List . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 624
9.4 GEXF . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 630
9.5 GML . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 633
9.6 Pickle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 638
9.7 GraphML . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 639
9.8 JSON . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 643
9.9 LEDA . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 648
9.10 YAML . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 649
9.11 SparseGraph6 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 651
9.12 Pajek . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 657
9.13 GIS Shapefile . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 659

10 Drawing 661
10.1 Matplotlib . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 661
10.2 Graphviz AGraph (dot) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 670
10.3 Graphviz with pydot . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 672
10.4 Graph Layout . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 675

11 Randomness 681

iii
12 Exceptions 683
12.1 Exceptions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 683

13 Utilities 685
13.1 Helper Functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 685
13.2 Data Structures and Algorithms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 687
13.3 Random Sequence Generators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 687
13.4 Decorators . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 689
13.5 Cuthill-Mckee Ordering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 692
13.6 Context Managers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 694

14 Glossary 695

A Tutorial 697
A.1 Creating a graph . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 697
A.2 Nodes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 697
A.3 Edges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 698
A.4 What to use as nodes and edges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699
A.5 Accessing edges and neighbors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 699
A.6 Adding attributes to graphs, nodes, and edges . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 700
A.7 Directed graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 701
A.8 Multigraphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702
A.9 Graph generators and graph operations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 702
A.10 Analyzing graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 703
A.11 Drawing graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 703

Bibliography 707

Python Module Index 709

Index 713

iv
CHAPTER

ONE

INTRODUCTION

The structure of NetworkX can be seen by the organization of its source code. The package provides classes for graph
objects, generators to create standard graphs, IO routines for reading in existing datasets, algorithms to analyze the
resulting networks and some basic drawing tools.
Most of the NetworkX API is provided by functions which take a graph object as an argument. Methods of the graph
object are limited to basic manipulation and reporting. This provides modularity of code and documentation. It also
makes it easier for newcomers to learn about the package in stages. The source code for each module is meant to be
easy to read and reading this Python code is actually a good way to learn more about network algorithms, but we have
put a lot of effort into making the documentation sufficient and friendly. If you have suggestions or questions please
contact us by joining the NetworkX Google group.
Classes are named using CamelCase (capital letters at the start of each word). functions, methods and variable
names are lower_case_underscore (lowercase with an underscore representing a space between words).

1.1 NetworkX Basics

After starting Python, import the networkx module with (the recommended way)

>>> import networkx as nx

To save repetition, in the documentation we assume that NetworkX has been imported this way.
If importing networkx fails, it means that Python cannot find the installed module. Check your installation and your
PYTHONPATH.
The following basic graph types are provided as Python classes:
Graph This class implements an undirected graph. It ignores multiple edges between two nodes. It does allow
self-loop edges between a node and itself.
DiGraph Directed graphs, that is, graphs with directed edges. Provides operations common to directed graphs, (a
subclass of Graph).
MultiGraph A flexible graph class that allows multiple undirected edges between pairs of nodes. The additional
flexibility leads to some degradation in performance, though usually not significant.
MultiDiGraph A directed version of a MultiGraph.
Empty graph-like objects are created with

>>> G = nx.Graph()
>>> G = nx.DiGraph()
>>> G = nx.MultiGraph()
>>> G = nx.MultiDiGraph()

1
NetworkX Reference, Release 2.3rc1.dev20190222214247

All graph classes allow any hashable object as a node. Hashable objects include strings, tuples, integers, and more.
Arbitrary edge attributes such as weights and labels can be associated with an edge.
The graph internal data structures are based on an adjacency list representation and implemented using Python dictio-
nary datastructures. The graph adjacency structure is implemented as a Python dictionary of dictionaries; the outer
dictionary is keyed by nodes to values that are themselves dictionaries keyed by neighboring node to the edge at-
tributes associated with that edge. This “dict-of-dicts” structure allows fast addition, deletion, and lookup of nodes
and neighbors in large graphs. The underlying datastructure is accessed directly by methods (the programming in-
terface “API”) in the class definitions. All functions, on the other hand, manipulate graph-like objects solely via
those API methods and not by acting directly on the datastructure. This design allows for possible replacement of the
‘dicts-of-dicts’-based datastructure with an alternative datastructure that implements the same methods.

1.2 Graphs

The first choice to be made when using NetworkX is what type of graph object to use. A graph (network) is a collection
of nodes together with a collection of edges that are pairs of nodes. Attributes are often associated with nodes and/or
edges. NetworkX graph objects come in different flavors depending on two main properties of the network:
• Directed: Are the edges directed? Does the order of the edge pairs (𝑢, 𝑣) matter? A directed graph is specified
by the “Di” prefix in the class name, e.g. DiGraph(). We make this distinction because many classical graph
properties are defined differently for directed graphs.
• Multi-edges: Are multiple edges allowed between each pair of nodes? As you might imagine, multiple edges
requires a different data structure, though clever users could design edge data attributes to support this function-
ality. We provide a standard data structure and interface for this type of graph using the prefix “Multi”, e.g.,
MultiGraph().
The basic graph classes are named: Graph, DiGraph, MultiGraph, and MultiDiGraph

1.2.1 Nodes and Edges

The next choice you have to make when specifying a graph is what kinds of nodes and edges to use.
If the topology of the network is all you care about then using integers or strings as the nodes makes sense and you
need not worry about edge data. If you have a data structure already in place to describe nodes you can simply use
that structure as your nodes provided it is hashable. If it is not hashable you can use a unique identifier to represent
the node and assign the data as a node attribute.
Edges often have data associated with them. Arbitrary data can be associated with edges as an edge attribute. If the
data is numeric and the intent is to represent a weighted graph then use the ‘weight’ keyword for the attribute. Some
of the graph algorithms, such as Dijkstra’s shortest path algorithm, use this attribute name by default to get the weight
for each edge.
Attributes can be assigned to an edge by using keyword/value pairs when adding edges. You can use any keyword to
name your attribute and can then query the edge data using that attribute keyword.
Once you’ve decided how to encode the nodes and edges, and whether you have an undirected/directed graph with or
without multiedges you are ready to build your network.

1.3 Graph Creation

NetworkX graph objects can be created in one of three ways:


• Graph generators—standard algorithms to create network topologies.

2 Chapter 1. Introduction
NetworkX Reference, Release 2.3rc1.dev20190222214247

• Importing data from pre-existing (usually file) sources.


• Adding edges and nodes explicitly.
Explicit addition and removal of nodes/edges is the easiest to describe. Each graph object supplies methods to manip-
ulate the graph. For example,

>>> import networkx as nx


>>> G = nx.Graph()
>>> G.add_edge(1, 2) # default edge data=1
>>> G.add_edge(2, 3, weight=0.9) # specify edge data

Edge attributes can be anything:

>>> import math


>>> G.add_edge('y', 'x', function=math.cos)
>>> G.add_node(math.cos) # any hashable can be a node

You can add many edges at one time:

>>> elist = [(1, 2), (2, 3), (1, 4), (4, 2)]
>>> G.add_edges_from(elist)
>>> elist = [('a', 'b', 5.0), ('b', 'c', 3.0), ('a', 'c', 1.0), ('c', 'd', 7.3)]
>>> G.add_weighted_edges_from(elist)

See the Tutorial for more examples.


Some basic graph operations such as union and intersection are described in the operators module documentation.
Graph generators such as binomial_graph() and erdos_renyi_graph() are provided in the graph genera-
tors subpackage.
For importing network data from formats such as GML, GraphML, edge list text files see the reading and writing
graphs subpackage.

1.4 Graph Reporting

Class views provide basic reporting of nodes, neighbors, edges and degree. These views provide iteration over the
properties as well as membership queries and data attribute lookup. The views refer to the graph data structure so
changes to the graph are reflected in the views. This is analogous to dictionary views in Python 3. If you want to
change the graph while iterating you will need to use e.g. for e in list(G.edges):. The views provide
set-like operations, e.g. union and intersection, as well as dict-like lookup and iteration of the data attributes us-
ing G.edges[u, v]['color'] and for e, datadict in G.edges.items():. Methods G.edges.
items() and G.edges.values() are familiar from python dicts. In addition G.edges.data() provides
specific attribute iteration e.g. for e, e_color in G.edges.data('color'):.
The basic graph relationship of an edge can be obtained in two ways. One can look for neighbors of a node or one can
look for edges. We jokingly refer to people who focus on nodes/neighbors as node-centric and people who focus on
edges as edge-centric. The designers of NetworkX tend to be node-centric and view edges as a relationship between
nodes. You can see this by our choice of lookup notation like G[u] providing neighbors (adjacency) while edge
lookup is G.edges[u, v]. Most data structures for sparse graphs are essentially adjacency lists and so fit this
perspective. In the end, of course, it doesn’t really matter which way you examine the graph. G.edges removes
duplicate representations of undirected edges while neighbor reporting across all nodes will naturally report both
directions.
Any properties that are more complicated than edges, neighbors and degree are provided by functions. For exam-
ple nx.triangles(G, n) gives the number of triangles which include node n as a vertex. These functions are

1.4. Graph Reporting 3


NetworkX Reference, Release 2.3rc1.dev20190222214247

grouped in the code and documentation under the term algorithms.

1.5 Algorithms

A number of graph algorithms are provided with NetworkX. These include shortest path, and breadth first search (see
traversal), clustering and isomorphism algorithms and others. There are many that we have not developed yet too. If
you implement a graph algorithm that might be useful for others please let us know through the NetworkX Google
group or the Github Developer Zone.
As an example here is code to use Dijkstra’s algorithm to find the shortest weighted path:

>>> G = nx.Graph()
>>> e = [('a', 'b', 0.3), ('b', 'c', 0.9), ('a', 'c', 0.5), ('c', 'd', 1.2)]
>>> G.add_weighted_edges_from(e)
>>> print(nx.dijkstra_path(G, 'a', 'd'))
['a', 'c', 'd']

1.6 Drawing

While NetworkX is not designed as a network drawing tool, we provide a simple interface to drawing packages
and some simple layout algorithms. We interface to the excellent Graphviz layout tools like dot and neato with the
(suggested) pygraphviz package or the pydot interface. Drawing can be done using external programs or the Matplotlib
Python package. Interactive GUI interfaces are possible, though not provided. The drawing tools are provided in the
module drawing.
The basic drawing functions essentially place the nodes on a scatterplot using the positions you provide via a dictionary
or the positions are computed with a layout function. The edges are lines between those dots.

>>> import matplotlib.pyplot as plt


>>> G = nx.cubical_graph()
>>> plt.subplot(121)
<matplotlib.axes._subplots.AxesSubplot object at ...>
>>> nx.draw(G) # default spring_layout
>>> plt.subplot(122)
<matplotlib.axes._subplots.AxesSubplot object at ...>
>>> nx.draw(G, pos=nx.circular_layout(G), nodecolor='r', edge_color='b')

4 Chapter 1. Introduction
NetworkX Reference, Release 2.3rc1.dev20190222214247

See the examples for more ideas.

1.7 Data Structure

NetworkX uses a “dictionary of dictionaries of dictionaries” as the basic network data structure. This allows fast
lookup with reasonable storage for large sparse networks. The keys are nodes so G[u] returns an adjacency dictionary
keyed by neighbor to the edge attribute dictionary. A view of the adjacency data structure is provided by the dict-like
object G.adj as e.g. for node, nbrsdict in G.adj.items():. The expression G[u][v] returns the
edge attribute dictionary itself. A dictionary of lists would have also been possible, but not allow fast edge detection
nor convenient storage of edge data.
Advantages of dict-of-dicts-of-dicts data structure:
• Find edges and remove edges with two dictionary look-ups.
• Prefer to “lists” because of fast lookup with sparse storage.
• Prefer to “sets” since data can be attached to edge.
• G[u][v] returns the edge attribute dictionary.
• n in G tests if node n is in graph G.
• for n in G: iterates through the graph.
• for nbr in G[n]: iterates through neighbors.

1.7. Data Structure 5


NetworkX Reference, Release 2.3rc1.dev20190222214247

As an example, here is a representation of an undirected graph with the edges (𝐴, 𝐵) and (𝐵, 𝐶).

>>> G = nx.Graph()
>>> G.add_edge('A', 'B')
>>> G.add_edge('B', 'C')
>>> print(G.adj)
{'A': {'B': {}}, 'B': {'A': {}, 'C': {}}, 'C': {'B': {}}}

The data structure gets morphed slightly for each base graph class. For DiGraph two dict-of-dicts-of-dicts structures
are provided, one for successors (G.succ) and one for predecessors (G.pred). For MultiGraph/MultiDiGraph
we use a dict-of-dicts-of-dicts-of-dicts1 where the third dictionary is keyed by an edge key identifier to the fourth
dictionary which contains the edge attributes for that edge between the two nodes.
Graphs provide two interfaces to the edge data attributes: adjacency and edges. So G[u][v]['width'] is the same
as G.edges[u, v]['width'].

>>> G = nx.Graph()
>>> G.add_edge(1, 2, color='red', weight=0.84, size=300)
>>> print(G[1][2]['size'])
300
>>> print(G.edges[1, 2]['color'])
red

• ;
• ;
• .

1 “It’s dictionaries all the way down.”

6 Chapter 1. Introduction
CHAPTER

TWO

GRAPH TYPES

NetworkX provides data structures and methods for storing graphs.


All NetworkX graph classes allow (hashable) Python objects as nodes and any Python object can be assigned as an
edge attribute.
The choice of graph class depends on the structure of the graph you want to represent.

2.1 Which graph class should I use?

Graph Type NetworkX Class


Undirected Simple Graph
Directed Simple DiGraph
With Self-loops Graph, DiGraph
With Parallel edges MultiGraph, MultiDiGraph

2.2 Basic graph types

2.2.1 Graph—Undirected graphs with self loops

Overview

class Graph(incoming_graph_data=None, **attr)


Base class for undirected graphs.
A Graph stores nodes and edges with optional data, or attributes.
Graphs hold undirected edges. Self loops are allowed but multiple (parallel) edges are not.
Nodes can be arbitrary (hashable) Python objects with optional key/value attributes. By convention None is not
used as a node.
Edges are represented as links between nodes with optional key/value attributes.
Parameters
• incoming_graph_data (input graph (optional, default: None)) – Data to initialize graph. If
None (default) an empty graph is created. The data can be any format that is supported by
the to_networkx_graph() function, currently including edge list, dict of dicts, dict of lists,
NetworkX graph, NumPy matrix or 2d ndarray, SciPy sparse matrix, or PyGraphviz graph.

7
NetworkX Reference, Release 2.3rc1.dev20190222214247

• attr (keyword arguments, optional (default= no attributes)) – Attributes to add to graph as


key=value pairs.
See also:
DiGraph, MultiGraph, MultiDiGraph, OrderedGraph

Examples

Create an empty graph structure (a “null graph”) with no nodes and no edges.

>>> G = nx.Graph()

G can be grown in several ways.


Nodes:
Add one node at a time:

>>> G.add_node(1)

Add the nodes from any container (a list, dict, set or even the lines from a file or the nodes from another graph).

>>> G.add_nodes_from([2, 3])


>>> G.add_nodes_from(range(100, 110))
>>> H = nx.path_graph(10)
>>> G.add_nodes_from(H)

In addition to strings and integers any hashable Python object (except None) can represent a node, e.g. a
customized node object, or even another Graph.

>>> G.add_node(H)

Edges:
G can also be grown by adding edges.
Add one edge,

>>> G.add_edge(1, 2)

a list of edges,

>>> G.add_edges_from([(1, 2), (1, 3)])

or a collection of edges,

>>> G.add_edges_from(H.edges)

If some edges connect nodes not yet in the graph, the nodes are added automatically. There are no errors when
adding nodes or edges that already exist.
Attributes:
Each graph, node, and edge can hold key/value attribute pairs in an associated attribute dictionary (the keys
must be hashable). By default these are empty, but can be added or changed using add_edge, add_node or direct
manipulation of the attribute dictionaries named graph, node and edge respectively.

8 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.Graph(day="Friday")
>>> G.graph
{'day': 'Friday'}

Add node attributes using add_node(), add_nodes_from() or G.nodes


>>> G.add_node(1, time='5pm')
>>> G.add_nodes_from([3], time='2pm')
>>> G.nodes[1]
{'time': '5pm'}
>>> G.nodes[1]['room'] = 714 # node must exist already to use G.nodes
>>> del G.nodes[1]['room'] # remove attribute
>>> list(G.nodes(data=True))
[(1, {'time': '5pm'}), (3, {'time': '2pm'})]

Add edge attributes using add_edge(), add_edges_from(), subscript notation, or G.edges.


>>> G.add_edge(1, 2, weight=4.7 )
>>> G.add_edges_from([(3, 4), (4, 5)], color='red')
>>> G.add_edges_from([(1, 2, {'color': 'blue'}), (2, 3, {'weight': 8})])
>>> G[1][2]['weight'] = 4.7
>>> G.edges[1, 2]['weight'] = 4

Warning: we protect the graph data structure by making G.edges a read-only dict-like structure. However, you
can assign to attributes in e.g. G.edges[1, 2]. Thus, use 2 sets of brackets to add/change data attributes:
G.edges[1, 2]['weight'] = 4 (For multigraphs: MG.edges[u, v, key][name] = value).
Shortcuts:
Many common graph features allow python syntax to speed reporting.
>>> 1 in G # check if node in graph
True
>>> [n for n in G if n < 3] # iterate through nodes
[1, 2]
>>> len(G) # number of nodes in graph
5

Often the best way to traverse all edges of a graph is via the neighbors. The neighbors are reported as an
adjacency-dict G.adj or G.adjacency()
>>> for n, nbrsdict in G.adjacency():
... for nbr, eattr in nbrsdict.items():
... if 'weight' in eattr:
... # Do something useful with the edges
... pass

But the edges() method is often more convenient:


>>> for u, v, weight in G.edges.data('weight'):
... if weight is not None:
... # Do something useful with the edges
... pass

Reporting:
Simple graph information is obtained using object-attributes and methods. Reporting typically provides
views instead of containers to reduce memory usage. The views update as the graph is updated simi-
larly to dict-views. The objects nodes, `edges and adj provide access to data attributes via lookup

2.2. Basic graph types 9


NetworkX Reference, Release 2.3rc1.dev20190222214247

(e.g. nodes[n], `edges[u, v], adj[u][v]) and iteration (e.g. nodes.items(), nodes.
data('color'), nodes.data('color', default='blue') and similarly for edges) Views exist
for nodes, edges, neighbors()/adj and degree.
For details on these and other miscellaneous methods, see below.
Subclasses (Advanced):
The Graph class uses a dict-of-dict-of-dict data structure. The outer dict (node_dict) holds adjacency information
keyed by node. The next dict (adjlist_dict) represents the adjacency information and holds edge data keyed by
neighbor. The inner dict (edge_attr_dict) represents the edge data and holds edge attribute values keyed by
attribute names.
Each of these three dicts can be replaced in a subclass by a user defined dict-like object. In general, the
dict-like features should be maintained but extra features can be added. To replace one of the dicts create
a new graph class by changing the class(!) variable holding the factory for that dict-like structure. The vari-
able names are node_dict_factory, node_attr_dict_factory, adjlist_inner_dict_factory, adjlist_outer_dict_factory,
edge_attr_dict_factory and graph_attr_dict_factory.
node_dict_factory [function, (default: dict)] Factory function to be used to create the dict containing node
attributes, keyed by node id. It should require no arguments and return a dict-like object
node_attr_dict_factory: function, (default: dict) Factory function to be used to create the node attribute dict
which holds attribute values keyed by attribute name. It should require no arguments and return a dict-like
object
adjlist_outer_dict_factory [function, (default: dict)] Factory function to be used to create the outer-most dict
in the data structure that holds adjacency info keyed by node. It should require no arguments and return a
dict-like object.
adjlist_inner_dict_factory [function, (default: dict)] Factory function to be used to create the adjacency list
dict which holds edge data keyed by neighbor. It should require no arguments and return a dict-like object
edge_attr_dict_factory [function, (default: dict)] Factory function to be used to create the edge attribute dict
which holds attribute values keyed by attribute name. It should require no arguments and return a dict-like
object.
graph_attr_dict_factory [function, (default: dict)] Factory function to be used to create the graph attribute
dict which holds attribute values keyed by attribute name. It should require no arguments and return a
dict-like object.
Typically, if your extension doesn’t impact the data structure all methods will inherit without issue except:
to_directed/to_undirected. By default these methods create a DiGraph/Graph class and you probably
want them to create your extension of a DiGraph/Graph. To facilitate this we define two class variables that you
can set in your subclass.
to_directed_class [callable, (default: DiGraph or MultiDiGraph)] Class to create a new graph structure in the
to_directed method. If None, a NetworkX class (DiGraph or MultiDiGraph) is used.
to_undirected_class [callable, (default: Graph or MultiGraph)] Class to create a new graph structure in the
to_undirected method. If None, a NetworkX class (Graph or MultiGraph) is used.

Examples

Create a low memory graph class that effectively disallows edge attributes by using a single attribute dict for all
edges. This reduces the memory used, but you lose edge attributes.

>>> class ThinGraph(nx.Graph):


... all_edge_dict = {'weight': 1}
(continues on next page)

10 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


... def single_edge_dict(self):
... return self.all_edge_dict
... edge_attr_dict_factory = single_edge_dict
>>> G = ThinGraph()
>>> G.add_edge(2, 1)
>>> G[2][1]
{'weight': 1}
>>> G.add_edge(2, 2)
>>> G[2][1] is G[2][2]
True

Please see ordered for more examples of creating graph subclasses by overwriting the base class dict with
a dictionary-like object.

Methods

Adding and removing nodes and edges

Graph.__init__([incoming_graph_data]) Initialize a graph with edges, name, or graph attributes.


Graph.add_node(node_for_adding, **attr) Add a single node node_for_adding and update
node attributes.
Graph.add_nodes_from(nodes_for_adding, Add multiple nodes.
**attr)
Graph.remove_node(n) Remove node n.
Graph.remove_nodes_from(nodes) Remove multiple nodes.
Graph.add_edge(u_of_edge, v_of_edge, **attr) Add an edge between u and v.
Graph.add_edges_from(ebunch_to_add, **attr) Add all the edges in ebunch_to_add.
Graph.add_weighted_edges_from(ebunch_to_add) Add weighted edges in ebunch_to_add with speci-
fied weight attr
Graph.remove_edge(u, v) Remove the edge between u and v.
Graph.remove_edges_from(ebunch) Remove all edges specified in ebunch.
Graph.update([edges, nodes]) Update the graph using nodes/edges/graphs as input.
Graph.clear() Remove all nodes and edges from the graph.

networkx.Graph.__init__

Graph.__init__(incoming_graph_data=None, **attr)
Initialize a graph with edges, name, or graph attributes.
Parameters
• incoming_graph_data (input graph (optional, default: None)) – Data to initialize graph. If
None (default) an empty graph is created. The data can be an edge list, or any NetworkX
graph object. If the corresponding optional Python packages are installed the data can also
be a NumPy matrix or 2d ndarray, a SciPy sparse matrix, or a PyGraphviz graph.
• attr (keyword arguments, optional (default= no attributes)) – Attributes to add to graph as
key=value pairs.
See also:
convert()

2.2. Basic graph types 11


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G = nx.Graph(name='my graph')
>>> e = [(1, 2), (2, 3), (3, 4)] # list of edges
>>> G = nx.Graph(e)

Arbitrary graph attribute pairs (key=value) may be assigned

>>> G = nx.Graph(e, day="Friday")


>>> G.graph
{'day': 'Friday'}

networkx.Graph.add_node

Graph.add_node(node_for_adding, **attr)
Add a single node node_for_adding and update node attributes.
Parameters
• node_for_adding (node) – A node can be any hashable Python object except None.
• attr (keyword arguments, optional) – Set or change node attributes using key=value.
See also:
add_nodes_from()

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_node(1)
>>> G.add_node('Hello')
>>> K3 = nx.Graph([(0, 1), (1, 2), (2, 0)])
>>> G.add_node(K3)
>>> G.number_of_nodes()
3

Use keywords set/change node attributes:

>>> G.add_node(1, size=10)


>>> G.add_node(3, weight=0.4, UTM=('13S', 382871, 3972649))

Notes

A hashable object is one that can be used as a key in a Python dictionary. This includes strings, numbers, tuples
of strings and numbers, etc.
On many platforms hashable items also include mutables such as NetworkX Graphs, though one should be
careful that the hash doesn’t change on mutables.

12 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.Graph.add_nodes_from

Graph.add_nodes_from(nodes_for_adding, **attr)
Add multiple nodes.
Parameters
• nodes_for_adding (iterable container) – A container of nodes (list, dict, set, etc.). OR A
container of (node, attribute dict) tuples. Node attributes are updated using the attribute dict.
• attr (keyword arguments, optional (default= no attributes)) – Update attributes for all nodes
in nodes. Node attributes specified in nodes as a tuple take precedence over attributes spec-
ified via keyword arguments.
See also:
add_node()

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_nodes_from('Hello')
>>> K3 = nx.Graph([(0, 1), (1, 2), (2, 0)])
>>> G.add_nodes_from(K3)
>>> sorted(G.nodes(), key=str)
[0, 1, 2, 'H', 'e', 'l', 'o']

Use keywords to update specific node attributes for every node.

>>> G.add_nodes_from([1, 2], size=10)


>>> G.add_nodes_from([3, 4], weight=0.4)

Use (node, attrdict) tuples to update attributes for specific nodes.

>>> G.add_nodes_from([(1, dict(size=11)), (2, {'color':'blue'})])


>>> G.nodes[1]['size']
11
>>> H = nx.Graph()
>>> H.add_nodes_from(G.nodes(data=True))
>>> H.nodes[1]['size']
11

networkx.Graph.remove_node

Graph.remove_node(n)
Remove node n.
Removes the node n and all adjacent edges. Attempting to remove a non-existent node will raise an exception.
Parameters n (node) – A node in the graph
Raises NetworkXError – If n is not in the graph.
See also:
remove_nodes_from()

2.2. Basic graph types 13


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> list(G.edges)
[(0, 1), (1, 2)]
>>> G.remove_node(1)
>>> list(G.edges)
[]

networkx.Graph.remove_nodes_from

Graph.remove_nodes_from(nodes)
Remove multiple nodes.
Parameters nodes (iterable container) – A container of nodes (list, dict, set, etc.). If a node in the
container is not in the graph it is silently ignored.
See also:
remove_node()

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> e = list(G.nodes)
>>> e
[0, 1, 2]
>>> G.remove_nodes_from(e)
>>> list(G.nodes)
[]

networkx.Graph.add_edge

Graph.add_edge(u_of_edge, v_of_edge, **attr)


Add an edge between u and v.
The nodes u and v will be automatically added if they are not already in the graph.
Edge attributes can be specified with keywords or by directly accessing the edge’s attribute dictionary. See
examples below.
Parameters
• u, v (nodes) – Nodes can be, for example, strings or numbers. Nodes must be hashable (and
not None) Python objects.
• attr (keyword arguments, optional) – Edge data (or labels or objects) can be assigned using
keyword arguments.
See also:

add_edges_from() add a collection of edges

14 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Adding an edge that already exists updates the edge data.


Many NetworkX algorithms designed for weighted graphs use an edge attribute (by default weight) to hold a
numerical value.

Examples

The following all add the edge e=(1, 2) to graph G:

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> e = (1, 2)
>>> G.add_edge(1, 2) # explicit two-node form
>>> G.add_edge(*e) # single edge as tuple of two nodes
>>> G.add_edges_from([(1, 2)]) # add edges from iterable container

Associate data to edges using keywords:

>>> G.add_edge(1, 2, weight=3)


>>> G.add_edge(1, 3, weight=7, capacity=15, length=342.7)

For non-string attribute keys, use subscript notation.

>>> G.add_edge(1, 2)
>>> G[1][2].update({0: 5})
>>> G.edges[1, 2].update({0: 5})

networkx.Graph.add_edges_from

Graph.add_edges_from(ebunch_to_add, **attr)
Add all the edges in ebunch_to_add.
Parameters
• ebunch_to_add (container of edges) – Each edge given in the container will be added to
the graph. The edges must be given as as 2-tuples (u, v) or 3-tuples (u, v, d) where d is a
dictionary containing edge data.
• attr (keyword arguments, optional) – Edge data (or labels or objects) can be assigned using
keyword arguments.
See also:

add_edge() add a single edge


add_weighted_edges_from() convenient way to add weighted edges

Notes

Adding the same edge twice has no effect but any edge data will be updated when each duplicate edge is added.
Edge attributes specified in an ebunch take precedence over attributes specified via keyword arguments.

2.2. Basic graph types 15


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_edges_from([(0, 1), (1, 2)]) # using a list of edge tuples
>>> e = zip(range(0, 3), range(1, 4))
>>> G.add_edges_from(e) # Add the path graph 0-1-2-3

Associate data to edges

>>> G.add_edges_from([(1, 2), (2, 3)], weight=3)


>>> G.add_edges_from([(3, 4), (1, 4)], label='WN2898')

networkx.Graph.add_weighted_edges_from

Graph.add_weighted_edges_from(ebunch_to_add, weight=’weight’, **attr)


Add weighted edges in ebunch_to_add with specified weight attr
Parameters
• ebunch_to_add (container of edges) – Each edge given in the list or container will be added
to the graph. The edges must be given as 3-tuples (u, v, w) where w is a number.
• weight (string, optional (default= ‘weight’)) – The attribute name for the edge weights to
be added.
• attr (keyword arguments, optional (default= no attributes)) – Edge attributes to add/update
for all edges.
See also:

add_edge() add a single edge


add_edges_from() add multiple edges

Notes

Adding the same edge twice for Graph/DiGraph simply updates the edge data. For MultiGraph/MultiDiGraph,
duplicate edges are stored.

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_weighted_edges_from([(0, 1, 3.0), (1, 2, 7.5)])

networkx.Graph.remove_edge

Graph.remove_edge(u, v)
Remove the edge between u and v.
Parameters u, v (nodes) – Remove the edge between nodes u and v.
Raises NetworkXError – If there is not an edge between u and v.
See also:

16 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

remove_edges_from() remove a collection of edges

Examples

>>> G = nx.path_graph(4) # or DiGraph, etc


>>> G.remove_edge(0, 1)
>>> e = (1, 2)
>>> G.remove_edge(*e) # unpacks e from an edge tuple
>>> e = (2, 3, {'weight':7}) # an edge with attribute data
>>> G.remove_edge(*e[:2]) # select first part of edge tuple

networkx.Graph.remove_edges_from

Graph.remove_edges_from(ebunch)
Remove all edges specified in ebunch.
Parameters ebunch (list or container of edge tuples) – Each edge given in the list or container will
be removed from the graph. The edges can be:
• 2-tuples (u, v) edge between u and v.
• 3-tuples (u, v, k) where k is ignored.
See also:

remove_edge() remove a single edge

Notes

Will fail silently if an edge in ebunch is not in the graph.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> ebunch=[(1, 2), (2, 3)]
>>> G.remove_edges_from(ebunch)

networkx.Graph.update

Graph.update(edges=None, nodes=None)
Update the graph using nodes/edges/graphs as input.
Like dict.update, this method takes a graph as input, adding the graph’s noes and edges to this graph. It can also
take two inputs: edges and nodes. Finally it can take either edges or nodes. To specify only nodes the keyword
nodes must be used.
The collections of edges and nodes are treated similarly to the add_edges_from/add_nodes_from methods.
When iterated, they should yield 2-tuples (u, v) or 3-tuples (u, v, datadict).
Parameters

2.2. Basic graph types 17


NetworkX Reference, Release 2.3rc1.dev20190222214247

• edges (Graph object, collection of edges, or None) – The first parameter can be a graph or
some edges. If it has attributes nodes and edges, then it is taken to be a Graph-like object
and those attributes are used as collections of nodes and edges to be added to the graph. If
the first parameter does not have those attributes, it is treated as a collection of edges and
added to the graph. If the first argument is None, no edges are added.
• nodes (collection of nodes, or None) – The second parameter is treated as a collection of
nodes to be added to the graph unless it is None. If edges is None and nodes is
None an exception is raised. If the first parameter is a Graph, then nodes is ignored.

Examples

>>> G = nx.path_graph(5)
>>> G.update(nx.complete_graph(range(4,10)))
>>> from itertools import combinations
>>> edges = ((u, v, {'power': u * v})
... for u, v in combinations(range(10, 20), 2)
... if u * v < 225)
>>> nodes = [1000] # for singleton, use a container
>>> G.update(edges, nodes)

Notes

It you want to update the graph using an adjacency structure it is straightforward to obtain the edges/nodes from
adjacency. The following examples provide common cases, your adjacency may be slightly different and require
tweaks of these examples.

>>> # dict-of-set/list/tuple
>>> adj = {1: {2, 3}, 2: {1, 3}, 3: {1, 2}}
>>> e = [(u, v) for u, nbrs in adj.items() for v in nbrs]
>>> G.update(edges=e, nodes=adj)

>>> DG = nx.DiGraph()
>>> # dict-of-dict-of-attribute
>>> adj = {1: {2: 1.3, 3: 0.7}, 2: {1: 1.4}, 3: {1: 0.7}}
>>> e = [(u, v, {'weight': d}) for u, nbrs in adj.items()
... for v, d in nbrs.items()]
>>> DG.update(edges=e, nodes=adj)

>>> # dict-of-dict-of-dict
>>> adj = {1: {2: {'weight': 1.3}, 3: {'color': 0.7, 'weight':1.2}}}
>>> e = [(u, v, {'weight': d}) for u, nbrs in adj.items()
... for v, d in nbrs.items()]
>>> DG.update(edges=e, nodes=adj)

>>> # predecessor adjacency (dict-of-set)


>>> pred = {1: {2, 3}, 2: {3}, 3: {3}}
>>> e = [(v, u) for u, nbrs in pred.items() for v in nbrs]

>>> # MultiGraph dict-of-dict-of-dict-of-attribute


>>> MDG = nx.MultiDiGraph()
>>> adj = {1: {2: {0: {'weight': 1.3}, 1: {'weight': 1.2}}},
... 3: {2: {0: {'weight': 0.7}}}}
(continues on next page)

18 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> e = [(u, v, ekey, d) for u, nbrs in adj.items()
... for v, keydict in nbrs.items()
... for ekey, d in keydict.items()]
>>> MDG.update(edges=e)

See also:

add_edges_from() add multiple edges to a graph


add_nodes_from() add multiple nodes to a graph

networkx.Graph.clear

Graph.clear()
Remove all nodes and edges from the graph.
This also removes the name, and all graph, node, and edge attributes.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.clear()
>>> list(G.nodes)
[]
>>> list(G.edges)
[]

Reporting nodes edges and neighbors

Graph.nodes A NodeView of the Graph as G.nodes or G.nodes().


Graph.__iter__() Iterate over the nodes.
Graph.has_node(n) Returns True if the graph contains the node n.
Graph.__contains__(n) Returns True if n is a node, False otherwise.
Graph.edges An EdgeView of the Graph as G.edges or G.edges().
Graph.has_edge(u, v) Returns True if the edge (u, v) is in the graph.
Graph.get_edge_data(u, v[, default]) Returns the attribute dictionary associated with edge (u,
v).
Graph.neighbors(n) Returns an iterator over all neighbors of node n.
Graph.adj Graph adjacency object holding the neighbors of each
node.
Graph.__getitem__(n) Returns a dict of neighbors of node n.
Graph.adjacency() Returns an iterator over (node, adjacency dict) tuples for
all nodes.
Graph.nbunch_iter([nbunch]) Returns an iterator over nodes contained in nbunch that
are also in the graph.

2.2. Basic graph types 19


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.Graph.nodes

Graph.nodes
A NodeView of the Graph as G.nodes or G.nodes().
Can be used as G.nodes for data lookup and for set-like operations. Can also be used as G.
nodes(data='color', default=None) to return a NodeDataView which reports specific node data
but no set operations. It presents a dict-like interface as well with G.nodes.items() iterating over (node,
nodedata) 2-tuples and G.nodes[3]['foo'] providing the value of the foo attribute for node 3. In
addition, a view G.nodes.data('foo') provides a dict-like interface to the foo attribute of each node.
G.nodes.data('foo', default=1) provides a default for nodes that do not have attribute foo.
Parameters
• data (string or bool, optional (default=False)) – The node attribute returned in 2-tuple (n,
ddict[data]). If True, return entire node attribute dict as (n, ddict). If False, return just the
nodes n.
• default (value, optional (default=None)) – Value used for nodes that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns
Allows set-like operations over the nodes as well as node attribute dict lookup and calling to
get a NodeDataView. A NodeDataView iterates over (n, data) and has no set operations. A
NodeView iterates over n and includes set operations.
When called, if data is False, an iterator over nodes. Otherwise an iterator of 2-tuples (node,
attribute value) where the attribute is specified in data. If data is True then the attribute becomes
the entire data dictionary.
Return type NodeView

Notes

If your node data is not needed, it is simpler and equivalent to use the expression for n in G, or list(G).

Examples

There are two simple ways of getting a list of all nodes in the graph:

>>> G = nx.path_graph(3)
>>> list(G.nodes)
[0, 1, 2]
>>> list(G)
[0, 1, 2]

To get the node data along with the nodes:

>>> G.add_node(1, time='5pm')


>>> G.nodes[0]['foo'] = 'bar'
>>> list(G.nodes(data=True))
[(0, {'foo': 'bar'}), (1, {'time': '5pm'}), (2, {})]
>>> list(G.nodes.data())
[(0, {'foo': 'bar'}), (1, {'time': '5pm'}), (2, {})]

20 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> list(G.nodes(data='foo'))
[(0, 'bar'), (1, None), (2, None)]
>>> list(G.nodes.data('foo'))
[(0, 'bar'), (1, None), (2, None)]

>>> list(G.nodes(data='time'))
[(0, None), (1, '5pm'), (2, None)]
>>> list(G.nodes.data('time'))
[(0, None), (1, '5pm'), (2, None)]

>>> list(G.nodes(data='time', default='Not Available'))


[(0, 'Not Available'), (1, '5pm'), (2, 'Not Available')]
>>> list(G.nodes.data('time', default='Not Available'))
[(0, 'Not Available'), (1, '5pm'), (2, 'Not Available')]

If some of your nodes have an attribute and the rest are assumed to have a default attribute value you can create
a dictionary from node/attribute pairs using the default keyword argument to guarantee the value is never
None:

>>> G = nx.Graph()
>>> G.add_node(0)
>>> G.add_node(1, weight=2)
>>> G.add_node(2, weight=3)
>>> dict(G.nodes(data='weight', default=1))
{0: 1, 1: 2, 2: 3}

networkx.Graph.__iter__

Graph.__iter__()
Iterate over the nodes. Use: ‘for n in G’.
Returns niter – An iterator over all nodes in the graph.
Return type iterator

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> [n for n in G]
[0, 1, 2, 3]
>>> list(G)
[0, 1, 2, 3]

networkx.Graph.has_node

Graph.has_node(n)
Returns True if the graph contains the node n.
Identical to n in G
Parameters n (node)

2.2. Basic graph types 21


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.has_node(0)
True

It is more readable and simpler to use

>>> 0 in G
True

networkx.Graph.__contains__

Graph.__contains__(n)
Returns True if n is a node, False otherwise. Use: ‘n in G’.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> 1 in G
True

networkx.Graph.edges

Graph.edges
An EdgeView of the Graph as G.edges or G.edges().
edges(self, nbunch=None, data=False, default=None)
The EdgeView provides set-like operations on the edge-tuples as well as edge attribute lookup. When called,
it also provides an EdgeDataView object which allows control of access to edge attributes (but does not pro-
vide set-like operations). Hence, G.edges[u, v]['color'] provides the value of the color attribute for
edge (u, v) while for (u, v, c) in G.edges.data('color', default='red'): iterates
through all the edges yielding the color attribute with default 'red' if no color attribute exists.
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• data (string or bool, optional (default=False)) – The edge attribute returned in 3-tuple (u, v,
ddict[data]). If True, return edge attribute dict in 3-tuple (u, v, ddict). If False, return 2-tuple
(u, v).
• default (value, optional (default=None)) – Value used for edges that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns edges – A view of edge attributes, usually it iterates over (u, v) or (u, v, d) tuples of edges,
but can also be used for attribute lookup as edges[u, v]['foo'].
Return type EdgeView

22 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Nodes in nbunch that are not in the graph will be (quietly) ignored. For directed graphs this returns the out-edges.

Examples

>>> G = nx.path_graph(3) # or MultiGraph, etc


>>> G.add_edge(2, 3, weight=5)
>>> [e for e in G.edges]
[(0, 1), (1, 2), (2, 3)]
>>> G.edges.data() # default data is {} (empty dict)
EdgeDataView([(0, 1, {}), (1, 2, {}), (2, 3, {'weight': 5})])
>>> G.edges.data('weight', default=1)
EdgeDataView([(0, 1, 1), (1, 2, 1), (2, 3, 5)])
>>> G.edges([0, 3]) # only edges incident to these nodes
EdgeDataView([(0, 1), (3, 2)])
>>> G.edges(0) # only edges incident to a single node (use G.adj[0]?)
EdgeDataView([(0, 1)])

networkx.Graph.has_edge

Graph.has_edge(u, v)
Returns True if the edge (u, v) is in the graph.
This is the same as v in G[u] without KeyError exceptions.
Parameters u, v (nodes) – Nodes can be, for example, strings or numbers. Nodes must be hashable
(and not None) Python objects.
Returns edge_ind – True if edge is in the graph, False otherwise.
Return type bool

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.has_edge(0, 1) # using two nodes
True
>>> e = (0, 1)
>>> G.has_edge(*e) # e is a 2-tuple (u, v)
True
>>> e = (0, 1, {'weight':7})
>>> G.has_edge(*e[:2]) # e is a 3-tuple (u, v, data_dictionary)
True

The following syntax are equivalent:

>>> G.has_edge(0, 1)
True
>>> 1 in G[0] # though this gives KeyError if 0 not in G
True

2.2. Basic graph types 23


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.Graph.get_edge_data

Graph.get_edge_data(u, v, default=None)
Returns the attribute dictionary associated with edge (u, v).
This is identical to G[u][v] except the default is returned instead of an exception is the edge doesn’t exist.
Parameters
• u, v (nodes)
• default (any Python object (default=None)) – Value to return if the edge (u, v) is not found.
Returns edge_dict – The edge attribute dictionary.
Return type dictionary

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G[0][1]
{}

Warning: Assigning to G[u][v] is not permitted. But it is safe to assign attributes G[u][v]['foo']

>>> G[0][1]['weight'] = 7
>>> G[0][1]['weight']
7
>>> G[1][0]['weight']
7

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.get_edge_data(0, 1) # default edge data is {}
{}
>>> e = (0, 1)
>>> G.get_edge_data(*e) # tuple form
{}
>>> G.get_edge_data('a', 'b', default=0) # edge not in graph, return 0
0

networkx.Graph.neighbors

Graph.neighbors(n)
Returns an iterator over all neighbors of node n.
This is identical to iter(G[n])
Parameters n (node) – A node in the graph
Returns neighbors – An iterator over all neighbors of node n
Return type iterator
Raises NetworkXError – If the node n is not in the graph.

24 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> [n for n in G.neighbors(0)]
[1]

Notes

It is usually more convenient (and faster) to access the adjacency dictionary as G[n]:

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_edge('a', 'b', weight=7)
>>> G['a']
AtlasView({'b': {'weight': 7}})
>>> G = nx.path_graph(4)
>>> [n for n in G[0]]
[1]

networkx.Graph.adj

Graph.adj
Graph adjacency object holding the neighbors of each node.
This object is a read-only dict-like structure with node keys and neighbor-dict values. The neighbor-dict is keyed
by neighbor to the edge-data-dict. So G.adj[3][2]['color'] = 'blue' sets the color of the edge (3,
2) to "blue".
Iterating over G.adj behaves like a dict. Useful idioms include for nbr, datadict in G.adj[n].
items():.
The neighbor information is also provided by subscripting the graph. So for nbr, foovalue in
G[node].data('foo', default=1): works.
For directed graphs, G.adj holds outgoing (successor) info.

networkx.Graph.__getitem__

Graph.__getitem__(n)
Returns a dict of neighbors of node n. Use: ‘G[n]’.
Parameters n (node) – A node in the graph.
Returns adj_dict – The adjacency dictionary for nodes connected to n.
Return type dictionary

Notes

G[n] is the same as G.adj[n] and similar to G.neighbors(n) (which is an iterator over G.adj[n])

2.2. Basic graph types 25


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G[0]
AtlasView({1: {}})

networkx.Graph.adjacency

Graph.adjacency()
Returns an iterator over (node, adjacency dict) tuples for all nodes.
For directed graphs, only outgoing neighbors/adjacencies are included.
Returns adj_iter – An iterator over (node, adjacency dictionary) for all nodes in the graph.
Return type iterator

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> [(n, nbrdict) for n, nbrdict in G.adjacency()]
[(0, {1: {}}), (1, {0: {}, 2: {}}), (2, {1: {}, 3: {}}), (3, {2: {}})]

networkx.Graph.nbunch_iter

Graph.nbunch_iter(nbunch=None)
Returns an iterator over nodes contained in nbunch that are also in the graph.
The nodes in nbunch are checked for membership in the graph and if not are silently ignored.
Parameters nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
Returns niter – An iterator over nodes in nbunch that are also in the graph. If nbunch is None,
iterate over all nodes in the graph.
Return type iterator
Raises NetworkXError – If nbunch is not a node or or sequence of nodes. If a node in nbunch is
not hashable.
See also:
Graph.__iter__()

Notes

When nbunch is an iterator, the returned iterator yields values directly from nbunch, becoming exhausted when
nbunch is exhausted.
To test whether nbunch is a single node, one can use “if nbunch in self:”, even after processing with this routine.
If nbunch is not a node or a (possibly empty) sequence/iterator or None, a NetworkXError is raised. Also, if
any object in nbunch is not hashable, a NetworkXError is raised.

26 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Counting nodes edges and neighbors

Graph.order() Returns the number of nodes in the graph.


Graph.number_of_nodes() Returns the number of nodes in the graph.
Graph.__len__() Returns the number of nodes.
Graph.degree A DegreeView for the Graph as G.degree or G.degree().
Graph.size([weight]) Returns the number of edges or total of all edge weights.
Graph.number_of_edges([u, v]) Returns the number of edges between two nodes.

networkx.Graph.order

Graph.order()
Returns the number of nodes in the graph.
Returns nnodes – The number of nodes in the graph.
Return type int
See also:
number_of_nodes(), __len__()

networkx.Graph.number_of_nodes

Graph.number_of_nodes()
Returns the number of nodes in the graph.
Returns nnodes – The number of nodes in the graph.
Return type int
See also:
order(), __len__()

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> len(G)
3

networkx.Graph.__len__

Graph.__len__()
Returns the number of nodes. Use: ‘len(G)’.
Returns nnodes – The number of nodes in the graph.
Return type int

2.2. Basic graph types 27


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> len(G)
4

networkx.Graph.degree

Graph.degree
A DegreeView for the Graph as G.degree or G.degree().
The node degree is the number of edges adjacent to the node. The weighted node degree is the sum of the edge
weights for edges incident to that node.
This object provides an iterator for (node, degree) as well as lookup for the degree for a single node.
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• weight (string or None, optional (default=None)) – The name of an edge attribute that holds
the numerical value used as a weight. If None, then each edge has weight 1. The degree is
the sum of the edge weights adjacent to the node.
Returns
• If a single node is requested
• deg (int) – Degree of the node
• OR if multiple nodes are requested
• nd_view (A DegreeView object capable of iterating (node, degree) pairs)

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.degree[0] # node 0 has degree 1
1
>>> list(G.degree([0, 1, 2]))
[(0, 1), (1, 2), (2, 2)]

networkx.Graph.size

Graph.size(weight=None)
Returns the number of edges or total of all edge weights.
Parameters weight (string or None, optional (default=None)) – The edge attribute that holds the
numerical value used as a weight. If None, then each edge has weight 1.
Returns
size – The number of edges or (if weight keyword is provided) the total weight sum.
If weight is None, returns an int. Otherwise a float (or more general numeric if the weights are
more general).

28 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type numeric


See also:
number_of_edges()

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.size()
3

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_edge('a', 'b', weight=2)
>>> G.add_edge('b', 'c', weight=4)
>>> G.size()
2
>>> G.size(weight='weight')
6.0

networkx.Graph.number_of_edges

Graph.number_of_edges(u=None, v=None)
Returns the number of edges between two nodes.
Parameters u, v (nodes, optional (default=all edges)) – If u and v are specified, return the number
of edges between u and v. Otherwise return the total number of all edges.
Returns nedges – The number of edges in the graph. If nodes u and v are specified return the
number of edges between those nodes. If the graph is directed, this only returns the number of
edges from u to v.
Return type int
See also:
size()

Examples

For undirected graphs, this method counts the total number of edges in the graph:

>>> G = nx.path_graph(4)
>>> G.number_of_edges()
3

If you specify two nodes, this counts the total number of edges joining the two nodes:

>>> G.number_of_edges(0, 1)
1

For directed graphs, this method can count the total number of directed edges from u to v:

2.2. Basic graph types 29


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.DiGraph()
>>> G.add_edge(0, 1)
>>> G.add_edge(1, 0)
>>> G.number_of_edges(0, 1)
1

Making copies and subgraphs

Graph.copy([as_view]) Returns a copy of the graph.


Graph.to_undirected([as_view]) Returns an undirected copy of the graph.
Graph.to_directed([as_view]) Returns a directed representation of the graph.
Graph.subgraph(nodes) Returns a SubGraph view of the subgraph induced on
nodes.
Graph.edge_subgraph(edges) Returns the subgraph induced by the specified edges.

networkx.Graph.copy

Graph.copy(as_view=False)
Returns a copy of the graph.
The copy method by default returns an independent shallow copy of the graph and attributes. That is, if an
attribute is a container, that container is shared by the original an the copy. Use Python’s copy.deepcopy for
new containers.
If as_view is True then a view is returned instead of a copy.

Notes

All copies reproduce the graph structure, but data attributes may be handled in different ways. There are four
types of copies of a graph that people might want.
Deepcopy – A “deepcopy” copies the graph structure as well as all data attributes and any objects they might
contain. The entire graph object is new so that changes in the copy do not affect the original object. (see Python’s
copy.deepcopy)
Data Reference (Shallow) – For a shallow copy the graph structure is copied but the edge, node and graph at-
tribute dicts are references to those in the original graph. This saves time and memory but could cause confusion
if you change an attribute in one graph and it changes the attribute in the other. NetworkX does not provide this
level of shallow copy.
Independent Shallow – This copy creates new independent attribute dicts and then does a shallow copy of the
attributes. That is, any attributes that are containers are shared between the new graph and the original. This is
exactly what dict.copy() provides. You can obtain this style copy using:

>>> G = nx.path_graph(5)
>>> H = G.copy()
>>> H = G.copy(as_view=False)
>>> H = nx.Graph(G)
>>> H = G.__class__(G)

Fresh Data – For fresh data, the graph structure is copied while new empty data attribute dicts are created. The
resulting graph is independent of the original and it has no edge, node or graph attributes. Fresh copies are not
enabled. Instead use:

30 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> H = G.__class__()
>>> H.add_nodes_from(G)
>>> H.add_edges_from(G.edges)

View – Inspired by dict-views, graph-views act like read-only versions of the original graph, providing a copy
of the original structure without requiring any memory for copying the information.
See the Python copy module for more information on shallow and deep copies, https://docs.python.org/2/library/
copy.html.
Parameters as_view (bool, optional (default=False)) – If True, the returned graph-view provides a
read-only view of the original graph without actually copying any data.
Returns G – A copy of the graph.
Return type Graph
See also:

to_directed() return a directed copy of the graph.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> H = G.copy()

networkx.Graph.to_undirected

Graph.to_undirected(as_view=False)
Returns an undirected copy of the graph.
Parameters as_view (bool (optional, default=False)) – If True return a view of the original undi-
rected graph.
Returns G – A deepcopy of the graph.
Return type Graph/MultiGraph
See also:
Graph(), copy(), add_edge(), add_edges_from()

Notes

This returns a “deepcopy” of the edge, node, and graph attributes which attempts to completely copy all of the
data and references.
This is in contrast to the similar G = nx.DiGraph(D) which returns a shallow copy of the data.
See the Python copy module for more information on shallow and deep copies, https://docs.python.org/2/library/
copy.html.
Warning: If you have subclassed DiGraph to use dict-like objects in the data structure, those changes do not
transfer to the Graph created by this method.

2.2. Basic graph types 31


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(2) # or MultiGraph, etc


>>> H = G.to_directed()
>>> list(H.edges)
[(0, 1), (1, 0)]
>>> G2 = H.to_undirected()
>>> list(G2.edges)
[(0, 1)]

networkx.Graph.to_directed

Graph.to_directed(as_view=False)
Returns a directed representation of the graph.
Returns G – A directed graph with the same name, same nodes, and with each edge (u, v, data)
replaced by two directed edges (u, v, data) and (v, u, data).
Return type DiGraph

Notes

This returns a “deepcopy” of the edge, node, and graph attributes which attempts to completely copy all of the
data and references.
This is in contrast to the similar D=DiGraph(G) which returns a shallow copy of the data.
See the Python copy module for more information on shallow and deep copies, https://docs.python.org/2/library/
copy.html.
Warning: If you have subclassed Graph to use dict-like objects in the data structure, those changes do not
transfer to the DiGraph created by this method.

Examples

>>> G = nx.Graph() # or MultiGraph, etc


>>> G.add_edge(0, 1)
>>> H = G.to_directed()
>>> list(H.edges)
[(0, 1), (1, 0)]

If already directed, return a (deep) copy

>>> G = nx.DiGraph() # or MultiDiGraph, etc


>>> G.add_edge(0, 1)
>>> H = G.to_directed()
>>> list(H.edges)
[(0, 1)]

networkx.Graph.subgraph

Graph.subgraph(nodes)
Returns a SubGraph view of the subgraph induced on nodes.

32 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

The induced subgraph of the graph contains the nodes in nodes and the edges between those nodes.
Parameters nodes (list, iterable) – A container of nodes which will be iterated through once.
Returns G – A subgraph view of the graph. The graph structure cannot be changed but node/edge
attributes can and are shared with the original graph.
Return type SubGraph View

Notes

The graph, edge and node attributes are shared with the original graph. Changes to the graph structure is ruled
out by the view, but changes to attributes are reflected in the original graph.
To create a subgraph with its own copy of the edge/node attributes use: G.subgraph(nodes).copy()
For an inplace reduction of a graph to a subgraph you can remove nodes: G.remove_nodes_from([n for n in G
if n not in set(nodes)])
Subgraph views are sometimes NOT what you want. In most cases where you want to do more than simply look
at the induced edges, it makes more sense to just create the subgraph as its own graph with code like:

# Create a subgraph SG based on a (possibly multigraph) G


SG = G.__class__()
SG.add_nodes_from((n, G.nodes[n]) for n in largest_wcc)
if SG.is_multigraph:
SG.add_edges_from((n, nbr, key, d)
for n, nbrs in G.adj.items() if n in largest_wcc
for nbr, keydict in nbrs.items() if nbr in largest_wcc
for key, d in keydict.items())
else:
SG.add_edges_from((n, nbr, d)
for n, nbrs in G.adj.items() if n in largest_wcc
for nbr, d in nbrs.items() if nbr in largest_wcc)
SG.graph.update(G.graph)

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> H = G.subgraph([0, 1, 2])
>>> list(H.edges)
[(0, 1), (1, 2)]

networkx.Graph.edge_subgraph

Graph.edge_subgraph(edges)
Returns the subgraph induced by the specified edges.
The induced subgraph contains each edge in edges and each node incident to any one of those edges.
Parameters edges (iterable) – An iterable of edges in this graph.
Returns G – An edge-induced subgraph of this graph with the same edge attributes.
Return type Graph

2.2. Basic graph types 33


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The graph, edge, and node attributes in the returned subgraph view are references to the corresponding attributes
in the original graph. The view is read-only.
To create a full graph version of the subgraph with its own copy of the edge or node attributes, use:

>>> G.edge_subgraph(edges).copy()

Examples

>>> G = nx.path_graph(5)
>>> H = G.edge_subgraph([(0, 1), (3, 4)])
>>> list(H.nodes)
[0, 1, 3, 4]
>>> list(H.edges)
[(0, 1), (3, 4)]

2.2.2 DiGraph—Directed graphs with self loops

Overview

class DiGraph(incoming_graph_data=None, **attr)


Base class for directed graphs.
A DiGraph stores nodes and edges with optional data, or attributes.
DiGraphs hold directed edges. Self loops are allowed but multiple (parallel) edges are not.
Nodes can be arbitrary (hashable) Python objects with optional key/value attributes. By convention None is not
used as a node.
Edges are represented as links between nodes with optional key/value attributes.
Parameters
• incoming_graph_data (input graph (optional, default: None)) – Data to initialize graph. If
None (default) an empty graph is created. The data can be any format that is supported by
the to_networkx_graph() function, currently including edge list, dict of dicts, dict of lists,
NetworkX graph, NumPy matrix or 2d ndarray, SciPy sparse matrix, or PyGraphviz graph.
• attr (keyword arguments, optional (default= no attributes)) – Attributes to add to graph as
key=value pairs.
See also:
Graph, MultiGraph, MultiDiGraph, OrderedDiGraph

Examples

Create an empty graph structure (a “null graph”) with no nodes and no edges.

>>> G = nx.DiGraph()

34 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

G can be grown in several ways.


Nodes:
Add one node at a time:

>>> G.add_node(1)

Add the nodes from any container (a list, dict, set or even the lines from a file or the nodes from another graph).

>>> G.add_nodes_from([2, 3])


>>> G.add_nodes_from(range(100, 110))
>>> H = nx.path_graph(10)
>>> G.add_nodes_from(H)

In addition to strings and integers any hashable Python object (except None) can represent a node, e.g. a
customized node object, or even another Graph.

>>> G.add_node(H)

Edges:
G can also be grown by adding edges.
Add one edge,

>>> G.add_edge(1, 2)

a list of edges,

>>> G.add_edges_from([(1, 2), (1, 3)])

or a collection of edges,

>>> G.add_edges_from(H.edges)

If some edges connect nodes not yet in the graph, the nodes are added automatically. There are no errors when
adding nodes or edges that already exist.
Attributes:
Each graph, node, and edge can hold key/value attribute pairs in an associated attribute dictionary (the keys
must be hashable). By default these are empty, but can be added or changed using add_edge, add_node or direct
manipulation of the attribute dictionaries named graph, node and edge respectively.

>>> G = nx.DiGraph(day="Friday")
>>> G.graph
{'day': 'Friday'}

Add node attributes using add_node(), add_nodes_from() or G.nodes

>>> G.add_node(1, time='5pm')


>>> G.add_nodes_from([3], time='2pm')
>>> G.nodes[1]
{'time': '5pm'}
>>> G.nodes[1]['room'] = 714
>>> del G.nodes[1]['room'] # remove attribute
>>> list(G.nodes(data=True))
[(1, {'time': '5pm'}), (3, {'time': '2pm'})]

2.2. Basic graph types 35


NetworkX Reference, Release 2.3rc1.dev20190222214247

Add edge attributes using add_edge(), add_edges_from(), subscript notation, or G.edges.

>>> G.add_edge(1, 2, weight=4.7 )


>>> G.add_edges_from([(3, 4), (4, 5)], color='red')
>>> G.add_edges_from([(1, 2, {'color':'blue'}), (2, 3, {'weight':8})])
>>> G[1][2]['weight'] = 4.7
>>> G.edges[1, 2]['weight'] = 4

Warning: we protect the graph data structure by making G.edges[1, 2] a read-only dict-like structure.
However, you can assign to attributes in e.g. G.edges[1, 2]. Thus, use 2 sets of brackets to add/change data
attributes: G.edges[1, 2]['weight'] = 4 (For multigraphs: MG.edges[u, v, key][name] =
value).
Shortcuts:
Many common graph features allow python syntax to speed reporting.

>>> 1 in G # check if node in graph


True
>>> [n for n in G if n < 3] # iterate through nodes
[1, 2]
>>> len(G) # number of nodes in graph
5

Often the best way to traverse all edges of a graph is via the neighbors. The neighbors are reported as an
adjacency-dict G.adj or G.adjacency()

>>> for n, nbrsdict in G.adjacency():


... for nbr, eattr in nbrsdict.items():
... if 'weight' in eattr:
... # Do something useful with the edges
... pass

But the edges reporting object is often more convenient:

>>> for u, v, weight in G.edges(data='weight'):


... if weight is not None:
... # Do something useful with the edges
... pass

Reporting:
Simple graph information is obtained using object-attributes and methods. Reporting usually provides
views instead of containers to reduce memory usage. The views update as the graph is updated simi-
larly to dict-views. The objects nodes, `edges and adj provide access to data attributes via lookup
(e.g. nodes[n], `edges[u, v], adj[u][v]) and iteration (e.g. nodes.items(), nodes.
data('color'), nodes.data('color', default='blue') and similarly for edges) Views exist
for nodes, edges, neighbors()/adj and degree.
For details on these and other miscellaneous methods, see below.
Subclasses (Advanced):
The Graph class uses a dict-of-dict-of-dict data structure. The outer dict (node_dict) holds adjacency information
keyed by node. The next dict (adjlist_dict) represents the adjacency information and holds edge data keyed by
neighbor. The inner dict (edge_attr_dict) represents the edge data and holds edge attribute values keyed by
attribute names.
Each of these three dicts can be replaced in a subclass by a user defined dict-like object. In general, the
dict-like features should be maintained but extra features can be added. To replace one of the dicts create

36 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

a new graph class by changing the class(!) variable holding the factory for that dict-like structure. The vari-
able names are node_dict_factory, node_attr_dict_factory, adjlist_inner_dict_factory, adjlist_outer_dict_factory,
edge_attr_dict_factory and graph_attr_dict_factory.
node_dict_factory [function, (default: dict)] Factory function to be used to create the dict containing node
attributes, keyed by node id. It should require no arguments and return a dict-like object
node_attr_dict_factory: function, (default: dict) Factory function to be used to create the node attribute dict
which holds attribute values keyed by attribute name. It should require no arguments and return a dict-like
object
adjlist_outer_dict_factory [function, (default: dict)] Factory function to be used to create the outer-most dict
in the data structure that holds adjacency info keyed by node. It should require no arguments and return a
dict-like object.
adjlist_inner_dict_factory [function, optional (default: dict)] Factory function to be used to create the adja-
cency list dict which holds edge data keyed by neighbor. It should require no arguments and return a
dict-like object
edge_attr_dict_factory [function, optional (default: dict)] Factory function to be used to create the edge at-
tribute dict which holds attribute values keyed by attribute name. It should require no arguments and return
a dict-like object.
graph_attr_dict_factory [function, (default: dict)] Factory function to be used to create the graph attribute
dict which holds attribute values keyed by attribute name. It should require no arguments and return a
dict-like object.
Typically, if your extension doesn’t impact the data structure all methods will inherited without issue except:
to_directed/to_undirected. By default these methods create a DiGraph/Graph class and you probably
want them to create your extension of a DiGraph/Graph. To facilitate this we define two class variables that you
can set in your subclass.
to_directed_class [callable, (default: DiGraph or MultiDiGraph)] Class to create a new graph structure in the
to_directed method. If None, a NetworkX class (DiGraph or MultiDiGraph) is used.
to_undirected_class [callable, (default: Graph or MultiGraph)] Class to create a new graph structure in the
to_undirected method. If None, a NetworkX class (Graph or MultiGraph) is used.

Examples

Create a low memory graph class that effectively disallows edge attributes by using a single attribute dict for all
edges. This reduces the memory used, but you lose edge attributes.

>>> class ThinGraph(nx.Graph):


... all_edge_dict = {'weight': 1}
... def single_edge_dict(self):
... return self.all_edge_dict
... edge_attr_dict_factory = single_edge_dict
>>> G = ThinGraph()
>>> G.add_edge(2, 1)
>>> G[2][1]
{'weight': 1}
>>> G.add_edge(2, 2)
>>> G[2][1] is G[2][2]
True

Please see ordered for more examples of creating graph subclasses by overwriting the base class dict with
a dictionary-like object.

2.2. Basic graph types 37


NetworkX Reference, Release 2.3rc1.dev20190222214247

Methods

Adding and removing nodes and edges

DiGraph.__init__([incoming_graph_data]) Initialize a graph with edges, name, or graph attributes.


DiGraph.add_node(node_for_adding, **attr) Add a single node node_for_adding and update
node attributes.
DiGraph.add_nodes_from(nodes_for_adding, Add multiple nodes.
**attr)
DiGraph.remove_node(n) Remove node n.
DiGraph.remove_nodes_from(nodes) Remove multiple nodes.
DiGraph.add_edge(u_of_edge, v_of_edge, **attr) Add an edge between u and v.
DiGraph.add_edges_from(ebunch_to_add, Add all the edges in ebunch_to_add.
**attr)
DiGraph.add_weighted_edges_from(ebunch_to_add) Add weighted edges in ebunch_to_add with speci-
fied weight attr
DiGraph.remove_edge(u, v) Remove the edge between u and v.
DiGraph.remove_edges_from(ebunch) Remove all edges specified in ebunch.
DiGraph.update([edges, nodes]) Update the graph using nodes/edges/graphs as input.
DiGraph.clear() Remove all nodes and edges from the graph.

networkx.DiGraph.__init__

DiGraph.__init__(incoming_graph_data=None, **attr)
Initialize a graph with edges, name, or graph attributes.
Parameters
• incoming_graph_data (input graph (optional, default: None)) – Data to initialize graph. If
None (default) an empty graph is created. The data can be an edge list, or any NetworkX
graph object. If the corresponding optional Python packages are installed the data can also
be a NumPy matrix or 2d ndarray, a SciPy sparse matrix, or a PyGraphviz graph.
• attr (keyword arguments, optional (default= no attributes)) – Attributes to add to graph as
key=value pairs.
See also:
convert()

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G = nx.Graph(name='my graph')
>>> e = [(1, 2), (2, 3), (3, 4)] # list of edges
>>> G = nx.Graph(e)

Arbitrary graph attribute pairs (key=value) may be assigned

>>> G = nx.Graph(e, day="Friday")


>>> G.graph
{'day': 'Friday'}

38 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.DiGraph.add_node

DiGraph.add_node(node_for_adding, **attr)
Add a single node node_for_adding and update node attributes.
Parameters
• node_for_adding (node) – A node can be any hashable Python object except None.
• attr (keyword arguments, optional) – Set or change node attributes using key=value.
See also:
add_nodes_from()

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_node(1)
>>> G.add_node('Hello')
>>> K3 = nx.Graph([(0, 1), (1, 2), (2, 0)])
>>> G.add_node(K3)
>>> G.number_of_nodes()
3

Use keywords set/change node attributes:

>>> G.add_node(1, size=10)


>>> G.add_node(3, weight=0.4, UTM=('13S', 382871, 3972649))

Notes

A hashable object is one that can be used as a key in a Python dictionary. This includes strings, numbers, tuples
of strings and numbers, etc.
On many platforms hashable items also include mutables such as NetworkX Graphs, though one should be
careful that the hash doesn’t change on mutables.

networkx.DiGraph.add_nodes_from

DiGraph.add_nodes_from(nodes_for_adding, **attr)
Add multiple nodes.
Parameters
• nodes_for_adding (iterable container) – A container of nodes (list, dict, set, etc.). OR A
container of (node, attribute dict) tuples. Node attributes are updated using the attribute dict.
• attr (keyword arguments, optional (default= no attributes)) – Update attributes for all nodes
in nodes. Node attributes specified in nodes as a tuple take precedence over attributes spec-
ified via keyword arguments.
See also:
add_node()

2.2. Basic graph types 39


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_nodes_from('Hello')
>>> K3 = nx.Graph([(0, 1), (1, 2), (2, 0)])
>>> G.add_nodes_from(K3)
>>> sorted(G.nodes(), key=str)
[0, 1, 2, 'H', 'e', 'l', 'o']

Use keywords to update specific node attributes for every node.

>>> G.add_nodes_from([1, 2], size=10)


>>> G.add_nodes_from([3, 4], weight=0.4)

Use (node, attrdict) tuples to update attributes for specific nodes.

>>> G.add_nodes_from([(1, dict(size=11)), (2, {'color':'blue'})])


>>> G.nodes[1]['size']
11
>>> H = nx.Graph()
>>> H.add_nodes_from(G.nodes(data=True))
>>> H.nodes[1]['size']
11

networkx.DiGraph.remove_node

DiGraph.remove_node(n)
Remove node n.
Removes the node n and all adjacent edges. Attempting to remove a non-existent node will raise an exception.
Parameters n (node) – A node in the graph
Raises NetworkXError – If n is not in the graph.
See also:
remove_nodes_from()

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> list(G.edges)
[(0, 1), (1, 2)]
>>> G.remove_node(1)
>>> list(G.edges)
[]

networkx.DiGraph.remove_nodes_from

DiGraph.remove_nodes_from(nodes)
Remove multiple nodes.
Parameters nodes (iterable container) – A container of nodes (list, dict, set, etc.). If a node in the
container is not in the graph it is silently ignored.

40 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
remove_node()

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> e = list(G.nodes)
>>> e
[0, 1, 2]
>>> G.remove_nodes_from(e)
>>> list(G.nodes)
[]

networkx.DiGraph.add_edge

DiGraph.add_edge(u_of_edge, v_of_edge, **attr)


Add an edge between u and v.
The nodes u and v will be automatically added if they are not already in the graph.
Edge attributes can be specified with keywords or by directly accessing the edge’s attribute dictionary. See
examples below.
Parameters
• u, v (nodes) – Nodes can be, for example, strings or numbers. Nodes must be hashable (and
not None) Python objects.
• attr (keyword arguments, optional) – Edge data (or labels or objects) can be assigned using
keyword arguments.
See also:

add_edges_from() add a collection of edges

Notes

Adding an edge that already exists updates the edge data.


Many NetworkX algorithms designed for weighted graphs use an edge attribute (by default weight) to hold a
numerical value.

Examples

The following all add the edge e=(1, 2) to graph G:

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> e = (1, 2)
>>> G.add_edge(1, 2) # explicit two-node form
>>> G.add_edge(*e) # single edge as tuple of two nodes
>>> G.add_edges_from( [(1, 2)] ) # add edges from iterable container

Associate data to edges using keywords:

2.2. Basic graph types 41


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G.add_edge(1, 2, weight=3)


>>> G.add_edge(1, 3, weight=7, capacity=15, length=342.7)

For non-string attribute keys, use subscript notation.

>>> G.add_edge(1, 2)
>>> G[1][2].update({0: 5})
>>> G.edges[1, 2].update({0: 5})

networkx.DiGraph.add_edges_from

DiGraph.add_edges_from(ebunch_to_add, **attr)
Add all the edges in ebunch_to_add.
Parameters
• ebunch_to_add (container of edges) – Each edge given in the container will be added to the
graph. The edges must be given as 2-tuples (u, v) or 3-tuples (u, v, d) where d is a dictionary
containing edge data.
• attr (keyword arguments, optional) – Edge data (or labels or objects) can be assigned using
keyword arguments.
See also:

add_edge() add a single edge


add_weighted_edges_from() convenient way to add weighted edges

Notes

Adding the same edge twice has no effect but any edge data will be updated when each duplicate edge is added.
Edge attributes specified in an ebunch take precedence over attributes specified via keyword arguments.

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_edges_from([(0, 1), (1, 2)]) # using a list of edge tuples
>>> e = zip(range(0, 3), range(1, 4))
>>> G.add_edges_from(e) # Add the path graph 0-1-2-3

Associate data to edges

>>> G.add_edges_from([(1, 2), (2, 3)], weight=3)


>>> G.add_edges_from([(3, 4), (1, 4)], label='WN2898')

networkx.DiGraph.add_weighted_edges_from

DiGraph.add_weighted_edges_from(ebunch_to_add, weight=’weight’, **attr)


Add weighted edges in ebunch_to_add with specified weight attr
Parameters

42 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

• ebunch_to_add (container of edges) – Each edge given in the list or container will be added
to the graph. The edges must be given as 3-tuples (u, v, w) where w is a number.
• weight (string, optional (default= ‘weight’)) – The attribute name for the edge weights to
be added.
• attr (keyword arguments, optional (default= no attributes)) – Edge attributes to add/update
for all edges.
See also:

add_edge() add a single edge


add_edges_from() add multiple edges

Notes

Adding the same edge twice for Graph/DiGraph simply updates the edge data. For MultiGraph/MultiDiGraph,
duplicate edges are stored.

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_weighted_edges_from([(0, 1, 3.0), (1, 2, 7.5)])

networkx.DiGraph.remove_edge

DiGraph.remove_edge(u, v)
Remove the edge between u and v.
Parameters u, v (nodes) – Remove the edge between nodes u and v.
Raises NetworkXError – If there is not an edge between u and v.
See also:

remove_edges_from() remove a collection of edges

Examples

>>> G = nx.Graph() # or DiGraph, etc


>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.remove_edge(0, 1)
>>> e = (1, 2)
>>> G.remove_edge(*e) # unpacks e from an edge tuple
>>> e = (2, 3, {'weight':7}) # an edge with attribute data
>>> G.remove_edge(*e[:2]) # select first part of edge tuple

networkx.DiGraph.remove_edges_from

DiGraph.remove_edges_from(ebunch)
Remove all edges specified in ebunch.

2.2. Basic graph types 43


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters ebunch (list or container of edge tuples) – Each edge given in the list or container will
be removed from the graph. The edges can be:
• 2-tuples (u, v) edge between u and v.
• 3-tuples (u, v, k) where k is ignored.
See also:

remove_edge() remove a single edge

Notes

Will fail silently if an edge in ebunch is not in the graph.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> ebunch = [(1, 2), (2, 3)]
>>> G.remove_edges_from(ebunch)

networkx.DiGraph.update

DiGraph.update(edges=None, nodes=None)
Update the graph using nodes/edges/graphs as input.
Like dict.update, this method takes a graph as input, adding the graph’s noes and edges to this graph. It can also
take two inputs: edges and nodes. Finally it can take either edges or nodes. To specify only nodes the keyword
nodes must be used.
The collections of edges and nodes are treated similarly to the add_edges_from/add_nodes_from methods.
When iterated, they should yield 2-tuples (u, v) or 3-tuples (u, v, datadict).
Parameters
• edges (Graph object, collection of edges, or None) – The first parameter can be a graph or
some edges. If it has attributes nodes and edges, then it is taken to be a Graph-like object
and those attributes are used as collections of nodes and edges to be added to the graph. If
the first parameter does not have those attributes, it is treated as a collection of edges and
added to the graph. If the first argument is None, no edges are added.
• nodes (collection of nodes, or None) – The second parameter is treated as a collection of
nodes to be added to the graph unless it is None. If edges is None and nodes is
None an exception is raised. If the first parameter is a Graph, then nodes is ignored.

Examples

>>> G = nx.path_graph(5)
>>> G.update(nx.complete_graph(range(4,10)))
>>> from itertools import combinations
>>> edges = ((u, v, {'power': u * v})
... for u, v in combinations(range(10, 20), 2)
... if u * v < 225)
(continues on next page)

44 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> nodes = [1000] # for singleton, use a container
>>> G.update(edges, nodes)

Notes

It you want to update the graph using an adjacency structure it is straightforward to obtain the edges/nodes from
adjacency. The following examples provide common cases, your adjacency may be slightly different and require
tweaks of these examples.

>>> # dict-of-set/list/tuple
>>> adj = {1: {2, 3}, 2: {1, 3}, 3: {1, 2}}
>>> e = [(u, v) for u, nbrs in adj.items() for v in nbrs]
>>> G.update(edges=e, nodes=adj)

>>> DG = nx.DiGraph()
>>> # dict-of-dict-of-attribute
>>> adj = {1: {2: 1.3, 3: 0.7}, 2: {1: 1.4}, 3: {1: 0.7}}
>>> e = [(u, v, {'weight': d}) for u, nbrs in adj.items()
... for v, d in nbrs.items()]
>>> DG.update(edges=e, nodes=adj)

>>> # dict-of-dict-of-dict
>>> adj = {1: {2: {'weight': 1.3}, 3: {'color': 0.7, 'weight':1.2}}}
>>> e = [(u, v, {'weight': d}) for u, nbrs in adj.items()
... for v, d in nbrs.items()]
>>> DG.update(edges=e, nodes=adj)

>>> # predecessor adjacency (dict-of-set)


>>> pred = {1: {2, 3}, 2: {3}, 3: {3}}
>>> e = [(v, u) for u, nbrs in pred.items() for v in nbrs]

>>> # MultiGraph dict-of-dict-of-dict-of-attribute


>>> MDG = nx.MultiDiGraph()
>>> adj = {1: {2: {0: {'weight': 1.3}, 1: {'weight': 1.2}}},
... 3: {2: {0: {'weight': 0.7}}}}
>>> e = [(u, v, ekey, d) for u, nbrs in adj.items()
... for v, keydict in nbrs.items()
... for ekey, d in keydict.items()]
>>> MDG.update(edges=e)

See also:

add_edges_from() add multiple edges to a graph


add_nodes_from() add multiple nodes to a graph

networkx.DiGraph.clear

DiGraph.clear()
Remove all nodes and edges from the graph.
This also removes the name, and all graph, node, and edge attributes.

2.2. Basic graph types 45


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.clear()
>>> list(G.nodes)
[]
>>> list(G.edges)
[]

Reporting nodes edges and neighbors

DiGraph.nodes A NodeView of the Graph as G.nodes or G.nodes().


DiGraph.__iter__() Iterate over the nodes.
DiGraph.has_node(n) Returns True if the graph contains the node n.
DiGraph.__contains__(n) Returns True if n is a node, False otherwise.
DiGraph.edges An OutEdgeView of the DiGraph as G.edges or
G.edges().
DiGraph.out_edges An OutEdgeView of the DiGraph as G.edges or
G.edges().
DiGraph.in_edges An InEdgeView of the Graph as G.in_edges or
G.in_edges().
DiGraph.has_edge(u, v) Returns True if the edge (u, v) is in the graph.
DiGraph.get_edge_data(u, v[, default]) Returns the attribute dictionary associated with edge (u,
v).
DiGraph.neighbors(n) Returns an iterator over successor nodes of n.
DiGraph.adj Graph adjacency object holding the neighbors of each
node.
DiGraph.__getitem__(n) Returns a dict of neighbors of node n.
DiGraph.successors(n) Returns an iterator over successor nodes of n.
DiGraph.succ Graph adjacency object holding the successors of each
node.
DiGraph.predecessors(n) Returns an iterator over predecessor nodes of n.
DiGraph.pred Graph adjacency object holding the predecessors of
each node.
DiGraph.adjacency() Returns an iterator over (node, adjacency dict) tuples for
all nodes.
DiGraph.nbunch_iter([nbunch]) Returns an iterator over nodes contained in nbunch that
are also in the graph.

networkx.DiGraph.nodes

DiGraph.nodes
A NodeView of the Graph as G.nodes or G.nodes().
Can be used as G.nodes for data lookup and for set-like operations. Can also be used as G.
nodes(data='color', default=None) to return a NodeDataView which reports specific node data
but no set operations. It presents a dict-like interface as well with G.nodes.items() iterating over (node,
nodedata) 2-tuples and G.nodes[3]['foo'] providing the value of the foo attribute for node 3. In
addition, a view G.nodes.data('foo') provides a dict-like interface to the foo attribute of each node.
G.nodes.data('foo', default=1) provides a default for nodes that do not have attribute foo.
Parameters

46 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

• data (string or bool, optional (default=False)) – The node attribute returned in 2-tuple (n,
ddict[data]). If True, return entire node attribute dict as (n, ddict). If False, return just the
nodes n.
• default (value, optional (default=None)) – Value used for nodes that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns
Allows set-like operations over the nodes as well as node attribute dict lookup and calling to
get a NodeDataView. A NodeDataView iterates over (n, data) and has no set operations. A
NodeView iterates over n and includes set operations.
When called, if data is False, an iterator over nodes. Otherwise an iterator of 2-tuples (node,
attribute value) where the attribute is specified in data. If data is True then the attribute becomes
the entire data dictionary.
Return type NodeView

Notes

If your node data is not needed, it is simpler and equivalent to use the expression for n in G, or list(G).

Examples

There are two simple ways of getting a list of all nodes in the graph:

>>> G = nx.path_graph(3)
>>> list(G.nodes)
[0, 1, 2]
>>> list(G)
[0, 1, 2]

To get the node data along with the nodes:

>>> G.add_node(1, time='5pm')


>>> G.nodes[0]['foo'] = 'bar'
>>> list(G.nodes(data=True))
[(0, {'foo': 'bar'}), (1, {'time': '5pm'}), (2, {})]
>>> list(G.nodes.data())
[(0, {'foo': 'bar'}), (1, {'time': '5pm'}), (2, {})]

>>> list(G.nodes(data='foo'))
[(0, 'bar'), (1, None), (2, None)]
>>> list(G.nodes.data('foo'))
[(0, 'bar'), (1, None), (2, None)]

>>> list(G.nodes(data='time'))
[(0, None), (1, '5pm'), (2, None)]
>>> list(G.nodes.data('time'))
[(0, None), (1, '5pm'), (2, None)]

>>> list(G.nodes(data='time', default='Not Available'))


[(0, 'Not Available'), (1, '5pm'), (2, 'Not Available')]
>>> list(G.nodes.data('time', default='Not Available'))
[(0, 'Not Available'), (1, '5pm'), (2, 'Not Available')]

2.2. Basic graph types 47


NetworkX Reference, Release 2.3rc1.dev20190222214247

If some of your nodes have an attribute and the rest are assumed to have a default attribute value you can create
a dictionary from node/attribute pairs using the default keyword argument to guarantee the value is never
None:
>>> G = nx.Graph()
>>> G.add_node(0)
>>> G.add_node(1, weight=2)
>>> G.add_node(2, weight=3)
>>> dict(G.nodes(data='weight', default=1))
{0: 1, 1: 2, 2: 3}

networkx.DiGraph.__iter__

DiGraph.__iter__()
Iterate over the nodes. Use: ‘for n in G’.
Returns niter – An iterator over all nodes in the graph.
Return type iterator

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> [n for n in G]
[0, 1, 2, 3]
>>> list(G)
[0, 1, 2, 3]

networkx.DiGraph.has_node

DiGraph.has_node(n)
Returns True if the graph contains the node n.
Identical to n in G
Parameters n (node)

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.has_node(0)
True

It is more readable and simpler to use


>>> 0 in G
True

networkx.DiGraph.__contains__

DiGraph.__contains__(n)
Returns True if n is a node, False otherwise. Use: ‘n in G’.

48 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> 1 in G
True

networkx.DiGraph.edges

DiGraph.edges
An OutEdgeView of the DiGraph as G.edges or G.edges().
edges(self, nbunch=None, data=False, default=None)
The OutEdgeView provides set-like operations on the edge-tuples as well as edge attribute lookup. When called,
it also provides an EdgeDataView object which allows control of access to edge attributes (but does not pro-
vide set-like operations). Hence, G.edges[u, v]['color'] provides the value of the color attribute for
edge (u, v) while for (u, v, c) in G.edges.data('color', default='red'): iterates
through all the edges yielding the color attribute with default 'red' if no color attribute exists.
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• data (string or bool, optional (default=False)) – The edge attribute returned in 3-tuple (u, v,
ddict[data]). If True, return edge attribute dict in 3-tuple (u, v, ddict). If False, return 2-tuple
(u, v).
• default (value, optional (default=None)) – Value used for edges that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns edges – A view of edge attributes, usually it iterates over (u, v) or (u, v, d) tuples of edges,
but can also be used for attribute lookup as edges[u, v]['foo'].
Return type OutEdgeView
See also:
in_edges, out_edges

Notes

Nodes in nbunch that are not in the graph will be (quietly) ignored. For directed graphs this returns the out-edges.

Examples

>>> G = nx.DiGraph() # or MultiDiGraph, etc


>>> nx.add_path(G, [0, 1, 2])
>>> G.add_edge(2, 3, weight=5)
>>> [e for e in G.edges]
[(0, 1), (1, 2), (2, 3)]
>>> G.edges.data() # default data is {} (empty dict)
OutEdgeDataView([(0, 1, {}), (1, 2, {}), (2, 3, {'weight': 5})])
>>> G.edges.data('weight', default=1)
OutEdgeDataView([(0, 1, 1), (1, 2, 1), (2, 3, 5)])
>>> G.edges([0, 2]) # only edges incident to these nodes
(continues on next page)

2.2. Basic graph types 49


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


OutEdgeDataView([(0, 1), (2, 3)])
>>> G.edges(0) # only edges incident to a single node (use G.adj[0]?)
OutEdgeDataView([(0, 1)])

networkx.DiGraph.out_edges

DiGraph.out_edges
An OutEdgeView of the DiGraph as G.edges or G.edges().
edges(self, nbunch=None, data=False, default=None)
The OutEdgeView provides set-like operations on the edge-tuples as well as edge attribute lookup. When called,
it also provides an EdgeDataView object which allows control of access to edge attributes (but does not pro-
vide set-like operations). Hence, G.edges[u, v]['color'] provides the value of the color attribute for
edge (u, v) while for (u, v, c) in G.edges.data('color', default='red'): iterates
through all the edges yielding the color attribute with default 'red' if no color attribute exists.
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• data (string or bool, optional (default=False)) – The edge attribute returned in 3-tuple (u, v,
ddict[data]). If True, return edge attribute dict in 3-tuple (u, v, ddict). If False, return 2-tuple
(u, v).
• default (value, optional (default=None)) – Value used for edges that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns edges – A view of edge attributes, usually it iterates over (u, v) or (u, v, d) tuples of edges,
but can also be used for attribute lookup as edges[u, v]['foo'].
Return type OutEdgeView
See also:
in_edges, out_edges

Notes

Nodes in nbunch that are not in the graph will be (quietly) ignored. For directed graphs this returns the out-edges.

Examples

>>> G = nx.DiGraph() # or MultiDiGraph, etc


>>> nx.add_path(G, [0, 1, 2])
>>> G.add_edge(2, 3, weight=5)
>>> [e for e in G.edges]
[(0, 1), (1, 2), (2, 3)]
>>> G.edges.data() # default data is {} (empty dict)
OutEdgeDataView([(0, 1, {}), (1, 2, {}), (2, 3, {'weight': 5})])
>>> G.edges.data('weight', default=1)
OutEdgeDataView([(0, 1, 1), (1, 2, 1), (2, 3, 5)])
>>> G.edges([0, 2]) # only edges incident to these nodes
OutEdgeDataView([(0, 1), (2, 3)])
(continues on next page)

50 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> G.edges(0) # only edges incident to a single node (use G.adj[0]?)
OutEdgeDataView([(0, 1)])

networkx.DiGraph.in_edges

DiGraph.in_edges
An InEdgeView of the Graph as G.in_edges or G.in_edges().
in_edges(self, nbunch=None, data=False, default=None):
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• data (string or bool, optional (default=False)) – The edge attribute returned in 3-tuple (u, v,
ddict[data]). If True, return edge attribute dict in 3-tuple (u, v, ddict). If False, return 2-tuple
(u, v).
• default (value, optional (default=None)) – Value used for edges that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns in_edges – A view of edge attributes, usually it iterates over (u, v) or (u, v, d) tuples of
edges, but can also be used for attribute lookup as edges[u, v]['foo'].
Return type InEdgeView
See also:
edges

networkx.DiGraph.has_edge

DiGraph.has_edge(u, v)
Returns True if the edge (u, v) is in the graph.
This is the same as v in G[u] without KeyError exceptions.
Parameters u, v (nodes) – Nodes can be, for example, strings or numbers. Nodes must be hashable
(and not None) Python objects.
Returns edge_ind – True if edge is in the graph, False otherwise.
Return type bool

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.has_edge(0, 1) # using two nodes
True
>>> e = (0, 1)
>>> G.has_edge(*e) # e is a 2-tuple (u, v)
True
>>> e = (0, 1, {'weight':7})
>>> G.has_edge(*e[:2]) # e is a 3-tuple (u, v, data_dictionary)
True

2.2. Basic graph types 51


NetworkX Reference, Release 2.3rc1.dev20190222214247

The following syntax are equivalent:

>>> G.has_edge(0, 1)
True
>>> 1 in G[0] # though this gives KeyError if 0 not in G
True

networkx.DiGraph.get_edge_data

DiGraph.get_edge_data(u, v, default=None)
Returns the attribute dictionary associated with edge (u, v).
This is identical to G[u][v] except the default is returned instead of an exception is the edge doesn’t exist.
Parameters
• u, v (nodes)
• default (any Python object (default=None)) – Value to return if the edge (u, v) is not found.
Returns edge_dict – The edge attribute dictionary.
Return type dictionary

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G[0][1]
{}

Warning: Assigning to G[u][v] is not permitted. But it is safe to assign attributes G[u][v]['foo']

>>> G[0][1]['weight'] = 7
>>> G[0][1]['weight']
7
>>> G[1][0]['weight']
7

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.get_edge_data(0, 1) # default edge data is {}
{}
>>> e = (0, 1)
>>> G.get_edge_data(*e) # tuple form
{}
>>> G.get_edge_data('a', 'b', default=0) # edge not in graph, return 0
0

networkx.DiGraph.neighbors

DiGraph.neighbors(n)
Returns an iterator over successor nodes of n.
A successor of n is a node m such that there exists a directed edge from n to m.
Parameters n (node) – A node in the graph

52 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Raises NetworkXError – If n is not in the graph.


See also:
predecessors()

Notes

neighbors() and successors() are the same.

networkx.DiGraph.adj

DiGraph.adj
Graph adjacency object holding the neighbors of each node.
This object is a read-only dict-like structure with node keys and neighbor-dict values. The neighbor-dict is keyed
by neighbor to the edge-data-dict. So G.adj[3][2]['color'] = 'blue' sets the color of the edge (3,
2) to "blue".
Iterating over G.adj behaves like a dict. Useful idioms include for nbr, datadict in G.adj[n].
items():.
The neighbor information is also provided by subscripting the graph. So for nbr, foovalue in
G[node].data('foo', default=1): works.
For directed graphs, G.adj holds outgoing (successor) info.

networkx.DiGraph.__getitem__

DiGraph.__getitem__(n)
Returns a dict of neighbors of node n. Use: ‘G[n]’.
Parameters n (node) – A node in the graph.
Returns adj_dict – The adjacency dictionary for nodes connected to n.
Return type dictionary

Notes

G[n] is the same as G.adj[n] and similar to G.neighbors(n) (which is an iterator over G.adj[n])

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G[0]
AtlasView({1: {}})

networkx.DiGraph.successors

DiGraph.successors(n)
Returns an iterator over successor nodes of n.
A successor of n is a node m such that there exists a directed edge from n to m.

2.2. Basic graph types 53


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters n (node) – A node in the graph


Raises NetworkXError – If n is not in the graph.
See also:
predecessors()

Notes

neighbors() and successors() are the same.

networkx.DiGraph.succ

DiGraph.succ
Graph adjacency object holding the successors of each node.
This object is a read-only dict-like structure with node keys and neighbor-dict values. The neighbor-dict is
keyed by neighbor to the edge-data-dict. So G.succ[3][2]['color'] = 'blue' sets the color of the
edge (3, 2) to "blue".
Iterating over G.succ behaves like a dict. Useful idioms include for nbr, datadict in G.succ[n].
items():. A data-view not provided by dicts also exists: for nbr, foovalue in G.succ[node].
data('foo'): and a default can be set via a default argument to the data method.
The neighbor information is also provided by subscripting the graph. So for nbr, foovalue in
G[node].data('foo', default=1): works.
For directed graphs, G.adj is identical to G.succ.

networkx.DiGraph.predecessors

DiGraph.predecessors(n)
Returns an iterator over predecessor nodes of n.
A predecessor of n is a node m such that there exists a directed edge from m to n.
Parameters n (node) – A node in the graph
Raises NetworkXError – If n is not in the graph.
See also:
successors()

networkx.DiGraph.pred

DiGraph.pred
Graph adjacency object holding the predecessors of each node.
This object is a read-only dict-like structure with node keys and neighbor-dict values. The neighbor-dict is
keyed by neighbor to the edge-data-dict. So G.pred[2][3]['color'] = 'blue' sets the color of the
edge (3, 2) to "blue".
Iterating over G.pred behaves like a dict. Useful idioms include for nbr, datadict in G.pred[n].
items():. A data-view not provided by dicts also exists: for nbr, foovalue in G.pred[node].
data('foo'): A default can be set via a default argument to the data method.

54 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.DiGraph.adjacency

DiGraph.adjacency()
Returns an iterator over (node, adjacency dict) tuples for all nodes.
For directed graphs, only outgoing neighbors/adjacencies are included.
Returns adj_iter – An iterator over (node, adjacency dictionary) for all nodes in the graph.
Return type iterator

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> [(n, nbrdict) for n, nbrdict in G.adjacency()]
[(0, {1: {}}), (1, {0: {}, 2: {}}), (2, {1: {}, 3: {}}), (3, {2: {}})]

networkx.DiGraph.nbunch_iter

DiGraph.nbunch_iter(nbunch=None)
Returns an iterator over nodes contained in nbunch that are also in the graph.
The nodes in nbunch are checked for membership in the graph and if not are silently ignored.
Parameters nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
Returns niter – An iterator over nodes in nbunch that are also in the graph. If nbunch is None,
iterate over all nodes in the graph.
Return type iterator
Raises NetworkXError – If nbunch is not a node or or sequence of nodes. If a node in nbunch is
not hashable.
See also:
Graph.__iter__()

Notes

When nbunch is an iterator, the returned iterator yields values directly from nbunch, becoming exhausted when
nbunch is exhausted.
To test whether nbunch is a single node, one can use “if nbunch in self:”, even after processing with this routine.
If nbunch is not a node or a (possibly empty) sequence/iterator or None, a NetworkXError is raised. Also, if
any object in nbunch is not hashable, a NetworkXError is raised.

Counting nodes edges and neighbors

DiGraph.order() Returns the number of nodes in the graph.


DiGraph.number_of_nodes() Returns the number of nodes in the graph.
DiGraph.__len__() Returns the number of nodes.
Continued on next page

2.2. Basic graph types 55


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 7 – continued from previous page


DiGraph.degree A DegreeView for the Graph as G.degree or G.degree().
DiGraph.in_degree An InDegreeView for (node, in_degree) or in_degree
for single node.
DiGraph.out_degree An OutDegreeView for (node, out_degree)
DiGraph.size([weight]) Returns the number of edges or total of all edge weights.
DiGraph.number_of_edges([u, v]) Returns the number of edges between two nodes.

networkx.DiGraph.order

DiGraph.order()
Returns the number of nodes in the graph.
Returns nnodes – The number of nodes in the graph.
Return type int
See also:
number_of_nodes(), __len__()

networkx.DiGraph.number_of_nodes

DiGraph.number_of_nodes()
Returns the number of nodes in the graph.
Returns nnodes – The number of nodes in the graph.
Return type int
See also:
order(), __len__()

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> len(G)
3

networkx.DiGraph.__len__

DiGraph.__len__()
Returns the number of nodes. Use: ‘len(G)’.
Returns nnodes – The number of nodes in the graph.
Return type int

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> len(G)
4

56 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.DiGraph.degree

DiGraph.degree
A DegreeView for the Graph as G.degree or G.degree().
The node degree is the number of edges adjacent to the node. The weighted node degree is the sum of the edge
weights for edges incident to that node.
This object provides an iterator for (node, degree) as well as lookup for the degree for a single node.
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• weight (string or None, optional (default=None)) – The name of an edge attribute that holds
the numerical value used as a weight. If None, then each edge has weight 1. The degree is
the sum of the edge weights adjacent to the node.
Returns
• If a single node is requested
• deg (int) – Degree of the node
• OR if multiple nodes are requested
• nd_iter (iterator) – The iterator returns two-tuples of (node, degree).
See also:
in_degree, out_degree

Examples

>>> G = nx.DiGraph() # or MultiDiGraph


>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.degree(0) # node 0 with degree 1
1
>>> list(G.degree([0, 1, 2]))
[(0, 1), (1, 2), (2, 2)]

networkx.DiGraph.in_degree

DiGraph.in_degree
An InDegreeView for (node, in_degree) or in_degree for single node.
The node in_degree is the number of edges pointing to the node. The weighted node degree is the sum of the
edge weights for edges incident to that node.
This object provides an iteration over (node, in_degree) as well as lookup for the degree for a single node.
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• weight (string or None, optional (default=None)) – The name of an edge attribute that holds
the numerical value used as a weight. If None, then each edge has weight 1. The degree is
the sum of the edge weights adjacent to the node.

2.2. Basic graph types 57


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns
• If a single node is requested
• deg (int) – In-degree of the node
• OR if multiple nodes are requested
• nd_iter (iterator) – The iterator returns two-tuples of (node, in-degree).
See also:
degree, out_degree

Examples

>>> G = nx.DiGraph()
>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.in_degree(0) # node 0 with degree 0
0
>>> list(G.in_degree([0, 1, 2]))
[(0, 0), (1, 1), (2, 1)]

networkx.DiGraph.out_degree

DiGraph.out_degree
An OutDegreeView for (node, out_degree)
The node out_degree is the number of edges pointing out of the node. The weighted node degree is the sum of
the edge weights for edges incident to that node.
This object provides an iterator over (node, out_degree) as well as lookup for the degree for a single node.
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• weight (string or None, optional (default=None)) – The name of an edge attribute that holds
the numerical value used as a weight. If None, then each edge has weight 1. The degree is
the sum of the edge weights adjacent to the node.
Returns
• If a single node is requested
• deg (int) – Out-degree of the node
• OR if multiple nodes are requested
• nd_iter (iterator) – The iterator returns two-tuples of (node, out-degree).
See also:
degree, in_degree

Examples

58 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.DiGraph()
>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.out_degree(0) # node 0 with degree 1
1
>>> list(G.out_degree([0, 1, 2]))
[(0, 1), (1, 1), (2, 1)]

networkx.DiGraph.size

DiGraph.size(weight=None)
Returns the number of edges or total of all edge weights.
Parameters weight (string or None, optional (default=None)) – The edge attribute that holds the
numerical value used as a weight. If None, then each edge has weight 1.
Returns
size – The number of edges or (if weight keyword is provided) the total weight sum.
If weight is None, returns an int. Otherwise a float (or more general numeric if the weights are
more general).
Return type numeric
See also:
number_of_edges()

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.size()
3

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_edge('a', 'b', weight=2)
>>> G.add_edge('b', 'c', weight=4)
>>> G.size()
2
>>> G.size(weight='weight')
6.0

networkx.DiGraph.number_of_edges

DiGraph.number_of_edges(u=None, v=None)
Returns the number of edges between two nodes.
Parameters u, v (nodes, optional (default=all edges)) – If u and v are specified, return the number
of edges between u and v. Otherwise return the total number of all edges.
Returns nedges – The number of edges in the graph. If nodes u and v are specified return the
number of edges between those nodes. If the graph is directed, this only returns the number of
edges from u to v.
Return type int

2.2. Basic graph types 59


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
size()

Examples

For undirected graphs, this method counts the total number of edges in the graph:

>>> G = nx.path_graph(4)
>>> G.number_of_edges()
3

If you specify two nodes, this counts the total number of edges joining the two nodes:

>>> G.number_of_edges(0, 1)
1

For directed graphs, this method can count the total number of directed edges from u to v:

>>> G = nx.DiGraph()
>>> G.add_edge(0, 1)
>>> G.add_edge(1, 0)
>>> G.number_of_edges(0, 1)
1

Making copies and subgraphs

DiGraph.copy([as_view]) Returns a copy of the graph.


DiGraph.to_undirected([reciprocal, as_view]) Returns an undirected representation of the digraph.
DiGraph.to_directed([as_view]) Returns a directed representation of the graph.
DiGraph.subgraph(nodes) Returns a SubGraph view of the subgraph induced on
nodes.
DiGraph.edge_subgraph(edges) Returns the subgraph induced by the specified edges.
DiGraph.reverse([copy]) Returns the reverse of the graph.

networkx.DiGraph.copy

DiGraph.copy(as_view=False)
Returns a copy of the graph.
The copy method by default returns an independent shallow copy of the graph and attributes. That is, if an
attribute is a container, that container is shared by the original an the copy. Use Python’s copy.deepcopy for
new containers.
If as_view is True then a view is returned instead of a copy.

Notes

All copies reproduce the graph structure, but data attributes may be handled in different ways. There are four
types of copies of a graph that people might want.
Deepcopy – A “deepcopy” copies the graph structure as well as all data attributes and any objects they might
contain. The entire graph object is new so that changes in the copy do not affect the original object. (see Python’s

60 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

copy.deepcopy)
Data Reference (Shallow) – For a shallow copy the graph structure is copied but the edge, node and graph at-
tribute dicts are references to those in the original graph. This saves time and memory but could cause confusion
if you change an attribute in one graph and it changes the attribute in the other. NetworkX does not provide this
level of shallow copy.
Independent Shallow – This copy creates new independent attribute dicts and then does a shallow copy of the
attributes. That is, any attributes that are containers are shared between the new graph and the original. This is
exactly what dict.copy() provides. You can obtain this style copy using:

>>> G = nx.path_graph(5)
>>> H = G.copy()
>>> H = G.copy(as_view=False)
>>> H = nx.Graph(G)
>>> H = G.__class__(G)

Fresh Data – For fresh data, the graph structure is copied while new empty data attribute dicts are created. The
resulting graph is independent of the original and it has no edge, node or graph attributes. Fresh copies are not
enabled. Instead use:

>>> H = G.__class__()
>>> H.add_nodes_from(G)
>>> H.add_edges_from(G.edges)

View – Inspired by dict-views, graph-views act like read-only versions of the original graph, providing a copy
of the original structure without requiring any memory for copying the information.
See the Python copy module for more information on shallow and deep copies, https://docs.python.org/2/library/
copy.html.
Parameters as_view (bool, optional (default=False)) – If True, the returned graph-view provides a
read-only view of the original graph without actually copying any data.
Returns G – A copy of the graph.
Return type Graph
See also:

to_directed() return a directed copy of the graph.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> H = G.copy()

networkx.DiGraph.to_undirected

DiGraph.to_undirected(reciprocal=False, as_view=False)
Returns an undirected representation of the digraph.
Parameters
• reciprocal (bool (optional)) – If True only keep edges that appear in both directions in the
original digraph.

2.2. Basic graph types 61


NetworkX Reference, Release 2.3rc1.dev20190222214247

• as_view (bool (optional, default=False)) – If True return an undirected view of the original
directed graph.
Returns G – An undirected graph with the same name and nodes and with edge (u, v, data) if either
(u, v, data) or (v, u, data) is in the digraph. If both edges exist in digraph and their edge data is
different, only one edge is created with an arbitrary choice of which edge data to use. You must
check and correct for this manually if desired.
Return type Graph
See also:
Graph(), copy(), add_edge(), add_edges_from()

Notes

If edges in both directions (u, v) and (v, u) exist in the graph, attributes for the new undirected edge will be a
combination of the attributes of the directed edges. The edge data is updated in the (arbitrary) order that the
edges are encountered. For more customized control of the edge attributes use add_edge().
This returns a “deepcopy” of the edge, node, and graph attributes which attempts to completely copy all of the
data and references.
This is in contrast to the similar G=DiGraph(D) which returns a shallow copy of the data.
See the Python copy module for more information on shallow and deep copies, https://docs.python.org/2/library/
copy.html.
Warning: If you have subclassed DiGraph to use dict-like objects in the data structure, those changes do not
transfer to the Graph created by this method.

Examples

>>> G = nx.path_graph(2) # or MultiGraph, etc


>>> H = G.to_directed()
>>> list(H.edges)
[(0, 1), (1, 0)]
>>> G2 = H.to_undirected()
>>> list(G2.edges)
[(0, 1)]

networkx.DiGraph.to_directed

DiGraph.to_directed(as_view=False)
Returns a directed representation of the graph.
Returns G – A directed graph with the same name, same nodes, and with each edge (u, v, data)
replaced by two directed edges (u, v, data) and (v, u, data).
Return type DiGraph

Notes

This returns a “deepcopy” of the edge, node, and graph attributes which attempts to completely copy all of the
data and references.

62 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

This is in contrast to the similar D=DiGraph(G) which returns a shallow copy of the data.
See the Python copy module for more information on shallow and deep copies, https://docs.python.org/2/library/
copy.html.
Warning: If you have subclassed Graph to use dict-like objects in the data structure, those changes do not
transfer to the DiGraph created by this method.

Examples

>>> G = nx.Graph() # or MultiGraph, etc


>>> G.add_edge(0, 1)
>>> H = G.to_directed()
>>> list(H.edges)
[(0, 1), (1, 0)]

If already directed, return a (deep) copy

>>> G = nx.DiGraph() # or MultiDiGraph, etc


>>> G.add_edge(0, 1)
>>> H = G.to_directed()
>>> list(H.edges)
[(0, 1)]

networkx.DiGraph.subgraph

DiGraph.subgraph(nodes)
Returns a SubGraph view of the subgraph induced on nodes.
The induced subgraph of the graph contains the nodes in nodes and the edges between those nodes.
Parameters nodes (list, iterable) – A container of nodes which will be iterated through once.
Returns G – A subgraph view of the graph. The graph structure cannot be changed but node/edge
attributes can and are shared with the original graph.
Return type SubGraph View

Notes

The graph, edge and node attributes are shared with the original graph. Changes to the graph structure is ruled
out by the view, but changes to attributes are reflected in the original graph.
To create a subgraph with its own copy of the edge/node attributes use: G.subgraph(nodes).copy()
For an inplace reduction of a graph to a subgraph you can remove nodes: G.remove_nodes_from([n for n in G
if n not in set(nodes)])
Subgraph views are sometimes NOT what you want. In most cases where you want to do more than simply look
at the induced edges, it makes more sense to just create the subgraph as its own graph with code like:

# Create a subgraph SG based on a (possibly multigraph) G


SG = G.__class__()
SG.add_nodes_from((n, G.nodes[n]) for n in largest_wcc)
if SG.is_multigraph:
SG.add_edges_from((n, nbr, key, d)
(continues on next page)

2.2. Basic graph types 63


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


for n, nbrs in G.adj.items() if n in largest_wcc
for nbr, keydict in nbrs.items() if nbr in largest_wcc
for key, d in keydict.items())
else:
SG.add_edges_from((n, nbr, d)
for n, nbrs in G.adj.items() if n in largest_wcc
for nbr, d in nbrs.items() if nbr in largest_wcc)
SG.graph.update(G.graph)

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> H = G.subgraph([0, 1, 2])
>>> list(H.edges)
[(0, 1), (1, 2)]

networkx.DiGraph.edge_subgraph

DiGraph.edge_subgraph(edges)
Returns the subgraph induced by the specified edges.
The induced subgraph contains each edge in edges and each node incident to any one of those edges.
Parameters edges (iterable) – An iterable of edges in this graph.
Returns G – An edge-induced subgraph of this graph with the same edge attributes.
Return type Graph

Notes

The graph, edge, and node attributes in the returned subgraph view are references to the corresponding attributes
in the original graph. The view is read-only.
To create a full graph version of the subgraph with its own copy of the edge or node attributes, use:
>>> G.edge_subgraph(edges).copy()

Examples

>>> G = nx.path_graph(5)
>>> H = G.edge_subgraph([(0, 1), (3, 4)])
>>> list(H.nodes)
[0, 1, 3, 4]
>>> list(H.edges)
[(0, 1), (3, 4)]

networkx.DiGraph.reverse

DiGraph.reverse(copy=True)
Returns the reverse of the graph.

64 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

The reverse is a graph with the same nodes and edges but with the directions of the edges reversed.
Parameters copy (bool optional (default=True)) – If True, return a new DiGraph holding the re-
versed edges. If False, the reverse graph is created using a view of the original graph.

2.2.3 MultiGraph—Undirected graphs with self loops and parallel edges

Overview

class MultiGraph(incoming_graph_data=None, **attr)


An undirected graph class that can store multiedges.
Multiedges are multiple edges between two nodes. Each edge can hold optional data or attributes.
A MultiGraph holds undirected edges. Self loops are allowed.
Nodes can be arbitrary (hashable) Python objects with optional key/value attributes. By convention None is not
used as a node.
Edges are represented as links between nodes with optional key/value attributes.
Parameters
• incoming_graph_data (input graph (optional, default: None)) – Data to initialize graph. If
None (default) an empty graph is created. The data can be any format that is supported by
the to_networkx_graph() function, currently including edge list, dict of dicts, dict of lists,
NetworkX graph, NumPy matrix or 2d ndarray, SciPy sparse matrix, or PyGraphviz graph.
• attr (keyword arguments, optional (default= no attributes)) – Attributes to add to graph as
key=value pairs.
See also:
Graph, DiGraph, MultiDiGraph, OrderedMultiGraph

Examples

Create an empty graph structure (a “null graph”) with no nodes and no edges.

>>> G = nx.MultiGraph()

G can be grown in several ways.


Nodes:
Add one node at a time:

>>> G.add_node(1)

Add the nodes from any container (a list, dict, set or even the lines from a file or the nodes from another graph).

>>> G.add_nodes_from([2, 3])


>>> G.add_nodes_from(range(100, 110))
>>> H = nx.path_graph(10)
>>> G.add_nodes_from(H)

In addition to strings and integers any hashable Python object (except None) can represent a node, e.g. a
customized node object, or even another Graph.

2.2. Basic graph types 65


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G.add_node(H)

Edges:
G can also be grown by adding edges.
Add one edge,

>>> key = G.add_edge(1, 2)

a list of edges,

>>> keys = G.add_edges_from([(1, 2), (1, 3)])

or a collection of edges,

>>> keys = G.add_edges_from(H.edges)

If some edges connect nodes not yet in the graph, the nodes are added automatically. If an edge already exists,
an additional edge is created and stored using a key to identify the edge. By default the key is the lowest unused
integer.

>>> keys = G.add_edges_from([(4,5,{'route':28}), (4,5,{'route':37})])


>>> G[4]
AdjacencyView({3: {0: {}}, 5: {0: {}, 1: {'route': 28}, 2: {'route': 37}}})

Attributes:
Each graph, node, and edge can hold key/value attribute pairs in an associated attribute dictionary (the keys
must be hashable). By default these are empty, but can be added or changed using add_edge, add_node or direct
manipulation of the attribute dictionaries named graph, node and edge respectively.

>>> G = nx.MultiGraph(day="Friday")
>>> G.graph
{'day': 'Friday'}

Add node attributes using add_node(), add_nodes_from() or G.nodes

>>> G.add_node(1, time='5pm')


>>> G.add_nodes_from([3], time='2pm')
>>> G.nodes[1]
{'time': '5pm'}
>>> G.nodes[1]['room'] = 714
>>> del G.nodes[1]['room'] # remove attribute
>>> list(G.nodes(data=True))
[(1, {'time': '5pm'}), (3, {'time': '2pm'})]

Add edge attributes using add_edge(), add_edges_from(), subscript notation, or G.edges.

>>> key = G.add_edge(1, 2, weight=4.7 )


>>> keys = G.add_edges_from([(3, 4), (4, 5)], color='red')
>>> keys = G.add_edges_from([(1,2,{'color':'blue'}), (2,3,{'weight':8})])
>>> G[1][2][0]['weight'] = 4.7
>>> G.edges[1, 2, 0]['weight'] = 4

Warning: we protect the graph data structure by making G.edges[1, 2] a read-only dict-like structure.
However, you can assign to attributes in e.g. G.edges[1, 2]. Thus, use 2 sets of brackets to add/change data

66 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

attributes: G.edges[1, 2]['weight'] = 4 (For multigraphs: MG.edges[u, v, key][name] =


value).
Shortcuts:
Many common graph features allow python syntax to speed reporting.

>>> 1 in G # check if node in graph


True
>>> [n for n in G if n<3] # iterate through nodes
[1, 2]
>>> len(G) # number of nodes in graph
5
>>> G[1] # adjacency dict-like view keyed by neighbor to edge attributes
AdjacencyView({2: {0: {'weight': 4}, 1: {'color': 'blue'}}})

Often the best way to traverse all edges of a graph is via the neighbors. The neighbors are reported as an
adjacency-dict G.adj or G.adjacency().

>>> for n, nbrsdict in G.adjacency():


... for nbr, keydict in nbrsdict.items():
... for key, eattr in keydict.items():
... if 'weight' in eattr:
... # Do something useful with the edges
... pass

But the edges() method is often more convenient:

>>> for u, v, keys, weight in G.edges(data='weight', keys=True):


... if weight is not None:
... # Do something useful with the edges
... pass

Reporting:
Simple graph information is obtained using methods and object-attributes. Reporting usually provides
views instead of containers to reduce memory usage. The views update as the graph is updated simi-
larly to dict-views. The objects nodes, `edges and adj provide access to data attributes via lookup
(e.g. nodes[n], `edges[u, v], adj[u][v]) and iteration (e.g. nodes.items(), nodes.
data('color'), nodes.data('color', default='blue') and similarly for edges) Views exist
for nodes, edges, neighbors()/adj and degree.
For details on these and other miscellaneous methods, see below.
Subclasses (Advanced):
The MultiGraph class uses a dict-of-dict-of-dict-of-dict data structure. The outer dict (node_dict) holds adja-
cency information keyed by node. The next dict (adjlist_dict) represents the adjacency information and holds
edge_key dicts keyed by neighbor. The edge_key dict holds each edge_attr dict keyed by edge key. The inner
dict (edge_attr_dict) represents the edge data and holds edge attribute values keyed by attribute names.
Each of these four dicts in the dict-of-dict-of-dict-of-dict structure can be replaced by a user defined dict-like
object. In general, the dict-like features should be maintained but extra features can be added. To replace
one of the dicts create a new graph class by changing the class(!) variable holding the factory for that dict-
like structure. The variable names are node_dict_factory, node_attr_dict_factory, adjlist_inner_dict_factory,
adjlist_outer_dict_factory, edge_key_dict_factory, edge_attr_dict_factory and graph_attr_dict_factory.
node_dict_factory [function, (default: dict)] Factory function to be used to create the dict containing node
attributes, keyed by node id. It should require no arguments and return a dict-like object

2.2. Basic graph types 67


NetworkX Reference, Release 2.3rc1.dev20190222214247

node_attr_dict_factory: function, (default: dict) Factory function to be used to create the node attribute dict
which holds attribute values keyed by attribute name. It should require no arguments and return a dict-like
object
adjlist_outer_dict_factory [function, (default: dict)] Factory function to be used to create the outer-most dict
in the data structure that holds adjacency info keyed by node. It should require no arguments and return a
dict-like object.
adjlist_inner_dict_factory [function, (default: dict)] Factory function to be used to create the adjacency list
dict which holds multiedge key dicts keyed by neighbor. It should require no arguments and return a
dict-like object.
edge_key_dict_factory [function, (default: dict)] Factory function to be used to create the edge key dict which
holds edge data keyed by edge key. It should require no arguments and return a dict-like object.
edge_attr_dict_factory [function, (default: dict)] Factory function to be used to create the edge attribute dict
which holds attribute values keyed by attribute name. It should require no arguments and return a dict-like
object.
graph_attr_dict_factory [function, (default: dict)] Factory function to be used to create the graph attribute
dict which holds attribute values keyed by attribute name. It should require no arguments and return a
dict-like object.
Typically, if your extension doesn’t impact the data structure all methods will inherited without issue except:
to_directed/to_undirected. By default these methods create a DiGraph/Graph class and you probably
want them to create your extension of a DiGraph/Graph. To facilitate this we define two class variables that you
can set in your subclass.
to_directed_class [callable, (default: DiGraph or MultiDiGraph)] Class to create a new graph structure in the
to_directed method. If None, a NetworkX class (DiGraph or MultiDiGraph) is used.
to_undirected_class [callable, (default: Graph or MultiGraph)] Class to create a new graph structure in the
to_undirected method. If None, a NetworkX class (Graph or MultiGraph) is used.

Examples

Please see ordered for examples of creating graph subclasses by overwriting the base class dict with a
dictionary-like object.

Methods

Adding and removing nodes and edges

MultiGraph.__init__([incoming_graph_data]) Initialize a graph with edges, name, or graph attributes.


MultiGraph.add_node(node_for_adding, **attr)Add a single node node_for_adding and update
node attributes.
MultiGraph.add_nodes_from(nodes_for_adding, Add multiple nodes.
...)
MultiGraph.remove_node(n) Remove node n.
MultiGraph.remove_nodes_from(nodes) Remove multiple nodes.
MultiGraph.add_edge(u_for_edge, v_for_edge) Add an edge between u and v.
MultiGraph.add_edges_from(ebunch_to_add, Add all the edges in ebunch_to_add.
**attr)
Continued on next page

68 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 9 – continued from previous page


Add weighted edges in ebunch_to_add with speci-
MultiGraph.add_weighted_edges_from(ebunch_to_add)
fied weight attr
MultiGraph.new_edge_key(u, v) Returns an unused key for edges between nodes u and
v.
MultiGraph.remove_edge(u, v[, key]) Remove an edge between u and v.
MultiGraph.remove_edges_from(ebunch) Remove all edges specified in ebunch.
MultiGraph.update([edges, nodes]) Update the graph using nodes/edges/graphs as input.
MultiGraph.clear() Remove all nodes and edges from the graph.

networkx.MultiGraph.__init__

MultiGraph.__init__(incoming_graph_data=None, **attr)
Initialize a graph with edges, name, or graph attributes.
Parameters
• incoming_graph_data (input graph) – Data to initialize graph. If incom-
ing_graph_data=None (default) an empty graph is created. The data can be an edge list,
or any NetworkX graph object. If the corresponding optional Python packages are installed
the data can also be a NumPy matrix or 2d ndarray, a SciPy sparse matrix, or a PyGraphviz
graph.
• attr (keyword arguments, optional (default= no attributes)) – Attributes to add to graph as
key=value pairs.
See also:
convert()

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G = nx.Graph(name='my graph')
>>> e = [(1, 2), (2, 3), (3, 4)] # list of edges
>>> G = nx.Graph(e)

Arbitrary graph attribute pairs (key=value) may be assigned


>>> G = nx.Graph(e, day="Friday")
>>> G.graph
{'day': 'Friday'}

networkx.MultiGraph.add_node

MultiGraph.add_node(node_for_adding, **attr)
Add a single node node_for_adding and update node attributes.
Parameters
• node_for_adding (node) – A node can be any hashable Python object except None.
• attr (keyword arguments, optional) – Set or change node attributes using key=value.
See also:
add_nodes_from()

2.2. Basic graph types 69


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_node(1)
>>> G.add_node('Hello')
>>> K3 = nx.Graph([(0, 1), (1, 2), (2, 0)])
>>> G.add_node(K3)
>>> G.number_of_nodes()
3

Use keywords set/change node attributes:

>>> G.add_node(1, size=10)


>>> G.add_node(3, weight=0.4, UTM=('13S', 382871, 3972649))

Notes

A hashable object is one that can be used as a key in a Python dictionary. This includes strings, numbers, tuples
of strings and numbers, etc.
On many platforms hashable items also include mutables such as NetworkX Graphs, though one should be
careful that the hash doesn’t change on mutables.

networkx.MultiGraph.add_nodes_from

MultiGraph.add_nodes_from(nodes_for_adding, **attr)
Add multiple nodes.
Parameters
• nodes_for_adding (iterable container) – A container of nodes (list, dict, set, etc.). OR A
container of (node, attribute dict) tuples. Node attributes are updated using the attribute dict.
• attr (keyword arguments, optional (default= no attributes)) – Update attributes for all nodes
in nodes. Node attributes specified in nodes as a tuple take precedence over attributes spec-
ified via keyword arguments.
See also:
add_node()

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_nodes_from('Hello')
>>> K3 = nx.Graph([(0, 1), (1, 2), (2, 0)])
>>> G.add_nodes_from(K3)
>>> sorted(G.nodes(), key=str)
[0, 1, 2, 'H', 'e', 'l', 'o']

Use keywords to update specific node attributes for every node.

>>> G.add_nodes_from([1, 2], size=10)


>>> G.add_nodes_from([3, 4], weight=0.4)

70 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Use (node, attrdict) tuples to update attributes for specific nodes.


>>> G.add_nodes_from([(1, dict(size=11)), (2, {'color':'blue'})])
>>> G.nodes[1]['size']
11
>>> H = nx.Graph()
>>> H.add_nodes_from(G.nodes(data=True))
>>> H.nodes[1]['size']
11

networkx.MultiGraph.remove_node

MultiGraph.remove_node(n)
Remove node n.
Removes the node n and all adjacent edges. Attempting to remove a non-existent node will raise an exception.
Parameters n (node) – A node in the graph
Raises NetworkXError – If n is not in the graph.
See also:
remove_nodes_from()

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> list(G.edges)
[(0, 1), (1, 2)]
>>> G.remove_node(1)
>>> list(G.edges)
[]

networkx.MultiGraph.remove_nodes_from

MultiGraph.remove_nodes_from(nodes)
Remove multiple nodes.
Parameters nodes (iterable container) – A container of nodes (list, dict, set, etc.). If a node in the
container is not in the graph it is silently ignored.
See also:
remove_node()

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> e = list(G.nodes)
>>> e
[0, 1, 2]
>>> G.remove_nodes_from(e)
>>> list(G.nodes)
[]

2.2. Basic graph types 71


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiGraph.add_edge

MultiGraph.add_edge(u_for_edge, v_for_edge, key=None, **attr)


Add an edge between u and v.
The nodes u and v will be automatically added if they are not already in the graph.
Edge attributes can be specified with keywords or by directly accessing the edge’s attribute dictionary. See
examples below.
Parameters
• u_for_edge, v_for_edge (nodes) – Nodes can be, for example, strings or numbers. Nodes
must be hashable (and not None) Python objects.
• key (hashable identifier, optional (default=lowest unused integer)) – Used to distinguish
multiedges between a pair of nodes.
• attr (keyword arguments, optional) – Edge data (or labels or objects) can be assigned using
keyword arguments.
Returns
Return type The edge key assigned to the edge.
See also:

add_edges_from() add a collection of edges

Notes

To replace/update edge data, use the optional key argument to identify a unique edge. Otherwise a new edge
will be created.
NetworkX algorithms designed for weighted graphs cannot use multigraphs directly because it is not clear
how to handle multiedge weights. Convert to Graph using edge attribute ‘weight’ to enable weighted graph
algorithms.
Default keys are generated using the method new_edge_key(). This method can be overridden by subclass-
ing the base class and providing a custom new_edge_key() method.

Examples

The following all add the edge e=(1, 2) to graph G:

>>> G = nx.MultiGraph()
>>> e = (1, 2)
>>> ekey = G.add_edge(1, 2) # explicit two-node form
>>> G.add_edge(*e) # single edge as tuple of two nodes
1
>>> G.add_edges_from( [(1, 2)] ) # add edges from iterable container
[2]

Associate data to edges using keywords:

>>> ekey = G.add_edge(1, 2, weight=3)


>>> ekey = G.add_edge(1, 2, key=0, weight=4) # update data for key=0
>>> ekey = G.add_edge(1, 3, weight=7, capacity=15, length=342.7)

72 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

For non-string attribute keys, use subscript notation.

>>> ekey = G.add_edge(1, 2)


>>> G[1][2][0].update({0: 5})
>>> G.edges[1, 2, 0].update({0: 5})

networkx.MultiGraph.add_edges_from

MultiGraph.add_edges_from(ebunch_to_add, **attr)
Add all the edges in ebunch_to_add.
Parameters
• ebunch_to_add (container of edges) – Each edge given in the container will be added to
the graph. The edges can be:
– 2-tuples (u, v) or
– 3-tuples (u, v, d) for an edge data dict d, or
– 3-tuples (u, v, k) for not iterable key k, or
– 4-tuples (u, v, k, d) for an edge with data and key k
• attr (keyword arguments, optional) – Edge data (or labels or objects) can be assigned using
keyword arguments.
Returns
Return type A list of edge keys assigned to the edges in ebunch.
See also:

add_edge() add a single edge


add_weighted_edges_from() convenient way to add weighted edges

Notes

Adding the same edge twice has no effect but any edge data will be updated when each duplicate edge is added.
Edge attributes specified in an ebunch take precedence over attributes specified via keyword arguments.
Default keys are generated using the method new_edge_key(). This method can be overridden by subclass-
ing the base class and providing a custom new_edge_key() method.

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_edges_from([(0, 1), (1, 2)]) # using a list of edge tuples
>>> e = zip(range(0, 3), range(1, 4))
>>> G.add_edges_from(e) # Add the path graph 0-1-2-3

Associate data to edges

>>> G.add_edges_from([(1, 2), (2, 3)], weight=3)


>>> G.add_edges_from([(3, 4), (1, 4)], label='WN2898')

2.2. Basic graph types 73


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiGraph.add_weighted_edges_from

MultiGraph.add_weighted_edges_from(ebunch_to_add, weight=’weight’, **attr)


Add weighted edges in ebunch_to_add with specified weight attr
Parameters
• ebunch_to_add (container of edges) – Each edge given in the list or container will be added
to the graph. The edges must be given as 3-tuples (u, v, w) where w is a number.
• weight (string, optional (default= ‘weight’)) – The attribute name for the edge weights to
be added.
• attr (keyword arguments, optional (default= no attributes)) – Edge attributes to add/update
for all edges.
See also:

add_edge() add a single edge


add_edges_from() add multiple edges

Notes

Adding the same edge twice for Graph/DiGraph simply updates the edge data. For MultiGraph/MultiDiGraph,
duplicate edges are stored.

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_weighted_edges_from([(0, 1, 3.0), (1, 2, 7.5)])

networkx.MultiGraph.new_edge_key

MultiGraph.new_edge_key(u, v)
Returns an unused key for edges between nodes u and v.
The nodes u and v do not need to be already in the graph.

Notes

In the standard MultiGraph class the new key is the number of existing edges between u and v (increased
if necessary to ensure unused). The first edge will have key 0, then 1, etc. If an edge is removed further
new_edge_keys may not be in this order.
Parameters u, v (nodes)
Returns key
Return type int

74 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiGraph.remove_edge

MultiGraph.remove_edge(u, v, key=None)
Remove an edge between u and v.
Parameters
• u, v (nodes) – Remove an edge between nodes u and v.
• key (hashable identifier, optional (default=None)) – Used to distinguish multiple edges be-
tween a pair of nodes. If None remove a single (arbitrary) edge between u and v.
Raises NetworkXError – If there is not an edge between u and v, or if there is no edge with the
specified key.
See also:

remove_edges_from() remove a collection of edges

Examples

>>> G = nx.MultiGraph()
>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.remove_edge(0, 1)
>>> e = (1, 2)
>>> G.remove_edge(*e) # unpacks e from an edge tuple

For multiple edges

>>> G = nx.MultiGraph() # or MultiDiGraph, etc


>>> G.add_edges_from([(1, 2), (1, 2), (1, 2)]) # key_list returned
[0, 1, 2]
>>> G.remove_edge(1, 2) # remove a single (arbitrary) edge

For edges with keys

>>> G = nx.MultiGraph() # or MultiDiGraph, etc


>>> G.add_edge(1, 2, key='first')
'first'
>>> G.add_edge(1, 2, key='second')
'second'
>>> G.remove_edge(1, 2, key='second')

networkx.MultiGraph.remove_edges_from

MultiGraph.remove_edges_from(ebunch)
Remove all edges specified in ebunch.
Parameters ebunch (list or container of edge tuples) – Each edge given in the list or container will
be removed from the graph. The edges can be:
• 2-tuples (u, v) All edges between u and v are removed.
• 3-tuples (u, v, key) The edge identified by key is removed.
• 4-tuples (u, v, key, data) where data is ignored.
See also:

2.2. Basic graph types 75


NetworkX Reference, Release 2.3rc1.dev20190222214247

remove_edge() remove a single edge

Notes

Will fail silently if an edge in ebunch is not in the graph.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> ebunch=[(1, 2), (2, 3)]
>>> G.remove_edges_from(ebunch)

Removing multiple copies of edges

>>> G = nx.MultiGraph()
>>> keys = G.add_edges_from([(1, 2), (1, 2), (1, 2)])
>>> G.remove_edges_from([(1, 2), (1, 2)])
>>> list(G.edges())
[(1, 2)]
>>> G.remove_edges_from([(1, 2), (1, 2)]) # silently ignore extra copy
>>> list(G.edges) # now empty graph
[]

networkx.MultiGraph.update

MultiGraph.update(edges=None, nodes=None)
Update the graph using nodes/edges/graphs as input.
Like dict.update, this method takes a graph as input, adding the graph’s noes and edges to this graph. It can also
take two inputs: edges and nodes. Finally it can take either edges or nodes. To specify only nodes the keyword
nodes must be used.
The collections of edges and nodes are treated similarly to the add_edges_from/add_nodes_from methods.
When iterated, they should yield 2-tuples (u, v) or 3-tuples (u, v, datadict).
Parameters
• edges (Graph object, collection of edges, or None) – The first parameter can be a graph or
some edges. If it has attributes nodes and edges, then it is taken to be a Graph-like object
and those attributes are used as collections of nodes and edges to be added to the graph. If
the first parameter does not have those attributes, it is treated as a collection of edges and
added to the graph. If the first argument is None, no edges are added.
• nodes (collection of nodes, or None) – The second parameter is treated as a collection of
nodes to be added to the graph unless it is None. If edges is None and nodes is
None an exception is raised. If the first parameter is a Graph, then nodes is ignored.

Examples

>>> G = nx.path_graph(5)
>>> G.update(nx.complete_graph(range(4,10)))
>>> from itertools import combinations
>>> edges = ((u, v, {'power': u * v})
(continues on next page)

76 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


... for u, v in combinations(range(10, 20), 2)
... if u * v < 225)
>>> nodes = [1000] # for singleton, use a container
>>> G.update(edges, nodes)

Notes

It you want to update the graph using an adjacency structure it is straightforward to obtain the edges/nodes from
adjacency. The following examples provide common cases, your adjacency may be slightly different and require
tweaks of these examples.

>>> # dict-of-set/list/tuple
>>> adj = {1: {2, 3}, 2: {1, 3}, 3: {1, 2}}
>>> e = [(u, v) for u, nbrs in adj.items() for v in nbrs]
>>> G.update(edges=e, nodes=adj)

>>> DG = nx.DiGraph()
>>> # dict-of-dict-of-attribute
>>> adj = {1: {2: 1.3, 3: 0.7}, 2: {1: 1.4}, 3: {1: 0.7}}
>>> e = [(u, v, {'weight': d}) for u, nbrs in adj.items()
... for v, d in nbrs.items()]
>>> DG.update(edges=e, nodes=adj)

>>> # dict-of-dict-of-dict
>>> adj = {1: {2: {'weight': 1.3}, 3: {'color': 0.7, 'weight':1.2}}}
>>> e = [(u, v, {'weight': d}) for u, nbrs in adj.items()
... for v, d in nbrs.items()]
>>> DG.update(edges=e, nodes=adj)

>>> # predecessor adjacency (dict-of-set)


>>> pred = {1: {2, 3}, 2: {3}, 3: {3}}
>>> e = [(v, u) for u, nbrs in pred.items() for v in nbrs]

>>> # MultiGraph dict-of-dict-of-dict-of-attribute


>>> MDG = nx.MultiDiGraph()
>>> adj = {1: {2: {0: {'weight': 1.3}, 1: {'weight': 1.2}}},
... 3: {2: {0: {'weight': 0.7}}}}
>>> e = [(u, v, ekey, d) for u, nbrs in adj.items()
... for v, keydict in nbrs.items()
... for ekey, d in keydict.items()]
>>> MDG.update(edges=e)

See also:

add_edges_from() add multiple edges to a graph


add_nodes_from() add multiple nodes to a graph

networkx.MultiGraph.clear

MultiGraph.clear()
Remove all nodes and edges from the graph.

2.2. Basic graph types 77


NetworkX Reference, Release 2.3rc1.dev20190222214247

This also removes the name, and all graph, node, and edge attributes.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.clear()
>>> list(G.nodes)
[]
>>> list(G.edges)
[]

Reporting nodes edges and neighbors

MultiGraph.nodes A NodeView of the Graph as G.nodes or G.nodes().


MultiGraph.__iter__() Iterate over the nodes.
MultiGraph.has_node(n) Returns True if the graph contains the node n.
MultiGraph.__contains__(n) Returns True if n is a node, False otherwise.
MultiGraph.edges Returns an iterator over the edges.
MultiGraph.has_edge(u, v[, key]) Returns True if the graph has an edge between nodes u
and v.
MultiGraph.get_edge_data(u, v[, key, default]) Returns the attribute dictionary associated with edge (u,
v).
MultiGraph.neighbors(n) Returns an iterator over all neighbors of node n.
MultiGraph.adj Graph adjacency object holding the neighbors of each
node.
MultiGraph.__getitem__(n) Returns a dict of neighbors of node n.
MultiGraph.adjacency() Returns an iterator over (node, adjacency dict) tuples for
all nodes.
MultiGraph.nbunch_iter([nbunch]) Returns an iterator over nodes contained in nbunch that
are also in the graph.

networkx.MultiGraph.nodes

MultiGraph.nodes
A NodeView of the Graph as G.nodes or G.nodes().
Can be used as G.nodes for data lookup and for set-like operations. Can also be used as G.
nodes(data='color', default=None) to return a NodeDataView which reports specific node data
but no set operations. It presents a dict-like interface as well with G.nodes.items() iterating over (node,
nodedata) 2-tuples and G.nodes[3]['foo'] providing the value of the foo attribute for node 3. In
addition, a view G.nodes.data('foo') provides a dict-like interface to the foo attribute of each node.
G.nodes.data('foo', default=1) provides a default for nodes that do not have attribute foo.
Parameters
• data (string or bool, optional (default=False)) – The node attribute returned in 2-tuple (n,
ddict[data]). If True, return entire node attribute dict as (n, ddict). If False, return just the
nodes n.
• default (value, optional (default=None)) – Value used for nodes that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns

78 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Allows set-like operations over the nodes as well as node attribute dict lookup and calling to
get a NodeDataView. A NodeDataView iterates over (n, data) and has no set operations. A
NodeView iterates over n and includes set operations.
When called, if data is False, an iterator over nodes. Otherwise an iterator of 2-tuples (node,
attribute value) where the attribute is specified in data. If data is True then the attribute becomes
the entire data dictionary.
Return type NodeView

Notes

If your node data is not needed, it is simpler and equivalent to use the expression for n in G, or list(G).

Examples

There are two simple ways of getting a list of all nodes in the graph:
>>> G = nx.path_graph(3)
>>> list(G.nodes)
[0, 1, 2]
>>> list(G)
[0, 1, 2]

To get the node data along with the nodes:


>>> G.add_node(1, time='5pm')
>>> G.nodes[0]['foo'] = 'bar'
>>> list(G.nodes(data=True))
[(0, {'foo': 'bar'}), (1, {'time': '5pm'}), (2, {})]
>>> list(G.nodes.data())
[(0, {'foo': 'bar'}), (1, {'time': '5pm'}), (2, {})]

>>> list(G.nodes(data='foo'))
[(0, 'bar'), (1, None), (2, None)]
>>> list(G.nodes.data('foo'))
[(0, 'bar'), (1, None), (2, None)]

>>> list(G.nodes(data='time'))
[(0, None), (1, '5pm'), (2, None)]
>>> list(G.nodes.data('time'))
[(0, None), (1, '5pm'), (2, None)]

>>> list(G.nodes(data='time', default='Not Available'))


[(0, 'Not Available'), (1, '5pm'), (2, 'Not Available')]
>>> list(G.nodes.data('time', default='Not Available'))
[(0, 'Not Available'), (1, '5pm'), (2, 'Not Available')]

If some of your nodes have an attribute and the rest are assumed to have a default attribute value you can create
a dictionary from node/attribute pairs using the default keyword argument to guarantee the value is never
None:
>>> G = nx.Graph()
>>> G.add_node(0)
>>> G.add_node(1, weight=2)
(continues on next page)

2.2. Basic graph types 79


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> G.add_node(2, weight=3)
>>> dict(G.nodes(data='weight', default=1))
{0: 1, 1: 2, 2: 3}

networkx.MultiGraph.__iter__

MultiGraph.__iter__()
Iterate over the nodes. Use: ‘for n in G’.
Returns niter – An iterator over all nodes in the graph.
Return type iterator

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> [n for n in G]
[0, 1, 2, 3]
>>> list(G)
[0, 1, 2, 3]

networkx.MultiGraph.has_node

MultiGraph.has_node(n)
Returns True if the graph contains the node n.
Identical to n in G
Parameters n (node)

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.has_node(0)
True

It is more readable and simpler to use

>>> 0 in G
True

networkx.MultiGraph.__contains__

MultiGraph.__contains__(n)
Returns True if n is a node, False otherwise. Use: ‘n in G’.

80 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> 1 in G
True

networkx.MultiGraph.edges

MultiGraph.edges
Returns an iterator over the edges.
edges(self, nbunch=None, data=False, keys=False, default=None)
The EdgeView provides set-like operations on the edge-tuples as well as edge attribute lookup. When called,
it also provides an EdgeDataView object which allows control of access to edge attributes (but does not pro-
vide set-like operations). Hence, G.edges[u, v]['color'] provides the value of the color attribute for
edge (u, v) while for (u, v, c) in G.edges(data='color', default='red'): iterates
through all the edges yielding the color attribute.
Edges are returned as tuples with optional data and keys in the order (node, neighbor, key, data).
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• data (string or bool, optional (default=False)) – The edge attribute returned in 3-tuple (u, v,
ddict[data]). If True, return edge attribute dict in 3-tuple (u, v, ddict). If False, return 2-tuple
(u, v).
• keys (bool, optional (default=False)) – If True, return edge keys with each edge.
• default (value, optional (default=None)) – Value used for edges that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns edges – A view of edge attributes, usually it iterates over (u, v) (u, v, k) or (u, v, k, d) tuples
of edges, but can also be used for attribute lookup as edges[u, v, k]['foo'].
Return type MultiEdgeView

Notes

Nodes in nbunch that are not in the graph will be (quietly) ignored. For directed graphs this returns the out-edges.

Examples

>>> G = nx.MultiGraph() # or MultiDiGraph


>>> nx.add_path(G, [0, 1, 2])
>>> key = G.add_edge(2, 3, weight=5)
>>> [e for e in G.edges()]
[(0, 1), (1, 2), (2, 3)]
>>> G.edges.data() # default data is {} (empty dict)
MultiEdgeDataView([(0, 1, {}), (1, 2, {}), (2, 3, {'weight': 5})])
>>> G.edges.data('weight', default=1)
MultiEdgeDataView([(0, 1, 1), (1, 2, 1), (2, 3, 5)])
>>> G.edges(keys=True) # default keys are integers
(continues on next page)

2.2. Basic graph types 81


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


MultiEdgeView([(0, 1, 0), (1, 2, 0), (2, 3, 0)])
>>> G.edges.data(keys=True)
MultiEdgeDataView([(0, 1, 0, {}), (1, 2, 0, {}), (2, 3, 0, {'weight': 5})])
>>> G.edges.data('weight', default=1, keys=True)
MultiEdgeDataView([(0, 1, 0, 1), (1, 2, 0, 1), (2, 3, 0, 5)])
>>> G.edges([0, 3])
MultiEdgeDataView([(0, 1), (3, 2)])
>>> G.edges(0)
MultiEdgeDataView([(0, 1)])

networkx.MultiGraph.has_edge

MultiGraph.has_edge(u, v, key=None)
Returns True if the graph has an edge between nodes u and v.
This is the same as v in G[u] or key in G[u][v] without KeyError exceptions.
Parameters
• u, v (nodes) – Nodes can be, for example, strings or numbers.
• key (hashable identifier, optional (default=None)) – If specified return True only if the edge
with key is found.
Returns edge_ind – True if edge is in the graph, False otherwise.
Return type bool

Examples

Can be called either using two nodes u, v, an edge tuple (u, v), or an edge tuple (u, v, key).

>>> G = nx.MultiGraph() # or MultiDiGraph


>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.has_edge(0, 1) # using two nodes
True
>>> e = (0, 1)
>>> G.has_edge(*e) # e is a 2-tuple (u, v)
True
>>> G.add_edge(0, 1, key='a')
'a'
>>> G.has_edge(0, 1, key='a') # specify key
True
>>> e=(0, 1, 'a')
>>> G.has_edge(*e) # e is a 3-tuple (u, v, 'a')
True

The following syntax are equivalent:

>>> G.has_edge(0, 1)
True
>>> 1 in G[0] # though this gives :exc:`KeyError` if 0 not in G
True

82 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiGraph.get_edge_data

MultiGraph.get_edge_data(u, v, key=None, default=None)


Returns the attribute dictionary associated with edge (u, v).
This is identical to G[u][v][key] except the default is returned instead of an exception is the edge doesn’t
exist.
Parameters
• u, v (nodes)
• default (any Python object (default=None)) – Value to return if the edge (u, v) is not found.
• key (hashable identifier, optional (default=None)) – Return data only for the edge with
specified key.
Returns edge_dict – The edge attribute dictionary.
Return type dictionary

Examples

>>> G = nx.MultiGraph() # or MultiDiGraph


>>> key = G.add_edge(0, 1, key='a', weight=7)
>>> G[0][1]['a'] # key='a'
{'weight': 7}
>>> G.edges[0, 1, 'a'] # key='a'
{'weight': 7}

Warning: we protect the graph data structure by making G.edges and G[1][2] read-only dict-like structures.
However, you can assign values to attributes in e.g. G.edges[1, 2, 'a'] or G[1][2]['a'] using an
additional bracket as shown next. You need to specify all edge info to assign to the edge data associated with an
edge.

>>> G[0][1]['a']['weight'] = 10
>>> G.edges[0, 1, 'a']['weight'] = 10
>>> G[0][1]['a']['weight']
10
>>> G.edges[1, 0, 'a']['weight']
10

>>> G = nx.MultiGraph() # or MultiDiGraph


>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.get_edge_data(0, 1)
{0: {}}
>>> e = (0, 1)
>>> G.get_edge_data(*e) # tuple form
{0: {}}
>>> G.get_edge_data('a', 'b', default=0) # edge not in graph, return 0
0

networkx.MultiGraph.neighbors

MultiGraph.neighbors(n)
Returns an iterator over all neighbors of node n.

2.2. Basic graph types 83


NetworkX Reference, Release 2.3rc1.dev20190222214247

This is identical to iter(G[n])


Parameters n (node) – A node in the graph
Returns neighbors – An iterator over all neighbors of node n
Return type iterator
Raises NetworkXError – If the node n is not in the graph.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> [n for n in G.neighbors(0)]
[1]

Notes

It is usually more convenient (and faster) to access the adjacency dictionary as G[n]:

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_edge('a', 'b', weight=7)
>>> G['a']
AtlasView({'b': {'weight': 7}})
>>> G = nx.path_graph(4)
>>> [n for n in G[0]]
[1]

networkx.MultiGraph.adj

MultiGraph.adj
Graph adjacency object holding the neighbors of each node.
This object is a read-only dict-like structure with node keys and neighbor-dict values. The neighbor-dict is keyed
by neighbor to the edgekey-data-dict. So G.adj[3][2][0]['color'] = 'blue' sets the color of the
edge (3, 2, 0) to "blue".
Iterating over G.adj behaves like a dict. Useful idioms include for nbr, nbrdict in G.adj[n].
items():.
The neighbor information is also provided by subscripting the graph. So for nbr, foovalue in
G[node].data('foo', default=1): works.
For directed graphs, G.adj holds outgoing (successor) info.

networkx.MultiGraph.__getitem__

MultiGraph.__getitem__(n)
Returns a dict of neighbors of node n. Use: ‘G[n]’.
Parameters n (node) – A node in the graph.
Returns adj_dict – The adjacency dictionary for nodes connected to n.
Return type dictionary

84 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

G[n] is the same as G.adj[n] and similar to G.neighbors(n) (which is an iterator over G.adj[n])

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G[0]
AtlasView({1: {}})

networkx.MultiGraph.adjacency

MultiGraph.adjacency()
Returns an iterator over (node, adjacency dict) tuples for all nodes.
For directed graphs, only outgoing neighbors/adjacencies are included.
Returns adj_iter – An iterator over (node, adjacency dictionary) for all nodes in the graph.
Return type iterator

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> [(n, nbrdict) for n, nbrdict in G.adjacency()]
[(0, {1: {}}), (1, {0: {}, 2: {}}), (2, {1: {}, 3: {}}), (3, {2: {}})]

networkx.MultiGraph.nbunch_iter

MultiGraph.nbunch_iter(nbunch=None)
Returns an iterator over nodes contained in nbunch that are also in the graph.
The nodes in nbunch are checked for membership in the graph and if not are silently ignored.
Parameters nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
Returns niter – An iterator over nodes in nbunch that are also in the graph. If nbunch is None,
iterate over all nodes in the graph.
Return type iterator
Raises NetworkXError – If nbunch is not a node or or sequence of nodes. If a node in nbunch is
not hashable.
See also:
Graph.__iter__()

Notes

When nbunch is an iterator, the returned iterator yields values directly from nbunch, becoming exhausted when
nbunch is exhausted.

2.2. Basic graph types 85


NetworkX Reference, Release 2.3rc1.dev20190222214247

To test whether nbunch is a single node, one can use “if nbunch in self:”, even after processing with this routine.
If nbunch is not a node or a (possibly empty) sequence/iterator or None, a NetworkXError is raised. Also, if
any object in nbunch is not hashable, a NetworkXError is raised.

Counting nodes edges and neighbors

MultiGraph.order() Returns the number of nodes in the graph.


MultiGraph.number_of_nodes() Returns the number of nodes in the graph.
MultiGraph.__len__() Returns the number of nodes.
MultiGraph.degree A DegreeView for the Graph as G.degree or G.degree().
MultiGraph.size([weight]) Returns the number of edges or total of all edge weights.
MultiGraph.number_of_edges([u, v]) Returns the number of edges between two nodes.

networkx.MultiGraph.order

MultiGraph.order()
Returns the number of nodes in the graph.
Returns nnodes – The number of nodes in the graph.
Return type int
See also:
number_of_nodes(), __len__()

networkx.MultiGraph.number_of_nodes

MultiGraph.number_of_nodes()
Returns the number of nodes in the graph.
Returns nnodes – The number of nodes in the graph.
Return type int
See also:
order(), __len__()

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> len(G)
3

networkx.MultiGraph.__len__

MultiGraph.__len__()
Returns the number of nodes. Use: ‘len(G)’.
Returns nnodes – The number of nodes in the graph.
Return type int

86 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> len(G)
4

networkx.MultiGraph.degree

MultiGraph.degree
A DegreeView for the Graph as G.degree or G.degree().
The node degree is the number of edges adjacent to the node. The weighted node degree is the sum of the edge
weights for edges incident to that node.
This object provides an iterator for (node, degree) as well as lookup for the degree for a single node.
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• weight (string or None, optional (default=None)) – The name of an edge attribute that holds
the numerical value used as a weight. If None, then each edge has weight 1. The degree is
the sum of the edge weights adjacent to the node.
Returns
• If a single node is requested
• deg (int) – Degree of the node, if a single node is passed as argument.
• OR if multiple nodes are requested
• nd_iter (iterator) – The iterator returns two-tuples of (node, degree).

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.degree(0) # node 0 with degree 1
1
>>> list(G.degree([0, 1]))
[(0, 1), (1, 2)]

networkx.MultiGraph.size

MultiGraph.size(weight=None)
Returns the number of edges or total of all edge weights.
Parameters weight (string or None, optional (default=None)) – The edge attribute that holds the
numerical value used as a weight. If None, then each edge has weight 1.
Returns
size – The number of edges or (if weight keyword is provided) the total weight sum.
If weight is None, returns an int. Otherwise a float (or more general numeric if the weights are
more general).

2.2. Basic graph types 87


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type numeric


See also:
number_of_edges()

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.size()
3

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_edge('a', 'b', weight=2)
>>> G.add_edge('b', 'c', weight=4)
>>> G.size()
2
>>> G.size(weight='weight')
6.0

networkx.MultiGraph.number_of_edges

MultiGraph.number_of_edges(u=None, v=None)
Returns the number of edges between two nodes.
Parameters u, v (nodes, optional (Gefault=all edges)) – If u and v are specified, return the number
of edges between u and v. Otherwise return the total number of all edges.
Returns nedges – The number of edges in the graph. If nodes u and v are specified return the
number of edges between those nodes. If the graph is directed, this only returns the number of
edges from u to v.
Return type int
See also:
size()

Examples

For undirected multigraphs, this method counts the total number of edges in the graph:

>>> G = nx.MultiGraph()
>>> G.add_edges_from([(0, 1), (0, 1), (1, 2)])
[0, 1, 0]
>>> G.number_of_edges()
3

If you specify two nodes, this counts the total number of edges joining the two nodes:

>>> G.number_of_edges(0, 1)
2

For directed multigraphs, this method can count the total number of directed edges from u to v:

88 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.MultiDiGraph()
>>> G.add_edges_from([(0, 1), (0, 1), (1, 0)])
[0, 1, 0]
>>> G.number_of_edges(0, 1)
2
>>> G.number_of_edges(1, 0)
1

Making copies and subgraphs

MultiGraph.copy([as_view]) Returns a copy of the graph.


MultiGraph.to_undirected([as_view]) Returns an undirected copy of the graph.
MultiGraph.to_directed([as_view]) Returns a directed representation of the graph.
MultiGraph.subgraph(nodes) Returns a SubGraph view of the subgraph induced on
nodes.
MultiGraph.edge_subgraph(edges) Returns the subgraph induced by the specified edges.

networkx.MultiGraph.copy

MultiGraph.copy(as_view=False)
Returns a copy of the graph.
The copy method by default returns an independent shallow copy of the graph and attributes. That is, if an
attribute is a container, that container is shared by the original an the copy. Use Python’s copy.deepcopy for
new containers.
If as_view is True then a view is returned instead of a copy.

Notes

All copies reproduce the graph structure, but data attributes may be handled in different ways. There are four
types of copies of a graph that people might want.
Deepcopy – A “deepcopy” copies the graph structure as well as all data attributes and any objects they might
contain. The entire graph object is new so that changes in the copy do not affect the original object. (see Python’s
copy.deepcopy)
Data Reference (Shallow) – For a shallow copy the graph structure is copied but the edge, node and graph at-
tribute dicts are references to those in the original graph. This saves time and memory but could cause confusion
if you change an attribute in one graph and it changes the attribute in the other. NetworkX does not provide this
level of shallow copy.
Independent Shallow – This copy creates new independent attribute dicts and then does a shallow copy of the
attributes. That is, any attributes that are containers are shared between the new graph and the original. This is
exactly what dict.copy() provides. You can obtain this style copy using:

>>> G = nx.path_graph(5)
>>> H = G.copy()
>>> H = G.copy(as_view=False)
>>> H = nx.Graph(G)
>>> H = G.__class__(G)

2.2. Basic graph types 89


NetworkX Reference, Release 2.3rc1.dev20190222214247

Fresh Data – For fresh data, the graph structure is copied while new empty data attribute dicts are created. The
resulting graph is independent of the original and it has no edge, node or graph attributes. Fresh copies are not
enabled. Instead use:

>>> H = G.__class__()
>>> H.add_nodes_from(G)
>>> H.add_edges_from(G.edges)

View – Inspired by dict-views, graph-views act like read-only versions of the original graph, providing a copy
of the original structure without requiring any memory for copying the information.
See the Python copy module for more information on shallow and deep copies, https://docs.python.org/2/library/
copy.html.
Parameters as_view (bool, optional (default=False)) – If True, the returned graph-view provides a
read-only view of the original graph without actually copying any data.
Returns G – A copy of the graph.
Return type Graph
See also:

to_directed() return a directed copy of the graph.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> H = G.copy()

networkx.MultiGraph.to_undirected

MultiGraph.to_undirected(as_view=False)
Returns an undirected copy of the graph.
Returns G – A deepcopy of the graph.
Return type Graph/MultiGraph
See also:
copy(), add_edge(), add_edges_from()

Notes

This returns a “deepcopy” of the edge, node, and graph attributes which attempts to completely copy all of the
data and references.
This is in contrast to the similar G = nx.MultiGraph(D) which returns a shallow copy of the data.
See the Python copy module for more information on shallow and deep copies, https://docs.python.org/2/library/
copy.html.
Warning: If you have subclassed MultiiGraph to use dict-like objects in the data structure, those changes do not
transfer to the MultiGraph created by this method.

90 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(2) # or MultiGraph, etc


>>> H = G.to_directed()
>>> list(H.edges)
[(0, 1), (1, 0)]
>>> G2 = H.to_undirected()
>>> list(G2.edges)
[(0, 1)]

networkx.MultiGraph.to_directed

MultiGraph.to_directed(as_view=False)
Returns a directed representation of the graph.
Returns G – A directed graph with the same name, same nodes, and with each edge (u, v, data)
replaced by two directed edges (u, v, data) and (v, u, data).
Return type MultiDiGraph

Notes

This returns a “deepcopy” of the edge, node, and graph attributes which attempts to completely copy all of the
data and references.
This is in contrast to the similar D=DiGraph(G) which returns a shallow copy of the data.
See the Python copy module for more information on shallow and deep copies, https://docs.python.org/2/library/
copy.html.
Warning: If you have subclassed MultiGraph to use dict-like objects in the data structure, those changes do not
transfer to the MultiDiGraph created by this method.

Examples

>>> G = nx.Graph() # or MultiGraph, etc


>>> G.add_edge(0, 1)
>>> H = G.to_directed()
>>> list(H.edges)
[(0, 1), (1, 0)]

If already directed, return a (deep) copy

>>> G = nx.DiGraph() # or MultiDiGraph, etc


>>> G.add_edge(0, 1)
>>> H = G.to_directed()
>>> list(H.edges)
[(0, 1)]

networkx.MultiGraph.subgraph

MultiGraph.subgraph(nodes)
Returns a SubGraph view of the subgraph induced on nodes.

2.2. Basic graph types 91


NetworkX Reference, Release 2.3rc1.dev20190222214247

The induced subgraph of the graph contains the nodes in nodes and the edges between those nodes.
Parameters nodes (list, iterable) – A container of nodes which will be iterated through once.
Returns G – A subgraph view of the graph. The graph structure cannot be changed but node/edge
attributes can and are shared with the original graph.
Return type SubGraph View

Notes

The graph, edge and node attributes are shared with the original graph. Changes to the graph structure is ruled
out by the view, but changes to attributes are reflected in the original graph.
To create a subgraph with its own copy of the edge/node attributes use: G.subgraph(nodes).copy()
For an inplace reduction of a graph to a subgraph you can remove nodes: G.remove_nodes_from([n for n in G
if n not in set(nodes)])
Subgraph views are sometimes NOT what you want. In most cases where you want to do more than simply look
at the induced edges, it makes more sense to just create the subgraph as its own graph with code like:

# Create a subgraph SG based on a (possibly multigraph) G


SG = G.__class__()
SG.add_nodes_from((n, G.nodes[n]) for n in largest_wcc)
if SG.is_multigraph:
SG.add_edges_from((n, nbr, key, d)
for n, nbrs in G.adj.items() if n in largest_wcc
for nbr, keydict in nbrs.items() if nbr in largest_wcc
for key, d in keydict.items())
else:
SG.add_edges_from((n, nbr, d)
for n, nbrs in G.adj.items() if n in largest_wcc
for nbr, d in nbrs.items() if nbr in largest_wcc)
SG.graph.update(G.graph)

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> H = G.subgraph([0, 1, 2])
>>> list(H.edges)
[(0, 1), (1, 2)]

networkx.MultiGraph.edge_subgraph

MultiGraph.edge_subgraph(edges)
Returns the subgraph induced by the specified edges.
The induced subgraph contains each edge in edges and each node incident to any one of those edges.
Parameters edges (iterable) – An iterable of edges in this graph.
Returns G – An edge-induced subgraph of this graph with the same edge attributes.
Return type Graph

92 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The graph, edge, and node attributes in the returned subgraph view are references to the corresponding attributes
in the original graph. The view is read-only.
To create a full graph version of the subgraph with its own copy of the edge or node attributes, use:

>>> G.edge_subgraph(edges).copy()

Examples

>>> G = nx.path_graph(5)
>>> H = G.edge_subgraph([(0, 1), (3, 4)])
>>> list(H.nodes)
[0, 1, 3, 4]
>>> list(H.edges)
[(0, 1), (3, 4)]

2.2.4 MultiDiGraph—Directed graphs with self loops and parallel edges

Overview

class MultiDiGraph(incoming_graph_data=None, **attr)


A directed graph class that can store multiedges.
Multiedges are multiple edges between two nodes. Each edge can hold optional data or attributes.
A MultiDiGraph holds directed edges. Self loops are allowed.
Nodes can be arbitrary (hashable) Python objects with optional key/value attributes. By convention None is not
used as a node.
Edges are represented as links between nodes with optional key/value attributes.
Parameters
• incoming_graph_data (input graph (optional, default: None)) – Data to initialize graph. If
None (default) an empty graph is created. The data can be any format that is supported by
the to_networkx_graph() function, currently including edge list, dict of dicts, dict of lists,
NetworkX graph, NumPy matrix or 2d ndarray, SciPy sparse matrix, or PyGraphviz graph.
• attr (keyword arguments, optional (default= no attributes)) – Attributes to add to graph as
key=value pairs.
See also:
Graph, DiGraph, MultiGraph, OrderedMultiDiGraph

Examples

Create an empty graph structure (a “null graph”) with no nodes and no edges.

>>> G = nx.MultiDiGraph()

2.2. Basic graph types 93


NetworkX Reference, Release 2.3rc1.dev20190222214247

G can be grown in several ways.


Nodes:
Add one node at a time:
>>> G.add_node(1)

Add the nodes from any container (a list, dict, set or even the lines from a file or the nodes from another graph).
>>> G.add_nodes_from([2, 3])
>>> G.add_nodes_from(range(100, 110))
>>> H = nx.path_graph(10)
>>> G.add_nodes_from(H)

In addition to strings and integers any hashable Python object (except None) can represent a node, e.g. a
customized node object, or even another Graph.
>>> G.add_node(H)

Edges:
G can also be grown by adding edges.
Add one edge,
>>> key = G.add_edge(1, 2)

a list of edges,
>>> keys = G.add_edges_from([(1, 2), (1, 3)])

or a collection of edges,
>>> keys = G.add_edges_from(H.edges)

If some edges connect nodes not yet in the graph, the nodes are added automatically. If an edge already exists,
an additional edge is created and stored using a key to identify the edge. By default the key is the lowest unused
integer.
>>> keys = G.add_edges_from([(4,5,dict(route=282)), (4,5,dict(route=37))])
>>> G[4]
AdjacencyView({5: {0: {}, 1: {'route': 282}, 2: {'route': 37}}})

Attributes:
Each graph, node, and edge can hold key/value attribute pairs in an associated attribute dictionary (the keys
must be hashable). By default these are empty, but can be added or changed using add_edge, add_node or direct
manipulation of the attribute dictionaries named graph, node and edge respectively.
>>> G = nx.MultiDiGraph(day="Friday")
>>> G.graph
{'day': 'Friday'}

Add node attributes using add_node(), add_nodes_from() or G.nodes


>>> G.add_node(1, time='5pm')
>>> G.add_nodes_from([3], time='2pm')
>>> G.nodes[1]
(continues on next page)

94 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


{'time': '5pm'}
>>> G.nodes[1]['room'] = 714
>>> del G.nodes[1]['room'] # remove attribute
>>> list(G.nodes(data=True))
[(1, {'time': '5pm'}), (3, {'time': '2pm'})]

Add edge attributes using add_edge(), add_edges_from(), subscript notation, or G.edges.

>>> key = G.add_edge(1, 2, weight=4.7 )


>>> keys = G.add_edges_from([(3, 4), (4, 5)], color='red')
>>> keys = G.add_edges_from([(1,2,{'color':'blue'}), (2,3,{'weight':8})])
>>> G[1][2][0]['weight'] = 4.7
>>> G.edges[1, 2, 0]['weight'] = 4

Warning: we protect the graph data structure by making G.edges[1, 2] a read-only dict-like structure.
However, you can assign to attributes in e.g. G.edges[1, 2]. Thus, use 2 sets of brackets to add/change data
attributes: G.edges[1, 2]['weight'] = 4 (For multigraphs: MG.edges[u, v, key][name] =
value).
Shortcuts:
Many common graph features allow python syntax to speed reporting.

>>> 1 in G # check if node in graph


True
>>> [n for n in G if n<3] # iterate through nodes
[1, 2]
>>> len(G) # number of nodes in graph
5
>>> G[1] # adjacency dict-like view keyed by neighbor to edge attributes
AdjacencyView({2: {0: {'weight': 4}, 1: {'color': 'blue'}}})

Often the best way to traverse all edges of a graph is via the neighbors. The neighbors are available as an
adjacency-view G.adj object or via the method G.adjacency().

>>> for n, nbrsdict in G.adjacency():


... for nbr, keydict in nbrsdict.items():
... for key, eattr in keydict.items():
... if 'weight' in eattr:
... # Do something useful with the edges
... pass

But the edges() method is often more convenient:

>>> for u, v, keys, weight in G.edges(data='weight', keys=True):


... if weight is not None:
... # Do something useful with the edges
... pass

Reporting:
Simple graph information is obtained using methods and object-attributes. Reporting usually provides
views instead of containers to reduce memory usage. The views update as the graph is updated simi-
larly to dict-views. The objects nodes, `edges and adj provide access to data attributes via lookup
(e.g. nodes[n], `edges[u, v], adj[u][v]) and iteration (e.g. nodes.items(), nodes.
data('color'), nodes.data('color', default='blue') and similarly for edges) Views exist
for nodes, edges, neighbors()/adj and degree.

2.2. Basic graph types 95


NetworkX Reference, Release 2.3rc1.dev20190222214247

For details on these and other miscellaneous methods, see below.


Subclasses (Advanced):
The MultiDiGraph class uses a dict-of-dict-of-dict-of-dict structure. The outer dict (node_dict) holds adjacency
information keyed by node. The next dict (adjlist_dict) represents the adjacency information and holds edge_key
dicts keyed by neighbor. The edge_key dict holds each edge_attr dict keyed by edge key. The inner dict
(edge_attr_dict) represents the edge data and holds edge attribute values keyed by attribute names.
Each of these four dicts in the dict-of-dict-of-dict-of-dict structure can be replaced by a user defined dict-like
object. In general, the dict-like features should be maintained but extra features can be added. To replace
one of the dicts create a new graph class by changing the class(!) variable holding the factory for that dict-
like structure. The variable names are node_dict_factory, node_attr_dict_factory, adjlist_inner_dict_factory,
adjlist_outer_dict_factory, edge_key_dict_factory, edge_attr_dict_factory and graph_attr_dict_factory.
node_dict_factory [function, (default: dict)] Factory function to be used to create the dict containing node
attributes, keyed by node id. It should require no arguments and return a dict-like object
node_attr_dict_factory: function, (default: dict) Factory function to be used to create the node attribute dict
which holds attribute values keyed by attribute name. It should require no arguments and return a dict-like
object
adjlist_outer_dict_factory [function, (default: dict)] Factory function to be used to create the outer-most dict
in the data structure that holds adjacency info keyed by node. It should require no arguments and return a
dict-like object.
adjlist_inner_dict_factory [function, (default: dict)] Factory function to be used to create the adjacency list
dict which holds multiedge key dicts keyed by neighbor. It should require no arguments and return a
dict-like object.
edge_key_dict_factory [function, (default: dict)] Factory function to be used to create the edge key dict which
holds edge data keyed by edge key. It should require no arguments and return a dict-like object.
edge_attr_dict_factory [function, (default: dict)] Factory function to be used to create the edge attribute dict
which holds attribute values keyed by attribute name. It should require no arguments and return a dict-like
object.
graph_attr_dict_factory [function, (default: dict)] Factory function to be used to create the graph attribute
dict which holds attribute values keyed by attribute name. It should require no arguments and return a
dict-like object.
Typically, if your extension doesn’t impact the data structure all methods will inherited without issue except:
to_directed/to_undirected. By default these methods create a DiGraph/Graph class and you probably
want them to create your extension of a DiGraph/Graph. To facilitate this we define two class variables that you
can set in your subclass.
to_directed_class [callable, (default: DiGraph or MultiDiGraph)] Class to create a new graph structure in the
to_directed method. If None, a NetworkX class (DiGraph or MultiDiGraph) is used.
to_undirected_class [callable, (default: Graph or MultiGraph)] Class to create a new graph structure in the
to_undirected method. If None, a NetworkX class (Graph or MultiGraph) is used.

Examples

Please see ordered for examples of creating graph subclasses by overwriting the base class dict with a
dictionary-like object.

96 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Methods

Adding and Removing Nodes and Edges

MultiDiGraph.__init__([incoming_graph_data]) Initialize a graph with edges, name, or graph attributes.


MultiDiGraph.add_node(node_for_adding, Add a single node node_for_adding and update
**attr) node attributes.
MultiDiGraph.add_nodes_from(. . . ) Add multiple nodes.
MultiDiGraph.remove_node(n) Remove node n.
MultiDiGraph.remove_nodes_from(nodes) Remove multiple nodes.
MultiDiGraph.add_edge(u_for_edge, Add an edge between u and v.
v_for_edge)
MultiDiGraph.add_edges_from(ebunch_to_add, Add all the edges in ebunch_to_add.
...)
MultiDiGraph.add_weighted_edges_from(. . . [,Add weighted edges in ebunch_to_add with speci-
. . . ]) fied weight attr
MultiDiGraph.new_edge_key(u, v) Returns an unused key for edges between nodes u and
v.
MultiDiGraph.remove_edge(u, v[, key]) Remove an edge between u and v.
MultiDiGraph.remove_edges_from(ebunch) Remove all edges specified in ebunch.
MultiDiGraph.update([edges, nodes]) Update the graph using nodes/edges/graphs as input.
MultiDiGraph.clear() Remove all nodes and edges from the graph.

networkx.MultiDiGraph.__init__

MultiDiGraph.__init__(incoming_graph_data=None, **attr)
Initialize a graph with edges, name, or graph attributes.
Parameters
• incoming_graph_data (input graph) – Data to initialize graph. If incom-
ing_graph_data=None (default) an empty graph is created. The data can be an edge list,
or any NetworkX graph object. If the corresponding optional Python packages are installed
the data can also be a NumPy matrix or 2d ndarray, a SciPy sparse matrix, or a PyGraphviz
graph.
• attr (keyword arguments, optional (default= no attributes)) – Attributes to add to graph as
key=value pairs.
See also:
convert()

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G = nx.Graph(name='my graph')
>>> e = [(1, 2), (2, 3), (3, 4)] # list of edges
>>> G = nx.Graph(e)

Arbitrary graph attribute pairs (key=value) may be assigned

2.2. Basic graph types 97


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.Graph(e, day="Friday")


>>> G.graph
{'day': 'Friday'}

networkx.MultiDiGraph.add_node

MultiDiGraph.add_node(node_for_adding, **attr)
Add a single node node_for_adding and update node attributes.
Parameters
• node_for_adding (node) – A node can be any hashable Python object except None.
• attr (keyword arguments, optional) – Set or change node attributes using key=value.
See also:
add_nodes_from()

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_node(1)
>>> G.add_node('Hello')
>>> K3 = nx.Graph([(0, 1), (1, 2), (2, 0)])
>>> G.add_node(K3)
>>> G.number_of_nodes()
3

Use keywords set/change node attributes:


>>> G.add_node(1, size=10)
>>> G.add_node(3, weight=0.4, UTM=('13S', 382871, 3972649))

Notes

A hashable object is one that can be used as a key in a Python dictionary. This includes strings, numbers, tuples
of strings and numbers, etc.
On many platforms hashable items also include mutables such as NetworkX Graphs, though one should be
careful that the hash doesn’t change on mutables.

networkx.MultiDiGraph.add_nodes_from

MultiDiGraph.add_nodes_from(nodes_for_adding, **attr)
Add multiple nodes.
Parameters
• nodes_for_adding (iterable container) – A container of nodes (list, dict, set, etc.). OR A
container of (node, attribute dict) tuples. Node attributes are updated using the attribute dict.
• attr (keyword arguments, optional (default= no attributes)) – Update attributes for all nodes
in nodes. Node attributes specified in nodes as a tuple take precedence over attributes spec-
ified via keyword arguments.

98 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
add_node()

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_nodes_from('Hello')
>>> K3 = nx.Graph([(0, 1), (1, 2), (2, 0)])
>>> G.add_nodes_from(K3)
>>> sorted(G.nodes(), key=str)
[0, 1, 2, 'H', 'e', 'l', 'o']

Use keywords to update specific node attributes for every node.

>>> G.add_nodes_from([1, 2], size=10)


>>> G.add_nodes_from([3, 4], weight=0.4)

Use (node, attrdict) tuples to update attributes for specific nodes.

>>> G.add_nodes_from([(1, dict(size=11)), (2, {'color':'blue'})])


>>> G.nodes[1]['size']
11
>>> H = nx.Graph()
>>> H.add_nodes_from(G.nodes(data=True))
>>> H.nodes[1]['size']
11

networkx.MultiDiGraph.remove_node

MultiDiGraph.remove_node(n)
Remove node n.
Removes the node n and all adjacent edges. Attempting to remove a non-existent node will raise an exception.
Parameters n (node) – A node in the graph
Raises NetworkXError – If n is not in the graph.
See also:
remove_nodes_from()

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> list(G.edges)
[(0, 1), (1, 2)]
>>> G.remove_node(1)
>>> list(G.edges)
[]

2.2. Basic graph types 99


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiDiGraph.remove_nodes_from

MultiDiGraph.remove_nodes_from(nodes)
Remove multiple nodes.
Parameters nodes (iterable container) – A container of nodes (list, dict, set, etc.). If a node in the
container is not in the graph it is silently ignored.
See also:
remove_node()

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> e = list(G.nodes)
>>> e
[0, 1, 2]
>>> G.remove_nodes_from(e)
>>> list(G.nodes)
[]

networkx.MultiDiGraph.add_edge

MultiDiGraph.add_edge(u_for_edge, v_for_edge, key=None, **attr)


Add an edge between u and v.
The nodes u and v will be automatically added if they are not already in the graph.
Edge attributes can be specified with keywords or by directly accessing the edge’s attribute dictionary. See
examples below.
Parameters
• u_for_edge, v_for_edge (nodes) – Nodes can be, for example, strings or numbers. Nodes
must be hashable (and not None) Python objects.
• key (hashable identifier, optional (default=lowest unused integer)) – Used to distinguish
multiedges between a pair of nodes.
• attr_dict (dictionary, optional (default= no attributes)) – Dictionary of edge attributes.
Key/value pairs will update existing data associated with the edge.
• attr (keyword arguments, optional) – Edge data (or labels or objects) can be assigned using
keyword arguments.
Returns
Return type The edge key assigned to the edge.
See also:

add_edges_from() add a collection of edges

100 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

To replace/update edge data, use the optional key argument to identify a unique edge. Otherwise a new edge
will be created.
NetworkX algorithms designed for weighted graphs cannot use multigraphs directly because it is not clear
how to handle multiedge weights. Convert to Graph using edge attribute ‘weight’ to enable weighted graph
algorithms.
Default keys are generated using the method new_edge_key(). This method can be overridden by subclass-
ing the base class and providing a custom new_edge_key() method.

Examples

The following all add the edge e=(1, 2) to graph G:

>>> G = nx.MultiDiGraph()
>>> e = (1, 2)
>>> key = G.add_edge(1, 2) # explicit two-node form
>>> G.add_edge(*e) # single edge as tuple of two nodes
1
>>> G.add_edges_from( [(1, 2)] ) # add edges from iterable container
[2]

Associate data to edges using keywords:

>>> key = G.add_edge(1, 2, weight=3)


>>> key = G.add_edge(1, 2, key=0, weight=4) # update data for key=0
>>> key = G.add_edge(1, 3, weight=7, capacity=15, length=342.7)

For non-string attribute keys, use subscript notation.

>>> ekey = G.add_edge(1, 2)


>>> G[1][2][0].update({0: 5})
>>> G.edges[1, 2, 0].update({0: 5})

networkx.MultiDiGraph.add_edges_from

MultiDiGraph.add_edges_from(ebunch_to_add, **attr)
Add all the edges in ebunch_to_add.
Parameters
• ebunch_to_add (container of edges) – Each edge given in the container will be added to
the graph. The edges can be:
– 2-tuples (u, v) or
– 3-tuples (u, v, d) for an edge data dict d, or
– 3-tuples (u, v, k) for not iterable key k, or
– 4-tuples (u, v, k, d) for an edge with data and key k
• attr (keyword arguments, optional) – Edge data (or labels or objects) can be assigned using
keyword arguments.
Returns

2.2. Basic graph types 101


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type A list of edge keys assigned to the edges in ebunch.


See also:

add_edge() add a single edge


add_weighted_edges_from() convenient way to add weighted edges

Notes

Adding the same edge twice has no effect but any edge data will be updated when each duplicate edge is added.
Edge attributes specified in an ebunch take precedence over attributes specified via keyword arguments.
Default keys are generated using the method new_edge_key(). This method can be overridden by subclass-
ing the base class and providing a custom new_edge_key() method.

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_edges_from([(0, 1), (1, 2)]) # using a list of edge tuples
>>> e = zip(range(0, 3), range(1, 4))
>>> G.add_edges_from(e) # Add the path graph 0-1-2-3

Associate data to edges

>>> G.add_edges_from([(1, 2), (2, 3)], weight=3)


>>> G.add_edges_from([(3, 4), (1, 4)], label='WN2898')

networkx.MultiDiGraph.add_weighted_edges_from

MultiDiGraph.add_weighted_edges_from(ebunch_to_add, weight=’weight’, **attr)


Add weighted edges in ebunch_to_add with specified weight attr
Parameters
• ebunch_to_add (container of edges) – Each edge given in the list or container will be added
to the graph. The edges must be given as 3-tuples (u, v, w) where w is a number.
• weight (string, optional (default= ‘weight’)) – The attribute name for the edge weights to
be added.
• attr (keyword arguments, optional (default= no attributes)) – Edge attributes to add/update
for all edges.
See also:

add_edge() add a single edge


add_edges_from() add multiple edges

Notes

Adding the same edge twice for Graph/DiGraph simply updates the edge data. For MultiGraph/MultiDiGraph,
duplicate edges are stored.

102 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_weighted_edges_from([(0, 1, 3.0), (1, 2, 7.5)])

networkx.MultiDiGraph.new_edge_key

MultiDiGraph.new_edge_key(u, v)
Returns an unused key for edges between nodes u and v.
The nodes u and v do not need to be already in the graph.

Notes

In the standard MultiGraph class the new key is the number of existing edges between u and v (increased
if necessary to ensure unused). The first edge will have key 0, then 1, etc. If an edge is removed further
new_edge_keys may not be in this order.
Parameters u, v (nodes)
Returns key
Return type int

networkx.MultiDiGraph.remove_edge

MultiDiGraph.remove_edge(u, v, key=None)
Remove an edge between u and v.
Parameters
• u, v (nodes) – Remove an edge between nodes u and v.
• key (hashable identifier, optional (default=None)) – Used to distinguish multiple edges be-
tween a pair of nodes. If None remove a single (arbitrary) edge between u and v.
Raises NetworkXError – If there is not an edge between u and v, or if there is no edge with the
specified key.
See also:

remove_edges_from() remove a collection of edges

Examples

>>> G = nx.MultiDiGraph()
>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.remove_edge(0, 1)
>>> e = (1, 2)
>>> G.remove_edge(*e) # unpacks e from an edge tuple

For multiple edges

2.2. Basic graph types 103


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.MultiDiGraph()
>>> G.add_edges_from([(1, 2), (1, 2), (1, 2)]) # key_list returned
[0, 1, 2]
>>> G.remove_edge(1, 2) # remove a single (arbitrary) edge

For edges with keys

>>> G = nx.MultiDiGraph()
>>> G.add_edge(1, 2, key='first')
'first'
>>> G.add_edge(1, 2, key='second')
'second'
>>> G.remove_edge(1, 2, key='second')

networkx.MultiDiGraph.remove_edges_from

MultiDiGraph.remove_edges_from(ebunch)
Remove all edges specified in ebunch.
Parameters ebunch (list or container of edge tuples) – Each edge given in the list or container will
be removed from the graph. The edges can be:
• 2-tuples (u, v) All edges between u and v are removed.
• 3-tuples (u, v, key) The edge identified by key is removed.
• 4-tuples (u, v, key, data) where data is ignored.
See also:

remove_edge() remove a single edge

Notes

Will fail silently if an edge in ebunch is not in the graph.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> ebunch=[(1, 2), (2, 3)]
>>> G.remove_edges_from(ebunch)

Removing multiple copies of edges

>>> G = nx.MultiGraph()
>>> keys = G.add_edges_from([(1, 2), (1, 2), (1, 2)])
>>> G.remove_edges_from([(1, 2), (1, 2)])
>>> list(G.edges())
[(1, 2)]
>>> G.remove_edges_from([(1, 2), (1, 2)]) # silently ignore extra copy
>>> list(G.edges) # now empty graph
[]

104 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiDiGraph.update

MultiDiGraph.update(edges=None, nodes=None)
Update the graph using nodes/edges/graphs as input.
Like dict.update, this method takes a graph as input, adding the graph’s noes and edges to this graph. It can also
take two inputs: edges and nodes. Finally it can take either edges or nodes. To specify only nodes the keyword
nodes must be used.
The collections of edges and nodes are treated similarly to the add_edges_from/add_nodes_from methods.
When iterated, they should yield 2-tuples (u, v) or 3-tuples (u, v, datadict).
Parameters
• edges (Graph object, collection of edges, or None) – The first parameter can be a graph or
some edges. If it has attributes nodes and edges, then it is taken to be a Graph-like object
and those attributes are used as collections of nodes and edges to be added to the graph. If
the first parameter does not have those attributes, it is treated as a collection of edges and
added to the graph. If the first argument is None, no edges are added.
• nodes (collection of nodes, or None) – The second parameter is treated as a collection of
nodes to be added to the graph unless it is None. If edges is None and nodes is
None an exception is raised. If the first parameter is a Graph, then nodes is ignored.

Examples

>>> G = nx.path_graph(5)
>>> G.update(nx.complete_graph(range(4,10)))
>>> from itertools import combinations
>>> edges = ((u, v, {'power': u * v})
... for u, v in combinations(range(10, 20), 2)
... if u * v < 225)
>>> nodes = [1000] # for singleton, use a container
>>> G.update(edges, nodes)

Notes

It you want to update the graph using an adjacency structure it is straightforward to obtain the edges/nodes from
adjacency. The following examples provide common cases, your adjacency may be slightly different and require
tweaks of these examples.

>>> # dict-of-set/list/tuple
>>> adj = {1: {2, 3}, 2: {1, 3}, 3: {1, 2}}
>>> e = [(u, v) for u, nbrs in adj.items() for v in nbrs]
>>> G.update(edges=e, nodes=adj)

>>> DG = nx.DiGraph()
>>> # dict-of-dict-of-attribute
>>> adj = {1: {2: 1.3, 3: 0.7}, 2: {1: 1.4}, 3: {1: 0.7}}
>>> e = [(u, v, {'weight': d}) for u, nbrs in adj.items()
... for v, d in nbrs.items()]
>>> DG.update(edges=e, nodes=adj)

2.2. Basic graph types 105


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> # dict-of-dict-of-dict
>>> adj = {1: {2: {'weight': 1.3}, 3: {'color': 0.7, 'weight':1.2}}}
>>> e = [(u, v, {'weight': d}) for u, nbrs in adj.items()
... for v, d in nbrs.items()]
>>> DG.update(edges=e, nodes=adj)

>>> # predecessor adjacency (dict-of-set)


>>> pred = {1: {2, 3}, 2: {3}, 3: {3}}
>>> e = [(v, u) for u, nbrs in pred.items() for v in nbrs]

>>> # MultiGraph dict-of-dict-of-dict-of-attribute


>>> MDG = nx.MultiDiGraph()
>>> adj = {1: {2: {0: {'weight': 1.3}, 1: {'weight': 1.2}}},
... 3: {2: {0: {'weight': 0.7}}}}
>>> e = [(u, v, ekey, d) for u, nbrs in adj.items()
... for v, keydict in nbrs.items()
... for ekey, d in keydict.items()]
>>> MDG.update(edges=e)

See also:

add_edges_from() add multiple edges to a graph


add_nodes_from() add multiple nodes to a graph

networkx.MultiDiGraph.clear

MultiDiGraph.clear()
Remove all nodes and edges from the graph.
This also removes the name, and all graph, node, and edge attributes.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.clear()
>>> list(G.nodes)
[]
>>> list(G.edges)
[]

Reporting nodes edges and neighbors

MultiDiGraph.nodes A NodeView of the Graph as G.nodes or G.nodes().


MultiDiGraph.__iter__() Iterate over the nodes.
MultiDiGraph.has_node(n) Returns True if the graph contains the node n.
MultiDiGraph.__contains__(n) Returns True if n is a node, False otherwise.
MultiDiGraph.edges An OutMultiEdgeView of the Graph as G.edges or
G.edges().
Continued on next page

106 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 14 – continued from previous page


MultiDiGraph.out_edges An OutMultiEdgeView of the Graph as G.edges or
G.edges().
MultiDiGraph.in_edges An InMultiEdgeView of the Graph as G.in_edges or
G.in_edges().
MultiDiGraph.has_edge(u, v[, key]) Returns True if the graph has an edge between nodes u
and v.
MultiDiGraph.get_edge_data(u, v[, key, de- Returns the attribute dictionary associated with edge (u,
fault]) v).
MultiDiGraph.neighbors(n) Returns an iterator over successor nodes of n.
MultiDiGraph.adj Graph adjacency object holding the neighbors of each
node.
MultiDiGraph.__getitem__(n) Returns a dict of neighbors of node n.
MultiDiGraph.successors(n) Returns an iterator over successor nodes of n.
MultiDiGraph.succ Graph adjacency object holding the successors of each
node.
MultiDiGraph.predecessors(n) Returns an iterator over predecessor nodes of n.
MultiDiGraph.succ Graph adjacency object holding the successors of each
node.
MultiDiGraph.adjacency() Returns an iterator over (node, adjacency dict) tuples for
all nodes.
MultiDiGraph.nbunch_iter([nbunch]) Returns an iterator over nodes contained in nbunch that
are also in the graph.

networkx.MultiDiGraph.nodes

MultiDiGraph.nodes
A NodeView of the Graph as G.nodes or G.nodes().
Can be used as G.nodes for data lookup and for set-like operations. Can also be used as G.
nodes(data='color', default=None) to return a NodeDataView which reports specific node data
but no set operations. It presents a dict-like interface as well with G.nodes.items() iterating over (node,
nodedata) 2-tuples and G.nodes[3]['foo'] providing the value of the foo attribute for node 3. In
addition, a view G.nodes.data('foo') provides a dict-like interface to the foo attribute of each node.
G.nodes.data('foo', default=1) provides a default for nodes that do not have attribute foo.
Parameters
• data (string or bool, optional (default=False)) – The node attribute returned in 2-tuple (n,
ddict[data]). If True, return entire node attribute dict as (n, ddict). If False, return just the
nodes n.
• default (value, optional (default=None)) – Value used for nodes that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns
Allows set-like operations over the nodes as well as node attribute dict lookup and calling to
get a NodeDataView. A NodeDataView iterates over (n, data) and has no set operations. A
NodeView iterates over n and includes set operations.
When called, if data is False, an iterator over nodes. Otherwise an iterator of 2-tuples (node,
attribute value) where the attribute is specified in data. If data is True then the attribute becomes
the entire data dictionary.
Return type NodeView

2.2. Basic graph types 107


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

If your node data is not needed, it is simpler and equivalent to use the expression for n in G, or list(G).

Examples

There are two simple ways of getting a list of all nodes in the graph:

>>> G = nx.path_graph(3)
>>> list(G.nodes)
[0, 1, 2]
>>> list(G)
[0, 1, 2]

To get the node data along with the nodes:

>>> G.add_node(1, time='5pm')


>>> G.nodes[0]['foo'] = 'bar'
>>> list(G.nodes(data=True))
[(0, {'foo': 'bar'}), (1, {'time': '5pm'}), (2, {})]
>>> list(G.nodes.data())
[(0, {'foo': 'bar'}), (1, {'time': '5pm'}), (2, {})]

>>> list(G.nodes(data='foo'))
[(0, 'bar'), (1, None), (2, None)]
>>> list(G.nodes.data('foo'))
[(0, 'bar'), (1, None), (2, None)]

>>> list(G.nodes(data='time'))
[(0, None), (1, '5pm'), (2, None)]
>>> list(G.nodes.data('time'))
[(0, None), (1, '5pm'), (2, None)]

>>> list(G.nodes(data='time', default='Not Available'))


[(0, 'Not Available'), (1, '5pm'), (2, 'Not Available')]
>>> list(G.nodes.data('time', default='Not Available'))
[(0, 'Not Available'), (1, '5pm'), (2, 'Not Available')]

If some of your nodes have an attribute and the rest are assumed to have a default attribute value you can create
a dictionary from node/attribute pairs using the default keyword argument to guarantee the value is never
None:

>>> G = nx.Graph()
>>> G.add_node(0)
>>> G.add_node(1, weight=2)
>>> G.add_node(2, weight=3)
>>> dict(G.nodes(data='weight', default=1))
{0: 1, 1: 2, 2: 3}

networkx.MultiDiGraph.__iter__

MultiDiGraph.__iter__()
Iterate over the nodes. Use: ‘for n in G’.

108 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns niter – An iterator over all nodes in the graph.


Return type iterator

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> [n for n in G]
[0, 1, 2, 3]
>>> list(G)
[0, 1, 2, 3]

networkx.MultiDiGraph.has_node

MultiDiGraph.has_node(n)
Returns True if the graph contains the node n.
Identical to n in G
Parameters n (node)

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.has_node(0)
True

It is more readable and simpler to use

>>> 0 in G
True

networkx.MultiDiGraph.__contains__

MultiDiGraph.__contains__(n)
Returns True if n is a node, False otherwise. Use: ‘n in G’.

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> 1 in G
True

networkx.MultiDiGraph.edges

MultiDiGraph.edges
An OutMultiEdgeView of the Graph as G.edges or G.edges().
edges(self, nbunch=None, data=False, keys=False, default=None)

2.2. Basic graph types 109


NetworkX Reference, Release 2.3rc1.dev20190222214247

The OutMultiEdgeView provides set-like operations on the edge-tuples as well as edge attribute lookup. When
called, it also provides an EdgeDataView object which allows control of access to edge attributes (but does not
provide set-like operations). Hence, G.edges[u, v]['color'] provides the value of the color attribute
for edge (u, v) while for (u, v, c) in G.edges(data='color', default='red'): iter-
ates through all the edges yielding the color attribute with default 'red' if no color attribute exists.
Edges are returned as tuples with optional data and keys in the order (node, neighbor, key, data).
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• data (string or bool, optional (default=False)) – The edge attribute returned in 3-tuple (u, v,
ddict[data]). If True, return edge attribute dict in 3-tuple (u, v, ddict). If False, return 2-tuple
(u, v).
• keys (bool, optional (default=False)) – If True, return edge keys with each edge.
• default (value, optional (default=None)) – Value used for edges that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns edges – A view of edge attributes, usually it iterates over (u, v) (u, v, k) or (u, v, k, d) tuples
of edges, but can also be used for attribute lookup as edges[u, v, k]['foo'].
Return type EdgeView

Notes

Nodes in nbunch that are not in the graph will be (quietly) ignored. For directed graphs this returns the out-edges.

Examples

>>> G = nx.MultiDiGraph()
>>> nx.add_path(G, [0, 1, 2])
>>> key = G.add_edge(2, 3, weight=5)
>>> [e for e in G.edges()]
[(0, 1), (1, 2), (2, 3)]
>>> list(G.edges(data=True)) # default data is {} (empty dict)
[(0, 1, {}), (1, 2, {}), (2, 3, {'weight': 5})]
>>> list(G.edges(data='weight', default=1))
[(0, 1, 1), (1, 2, 1), (2, 3, 5)]
>>> list(G.edges(keys=True)) # default keys are integers
[(0, 1, 0), (1, 2, 0), (2, 3, 0)]
>>> list(G.edges(data=True, keys=True))
[(0, 1, 0, {}), (1, 2, 0, {}), (2, 3, 0, {'weight': 5})]
>>> list(G.edges(data='weight', default=1, keys=True))
[(0, 1, 0, 1), (1, 2, 0, 1), (2, 3, 0, 5)]
>>> list(G.edges([0, 2]))
[(0, 1), (2, 3)]
>>> list(G.edges(0))
[(0, 1)]

See also:
in_edges, out_edges

110 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiDiGraph.out_edges

MultiDiGraph.out_edges
An OutMultiEdgeView of the Graph as G.edges or G.edges().
edges(self, nbunch=None, data=False, keys=False, default=None)
The OutMultiEdgeView provides set-like operations on the edge-tuples as well as edge attribute lookup. When
called, it also provides an EdgeDataView object which allows control of access to edge attributes (but does not
provide set-like operations). Hence, G.edges[u, v]['color'] provides the value of the color attribute
for edge (u, v) while for (u, v, c) in G.edges(data='color', default='red'): iter-
ates through all the edges yielding the color attribute with default 'red' if no color attribute exists.
Edges are returned as tuples with optional data and keys in the order (node, neighbor, key, data).
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• data (string or bool, optional (default=False)) – The edge attribute returned in 3-tuple (u, v,
ddict[data]). If True, return edge attribute dict in 3-tuple (u, v, ddict). If False, return 2-tuple
(u, v).
• keys (bool, optional (default=False)) – If True, return edge keys with each edge.
• default (value, optional (default=None)) – Value used for edges that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns edges – A view of edge attributes, usually it iterates over (u, v) (u, v, k) or (u, v, k, d) tuples
of edges, but can also be used for attribute lookup as edges[u, v, k]['foo'].
Return type EdgeView

Notes

Nodes in nbunch that are not in the graph will be (quietly) ignored. For directed graphs this returns the out-edges.

Examples

>>> G = nx.MultiDiGraph()
>>> nx.add_path(G, [0, 1, 2])
>>> key = G.add_edge(2, 3, weight=5)
>>> [e for e in G.edges()]
[(0, 1), (1, 2), (2, 3)]
>>> list(G.edges(data=True)) # default data is {} (empty dict)
[(0, 1, {}), (1, 2, {}), (2, 3, {'weight': 5})]
>>> list(G.edges(data='weight', default=1))
[(0, 1, 1), (1, 2, 1), (2, 3, 5)]
>>> list(G.edges(keys=True)) # default keys are integers
[(0, 1, 0), (1, 2, 0), (2, 3, 0)]
>>> list(G.edges(data=True, keys=True))
[(0, 1, 0, {}), (1, 2, 0, {}), (2, 3, 0, {'weight': 5})]
>>> list(G.edges(data='weight', default=1, keys=True))
[(0, 1, 0, 1), (1, 2, 0, 1), (2, 3, 0, 5)]
>>> list(G.edges([0, 2]))
[(0, 1), (2, 3)]
(continues on next page)

2.2. Basic graph types 111


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> list(G.edges(0))
[(0, 1)]

See also:
in_edges, out_edges

networkx.MultiDiGraph.in_edges

MultiDiGraph.in_edges
An InMultiEdgeView of the Graph as G.in_edges or G.in_edges().
in_edges(self, nbunch=None, data=False, keys=False, default=None)
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• data (string or bool, optional (default=False)) – The edge attribute returned in 3-tuple (u, v,
ddict[data]). If True, return edge attribute dict in 3-tuple (u, v, ddict). If False, return 2-tuple
(u, v).
• keys (bool, optional (default=False)) – If True, return edge keys with each edge.
• default (value, optional (default=None)) – Value used for edges that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns in_edges – A view of edge attributes, usually it iterates over (u, v) or (u, v, k) or (u, v, k, d)
tuples of edges, but can also be used for attribute lookup as edges[u, v, k]['foo'].
Return type InMultiEdgeView
See also:
edges

networkx.MultiDiGraph.has_edge

MultiDiGraph.has_edge(u, v, key=None)
Returns True if the graph has an edge between nodes u and v.
This is the same as v in G[u] or key in G[u][v] without KeyError exceptions.
Parameters
• u, v (nodes) – Nodes can be, for example, strings or numbers.
• key (hashable identifier, optional (default=None)) – If specified return True only if the edge
with key is found.
Returns edge_ind – True if edge is in the graph, False otherwise.
Return type bool

112 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

Can be called either using two nodes u, v, an edge tuple (u, v), or an edge tuple (u, v, key).

>>> G = nx.MultiGraph() # or MultiDiGraph


>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.has_edge(0, 1) # using two nodes
True
>>> e = (0, 1)
>>> G.has_edge(*e) # e is a 2-tuple (u, v)
True
>>> G.add_edge(0, 1, key='a')
'a'
>>> G.has_edge(0, 1, key='a') # specify key
True
>>> e=(0, 1, 'a')
>>> G.has_edge(*e) # e is a 3-tuple (u, v, 'a')
True

The following syntax are equivalent:

>>> G.has_edge(0, 1)
True
>>> 1 in G[0] # though this gives :exc:`KeyError` if 0 not in G
True

networkx.MultiDiGraph.get_edge_data

MultiDiGraph.get_edge_data(u, v, key=None, default=None)


Returns the attribute dictionary associated with edge (u, v).
This is identical to G[u][v][key] except the default is returned instead of an exception is the edge doesn’t
exist.
Parameters
• u, v (nodes)
• default (any Python object (default=None)) – Value to return if the edge (u, v) is not found.
• key (hashable identifier, optional (default=None)) – Return data only for the edge with
specified key.
Returns edge_dict – The edge attribute dictionary.
Return type dictionary

Examples

>>> G = nx.MultiGraph() # or MultiDiGraph


>>> key = G.add_edge(0, 1, key='a', weight=7)
>>> G[0][1]['a'] # key='a'
{'weight': 7}
>>> G.edges[0, 1, 'a'] # key='a'
{'weight': 7}

2.2. Basic graph types 113


NetworkX Reference, Release 2.3rc1.dev20190222214247

Warning: we protect the graph data structure by making G.edges and G[1][2] read-only dict-like structures.
However, you can assign values to attributes in e.g. G.edges[1, 2, 'a'] or G[1][2]['a'] using an
additional bracket as shown next. You need to specify all edge info to assign to the edge data associated with an
edge.
>>> G[0][1]['a']['weight'] = 10
>>> G.edges[0, 1, 'a']['weight'] = 10
>>> G[0][1]['a']['weight']
10
>>> G.edges[1, 0, 'a']['weight']
10

>>> G = nx.MultiGraph() # or MultiDiGraph


>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.get_edge_data(0, 1)
{0: {}}
>>> e = (0, 1)
>>> G.get_edge_data(*e) # tuple form
{0: {}}
>>> G.get_edge_data('a', 'b', default=0) # edge not in graph, return 0
0

networkx.MultiDiGraph.neighbors

MultiDiGraph.neighbors(n)
Returns an iterator over successor nodes of n.
A successor of n is a node m such that there exists a directed edge from n to m.
Parameters n (node) – A node in the graph
Raises NetworkXError – If n is not in the graph.
See also:
predecessors()

Notes

neighbors() and successors() are the same.

networkx.MultiDiGraph.adj

MultiDiGraph.adj
Graph adjacency object holding the neighbors of each node.
This object is a read-only dict-like structure with node keys and neighbor-dict values. The neighbor-dict is keyed
by neighbor to the edgekey-dict. So G.adj[3][2][0]['color'] = 'blue' sets the color of the edge
(3, 2, 0) to "blue".
Iterating over G.adj behaves like a dict. Useful idioms include for nbr, datadict in G.adj[n].
items():.
The neighbor information is also provided by subscripting the graph. So for nbr, foovalue in
G[node].data('foo', default=1): works.
For directed graphs, G.adj holds outgoing (successor) info.

114 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiDiGraph.__getitem__

MultiDiGraph.__getitem__(n)
Returns a dict of neighbors of node n. Use: ‘G[n]’.
Parameters n (node) – A node in the graph.
Returns adj_dict – The adjacency dictionary for nodes connected to n.
Return type dictionary

Notes

G[n] is the same as G.adj[n] and similar to G.neighbors(n) (which is an iterator over G.adj[n])

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G[0]
AtlasView({1: {}})

networkx.MultiDiGraph.successors

MultiDiGraph.successors(n)
Returns an iterator over successor nodes of n.
A successor of n is a node m such that there exists a directed edge from n to m.
Parameters n (node) – A node in the graph
Raises NetworkXError – If n is not in the graph.
See also:
predecessors()

Notes

neighbors() and successors() are the same.

networkx.MultiDiGraph.succ

MultiDiGraph.succ
Graph adjacency object holding the successors of each node.
This object is a read-only dict-like structure with node keys and neighbor-dict values. The neighbor-dict is keyed
by neighbor to the edgekey-dict. So G.adj[3][2][0]['color'] = 'blue' sets the color of the edge
(3, 2, 0) to "blue".
Iterating over G.adj behaves like a dict. Useful idioms include for nbr, datadict in G.adj[n].
items():.
The neighbor information is also provided by subscripting the graph. So for nbr, foovalue in
G[node].data('foo', default=1): works.

2.2. Basic graph types 115


NetworkX Reference, Release 2.3rc1.dev20190222214247

For directed graphs, G.succ is identical to G.adj.

networkx.MultiDiGraph.predecessors

MultiDiGraph.predecessors(n)
Returns an iterator over predecessor nodes of n.
A predecessor of n is a node m such that there exists a directed edge from m to n.
Parameters n (node) – A node in the graph
Raises NetworkXError – If n is not in the graph.
See also:
successors()

networkx.MultiDiGraph.adjacency

MultiDiGraph.adjacency()
Returns an iterator over (node, adjacency dict) tuples for all nodes.
For directed graphs, only outgoing neighbors/adjacencies are included.
Returns adj_iter – An iterator over (node, adjacency dictionary) for all nodes in the graph.
Return type iterator

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> [(n, nbrdict) for n, nbrdict in G.adjacency()]
[(0, {1: {}}), (1, {0: {}, 2: {}}), (2, {1: {}, 3: {}}), (3, {2: {}})]

networkx.MultiDiGraph.nbunch_iter

MultiDiGraph.nbunch_iter(nbunch=None)
Returns an iterator over nodes contained in nbunch that are also in the graph.
The nodes in nbunch are checked for membership in the graph and if not are silently ignored.
Parameters nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
Returns niter – An iterator over nodes in nbunch that are also in the graph. If nbunch is None,
iterate over all nodes in the graph.
Return type iterator
Raises NetworkXError – If nbunch is not a node or or sequence of nodes. If a node in nbunch is
not hashable.
See also:
Graph.__iter__()

116 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

When nbunch is an iterator, the returned iterator yields values directly from nbunch, becoming exhausted when
nbunch is exhausted.
To test whether nbunch is a single node, one can use “if nbunch in self:”, even after processing with this routine.
If nbunch is not a node or a (possibly empty) sequence/iterator or None, a NetworkXError is raised. Also, if
any object in nbunch is not hashable, a NetworkXError is raised.

Counting nodes edges and neighbors

MultiDiGraph.order() Returns the number of nodes in the graph.


MultiDiGraph.number_of_nodes() Returns the number of nodes in the graph.
MultiDiGraph.__len__() Returns the number of nodes.
MultiDiGraph.degree A DegreeView for the Graph as G.degree or G.degree().
MultiDiGraph.in_degree A DegreeView for (node, in_degree) or in_degree for
single node.
MultiDiGraph.out_degree Returns an iterator for (node, out-degree) or out-degree
for single node.
MultiDiGraph.size([weight]) Returns the number of edges or total of all edge weights.
MultiDiGraph.number_of_edges([u, v]) Returns the number of edges between two nodes.

networkx.MultiDiGraph.order

MultiDiGraph.order()
Returns the number of nodes in the graph.
Returns nnodes – The number of nodes in the graph.
Return type int
See also:
number_of_nodes(), __len__()

networkx.MultiDiGraph.number_of_nodes

MultiDiGraph.number_of_nodes()
Returns the number of nodes in the graph.
Returns nnodes – The number of nodes in the graph.
Return type int
See also:
order(), __len__()

Examples

>>> G = nx.path_graph(3) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> len(G)
3

2.2. Basic graph types 117


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiDiGraph.__len__

MultiDiGraph.__len__()
Returns the number of nodes. Use: ‘len(G)’.
Returns nnodes – The number of nodes in the graph.
Return type int

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> len(G)
4

networkx.MultiDiGraph.degree

MultiDiGraph.degree
A DegreeView for the Graph as G.degree or G.degree().
The node degree is the number of edges adjacent to the node. The weighted node degree is the sum of the edge
weights for edges incident to that node.
This object provides an iterator for (node, degree) as well as lookup for the degree for a single node.
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• weight (string or None, optional (default=None)) – The name of an edge attribute that holds
the numerical value used as a weight. If None, then each edge has weight 1. The degree is
the sum of the edge weights adjacent to the node.
Returns
• If a single nodes is requested
• deg (int) – Degree of the node
• OR if multiple nodes are requested
• nd_iter (iterator) – The iterator returns two-tuples of (node, degree).
See also:
out_degree, in_degree

Examples

>>> G = nx.MultiDiGraph()
>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.degree(0) # node 0 with degree 1
1
>>> list(G.degree([0, 1, 2]))
[(0, 1), (1, 2), (2, 2)]

118 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiDiGraph.in_degree

MultiDiGraph.in_degree
A DegreeView for (node, in_degree) or in_degree for single node.
The node in-degree is the number of edges pointing in to the node. The weighted node degree is the sum of the
edge weights for edges incident to that node.
This object provides an iterator for (node, degree) as well as lookup for the degree for a single node.
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used as a weight. If None, then each edge has weight 1. The degree is the sum of
the edge weights adjacent to the node.
Returns
• If a single node is requested
• deg (int) – Degree of the node
• OR if multiple nodes are requested
• nd_iter (iterator) – The iterator returns two-tuples of (node, in-degree).
See also:
degree, out_degree

Examples

>>> G = nx.MultiDiGraph()
>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.in_degree(0) # node 0 with degree 0
0
>>> list(G.in_degree([0, 1, 2]))
[(0, 0), (1, 1), (2, 1)]

networkx.MultiDiGraph.out_degree

MultiDiGraph.out_degree
Returns an iterator for (node, out-degree) or out-degree for single node.
out_degree(self, nbunch=None, weight=None)
The node out-degree is the number of edges pointing out of the node. This function returns the out-degree for a
single node or an iterator for a bunch of nodes or if nothing is passed as argument.
Parameters
• nbunch (single node, container, or all nodes (default= all nodes)) – The view will only
report edges incident to these nodes.
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used as a weight. If None, then each edge has weight 1. The degree is the sum of
the edge weights.

2.2. Basic graph types 119


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns
• If a single node is requested
• deg (int) – Degree of the node
• OR if multiple nodes are requested
• nd_iter (iterator) – The iterator returns two-tuples of (node, out-degree).
See also:
degree, in_degree

Examples

>>> G = nx.MultiDiGraph()
>>> nx.add_path(G, [0, 1, 2, 3])
>>> G.out_degree(0) # node 0 with degree 1
1
>>> list(G.out_degree([0, 1, 2]))
[(0, 1), (1, 1), (2, 1)]

networkx.MultiDiGraph.size

MultiDiGraph.size(weight=None)
Returns the number of edges or total of all edge weights.
Parameters weight (string or None, optional (default=None)) – The edge attribute that holds the
numerical value used as a weight. If None, then each edge has weight 1.
Returns
size – The number of edges or (if weight keyword is provided) the total weight sum.
If weight is None, returns an int. Otherwise a float (or more general numeric if the weights are
more general).
Return type numeric
See also:
number_of_edges()

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.size()
3

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_edge('a', 'b', weight=2)
>>> G.add_edge('b', 'c', weight=4)
>>> G.size()
2
>>> G.size(weight='weight')
6.0

120 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiDiGraph.number_of_edges

MultiDiGraph.number_of_edges(u=None, v=None)
Returns the number of edges between two nodes.
Parameters u, v (nodes, optional (Gefault=all edges)) – If u and v are specified, return the number
of edges between u and v. Otherwise return the total number of all edges.
Returns nedges – The number of edges in the graph. If nodes u and v are specified return the
number of edges between those nodes. If the graph is directed, this only returns the number of
edges from u to v.
Return type int
See also:
size()

Examples

For undirected multigraphs, this method counts the total number of edges in the graph:

>>> G = nx.MultiGraph()
>>> G.add_edges_from([(0, 1), (0, 1), (1, 2)])
[0, 1, 0]
>>> G.number_of_edges()
3

If you specify two nodes, this counts the total number of edges joining the two nodes:

>>> G.number_of_edges(0, 1)
2

For directed multigraphs, this method can count the total number of directed edges from u to v:

>>> G = nx.MultiDiGraph()
>>> G.add_edges_from([(0, 1), (0, 1), (1, 0)])
[0, 1, 0]
>>> G.number_of_edges(0, 1)
2
>>> G.number_of_edges(1, 0)
1

Making copies and subgraphs

MultiDiGraph.copy([as_view]) Returns a copy of the graph.


MultiDiGraph.to_undirected([reciprocal, Returns an undirected representation of the digraph.
as_view])
MultiDiGraph.to_directed([as_view]) Returns a directed representation of the graph.
MultiDiGraph.subgraph(nodes) Returns a SubGraph view of the subgraph induced on
nodes.
MultiDiGraph.edge_subgraph(edges) Returns the subgraph induced by the specified edges.
MultiDiGraph.reverse([copy]) Returns the reverse of the graph.

2.2. Basic graph types 121


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiDiGraph.copy

MultiDiGraph.copy(as_view=False)
Returns a copy of the graph.
The copy method by default returns an independent shallow copy of the graph and attributes. That is, if an
attribute is a container, that container is shared by the original an the copy. Use Python’s copy.deepcopy for
new containers.
If as_view is True then a view is returned instead of a copy.

Notes

All copies reproduce the graph structure, but data attributes may be handled in different ways. There are four
types of copies of a graph that people might want.
Deepcopy – A “deepcopy” copies the graph structure as well as all data attributes and any objects they might
contain. The entire graph object is new so that changes in the copy do not affect the original object. (see Python’s
copy.deepcopy)
Data Reference (Shallow) – For a shallow copy the graph structure is copied but the edge, node and graph at-
tribute dicts are references to those in the original graph. This saves time and memory but could cause confusion
if you change an attribute in one graph and it changes the attribute in the other. NetworkX does not provide this
level of shallow copy.
Independent Shallow – This copy creates new independent attribute dicts and then does a shallow copy of the
attributes. That is, any attributes that are containers are shared between the new graph and the original. This is
exactly what dict.copy() provides. You can obtain this style copy using:

>>> G = nx.path_graph(5)
>>> H = G.copy()
>>> H = G.copy(as_view=False)
>>> H = nx.Graph(G)
>>> H = G.__class__(G)

Fresh Data – For fresh data, the graph structure is copied while new empty data attribute dicts are created. The
resulting graph is independent of the original and it has no edge, node or graph attributes. Fresh copies are not
enabled. Instead use:

>>> H = G.__class__()
>>> H.add_nodes_from(G)
>>> H.add_edges_from(G.edges)

View – Inspired by dict-views, graph-views act like read-only versions of the original graph, providing a copy
of the original structure without requiring any memory for copying the information.
See the Python copy module for more information on shallow and deep copies, https://docs.python.org/2/library/
copy.html.
Parameters as_view (bool, optional (default=False)) – If True, the returned graph-view provides a
read-only view of the original graph without actually copying any data.
Returns G – A copy of the graph.
Return type Graph
See also:

to_directed() return a directed copy of the graph.

122 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> H = G.copy()

networkx.MultiDiGraph.to_undirected

MultiDiGraph.to_undirected(reciprocal=False, as_view=False)
Returns an undirected representation of the digraph.
Parameters
• reciprocal (bool (optional)) – If True only keep edges that appear in both directions in the
original digraph.
• as_view (bool (optional, default=False)) – If True return an undirected view of the original
directed graph.
Returns G – An undirected graph with the same name and nodes and with edge (u, v, data) if either
(u, v, data) or (v, u, data) is in the digraph. If both edges exist in digraph and their edge data is
different, only one edge is created with an arbitrary choice of which edge data to use. You must
check and correct for this manually if desired.
Return type MultiGraph
See also:
MultiGraph(), copy(), add_edge(), add_edges_from()

Notes

This returns a “deepcopy” of the edge, node, and graph attributes which attempts to completely copy all of the
data and references.
This is in contrast to the similar D=MultiiGraph(G) which returns a shallow copy of the data.
See the Python copy module for more information on shallow and deep copies, https://docs.python.org/2/library/
copy.html.
Warning: If you have subclassed MultiDiGraph to use dict-like objects in the data structure, those changes do
not transfer to the MultiGraph created by this method.

Examples

>>> G = nx.path_graph(2) # or MultiGraph, etc


>>> H = G.to_directed()
>>> list(H.edges)
[(0, 1), (1, 0)]
>>> G2 = H.to_undirected()
>>> list(G2.edges)
[(0, 1)]

2.2. Basic graph types 123


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.MultiDiGraph.to_directed

MultiDiGraph.to_directed(as_view=False)
Returns a directed representation of the graph.
Returns G – A directed graph with the same name, same nodes, and with each edge (u, v, data)
replaced by two directed edges (u, v, data) and (v, u, data).
Return type MultiDiGraph

Notes

This returns a “deepcopy” of the edge, node, and graph attributes which attempts to completely copy all of the
data and references.
This is in contrast to the similar D=DiGraph(G) which returns a shallow copy of the data.
See the Python copy module for more information on shallow and deep copies, https://docs.python.org/2/library/
copy.html.
Warning: If you have subclassed MultiGraph to use dict-like objects in the data structure, those changes do not
transfer to the MultiDiGraph created by this method.

Examples

>>> G = nx.Graph() # or MultiGraph, etc


>>> G.add_edge(0, 1)
>>> H = G.to_directed()
>>> list(H.edges)
[(0, 1), (1, 0)]

If already directed, return a (deep) copy

>>> G = nx.DiGraph() # or MultiDiGraph, etc


>>> G.add_edge(0, 1)
>>> H = G.to_directed()
>>> list(H.edges)
[(0, 1)]

networkx.MultiDiGraph.subgraph

MultiDiGraph.subgraph(nodes)
Returns a SubGraph view of the subgraph induced on nodes.
The induced subgraph of the graph contains the nodes in nodes and the edges between those nodes.
Parameters nodes (list, iterable) – A container of nodes which will be iterated through once.
Returns G – A subgraph view of the graph. The graph structure cannot be changed but node/edge
attributes can and are shared with the original graph.
Return type SubGraph View

124 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The graph, edge and node attributes are shared with the original graph. Changes to the graph structure is ruled
out by the view, but changes to attributes are reflected in the original graph.
To create a subgraph with its own copy of the edge/node attributes use: G.subgraph(nodes).copy()
For an inplace reduction of a graph to a subgraph you can remove nodes: G.remove_nodes_from([n for n in G
if n not in set(nodes)])
Subgraph views are sometimes NOT what you want. In most cases where you want to do more than simply look
at the induced edges, it makes more sense to just create the subgraph as its own graph with code like:

# Create a subgraph SG based on a (possibly multigraph) G


SG = G.__class__()
SG.add_nodes_from((n, G.nodes[n]) for n in largest_wcc)
if SG.is_multigraph:
SG.add_edges_from((n, nbr, key, d)
for n, nbrs in G.adj.items() if n in largest_wcc
for nbr, keydict in nbrs.items() if nbr in largest_wcc
for key, d in keydict.items())
else:
SG.add_edges_from((n, nbr, d)
for n, nbrs in G.adj.items() if n in largest_wcc
for nbr, d in nbrs.items() if nbr in largest_wcc)
SG.graph.update(G.graph)

Examples

>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> H = G.subgraph([0, 1, 2])
>>> list(H.edges)
[(0, 1), (1, 2)]

networkx.MultiDiGraph.edge_subgraph

MultiDiGraph.edge_subgraph(edges)
Returns the subgraph induced by the specified edges.
The induced subgraph contains each edge in edges and each node incident to any one of those edges.
Parameters edges (iterable) – An iterable of edges in this graph.
Returns G – An edge-induced subgraph of this graph with the same edge attributes.
Return type Graph

Notes

The graph, edge, and node attributes in the returned subgraph view are references to the corresponding attributes
in the original graph. The view is read-only.
To create a full graph version of the subgraph with its own copy of the edge or node attributes, use:

>>> G.edge_subgraph(edges).copy()

2.2. Basic graph types 125


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(5)
>>> H = G.edge_subgraph([(0, 1), (3, 4)])
>>> list(H.nodes)
[0, 1, 3, 4]
>>> list(H.edges)
[(0, 1), (3, 4)]

networkx.MultiDiGraph.reverse

MultiDiGraph.reverse(copy=True)
Returns the reverse of the graph.
The reverse is a graph with the same nodes and edges but with the directions of the edges reversed.
Parameters copy (bool optional (default=True)) – If True, return a new DiGraph holding the re-
versed edges. If False, the reverse graph is created using a view of the original graph.

2.2.5 Ordered Graphs—Consistently ordered graphs

Consistently ordered variants of the default base classes. Note that if you are using Python 3.6+, you shouldn’t need
these classes because the dicts in Python 3.6+ are ordered. Note also that there are many differing expectations for the
word “ordered” and that these classes may not provide the order you expect. The intent here is to give a consistent
order not a particular order.
The Ordered (Di/Multi/MultiDi) Graphs give a consistent order for reporting of nodes and edges. The order of node
reporting agrees with node adding, but for edges, the order is not necessarily the order that the edges were added.
In general, you should use the default (i.e., unordered) graph classes. However, there are times (e.g., when testing)
when you may need the order preserved.
Special care is required when using subgraphs of the Ordered classes. The order of nodes in the subclass is not
necessarily the same order as the original class. In general it is probably better to avoid using subgraphs and replace
with code similar to:
# instead of SG = G.subgraph(ordered_nodes) SG=nx.OrderedGraph()
SG.add_nodes_from(ordered_nodes) SG.add_edges_from((u, v) for (u, v) in G.edges() if u in SG
if v in SG)
class OrderedGraph(incoming_graph_data=None, **attr)
Consistently ordered variant of Graph.
class OrderedDiGraph(incoming_graph_data=None, **attr)
Consistently ordered variant of DiGraph.
class OrderedMultiGraph(incoming_graph_data=None, **attr)
Consistently ordered variant of MultiGraph.
class OrderedMultiDiGraph(incoming_graph_data=None, **attr)
Consistently ordered variant of MultiDiGraph.

Note: NetworkX uses dicts to store the nodes and neighbors in a graph. So the reporting of nodes and edges for
the base graph classes will not necessarily be consistent across versions and platforms. If you need the order of nodes
and edges to be consistent (e.g., when writing automated tests), please see OrderedGraph, OrderedDiGraph,

126 Chapter 2. Graph types


NetworkX Reference, Release 2.3rc1.dev20190222214247

OrderedMultiGraph, or OrderedMultiDiGraph, which behave like the base graph classes but give a consis-
tent order for reporting of nodes and edges.

2.3 Graph Views

View of Graphs as SubGraph, Reverse, Directed, Undirected.


In some algorithms it is convenient to temporarily morph a graph to exclude some nodes or edges. It should be better
to do that via a view than to remove and then re-add. In other algorithms it is convenient to temporarily morph a graph
to reverse directed edges, or treat a directed graph as undirected, etc. This module provides those graph views.
The resulting views are essentially read-only graphs that report data from the orignal graph object. We provide an
attribute G._graph which points to the underlying graph object.
Note: Since graphviews look like graphs, one can end up with view-of-view-of-view chains. Be careful with chains
because they become very slow with about 15 nested views. For the common simple case of node induced subgraphs
created from the graph class, we short-cut the chain by returning a subgraph of the original graph directly rather than
a subgraph of a subgraph. We are careful not to disrupt any edge filter in the middle subgraph. In general, determining
how to short-cut the chain is tricky and much harder with restricted_views than with induced subgraphs. Often it is
easiest to use .copy() to avoid chains.

generic_graph_view(G[, create_using])
subgraph_view(G[, filter_node, filter_edge])
reverse_view(G)

2.3.1 networkx.classes.graphviews.generic_graph_view

generic_graph_view(G, create_using=None)

2.3.2 networkx.classes.graphviews.subgraph_view

subgraph_view(G, filter_node=<function no_filter>, filter_edge=<function no_filter>)

2.3.3 networkx.classes.graphviews.reverse_view

reverse_view(G)

2.3. Graph Views 127


NetworkX Reference, Release 2.3rc1.dev20190222214247

128 Chapter 2. Graph types


CHAPTER

THREE

ALGORITHMS

3.1 Approximations and Heuristics

Approximations of graph properties and Heuristic functions for optimization problems.

Warning: The approximation submodule is not imported in the top-level networkx.

These functions can be imported with from networkx.algorithms import


approximation.

3.1.1 Connectivity

Fast approximation for node connectivity

all_pairs_node_connectivity(G[, nbunch, Compute node connectivity between all pairs of nodes.


cutoff])
local_node_connectivity(G, source, target[, Compute node connectivity between source and target.
. . . ])
node_connectivity(G[, s, t]) Returns an approximation for node connectivity for a
graph or digraph G.

networkx.algorithms.approximation.connectivity.all_pairs_node_connectivity

all_pairs_node_connectivity(G, nbunch=None, cutoff=None)


Compute node connectivity between all pairs of nodes.
Pairwise or local node connectivity between two distinct and nonadjacent nodes is the minimum number of
nodes that must be removed (minimum separating cutset) to disconnect them. By Menger’s theorem, this is
equal to the number of node independent paths (paths that share no nodes other than source and target). Which
is what we compute in this function.
This algorithm is a fast approximation that gives an strict lower bound on the actual number of node independent
paths between two nodes1 . It works for both directed and undirected graphs.
Parameters
• G (NetworkX graph)
1 White, Douglas R., and Mark Newman. 2001 A Fast Algorithm for Node-Independent Paths. Santa Fe Institute Working Paper #01-07-035
http://eclectic.ss.uci.edu/~drwhite/working.pdf

129
NetworkX Reference, Release 2.3rc1.dev20190222214247

• nbunch (container) – Container of nodes. If provided node connectivity will be computed


only over pairs of nodes in nbunch.
• cutoff (integer) – Maximum node connectivity to consider. If None, the minimum degree
of source or target is used as a cutoff in each pair of nodes. Default value None.
Returns K – Dictionary, keyed by source and target, of pairwise node connectivity
Return type dictionary
See also:
local_node_connectivity(), all_pairs_node_connectivity()

References

networkx.algorithms.approximation.connectivity.local_node_connectivity

local_node_connectivity(G, source, target, cutoff=None)


Compute node connectivity between source and target.
Pairwise or local node connectivity between two distinct and nonadjacent nodes is the minimum number of
nodes that must be removed (minimum separating cutset) to disconnect them. By Menger’s theorem, this is
equal to the number of node independent paths (paths that share no nodes other than source and target). Which
is what we compute in this function.
This algorithm is a fast approximation that gives an strict lower bound on the actual number of node independent
paths between two nodes1 . It works for both directed and undirected graphs.
Parameters
• G (NetworkX graph)
• source (node) – Starting node for node connectivity
• target (node) – Ending node for node connectivity
• cutoff (integer) – Maximum node connectivity to consider. If None, the minimum degree
of source or target is used as a cutoff. Default value None.
Returns k – pairwise node connectivity
Return type integer

Examples

>>> # Platonic octahedral graph has node connectivity 4


>>> # for each non adjacent node pair
>>> from networkx.algorithms import approximation as approx
>>> G = nx.octahedral_graph()
>>> approx.local_node_connectivity(G, 0, 5)
4

1 White, Douglas R., and Mark Newman. 2001 A Fast Algorithm for Node-Independent Paths. Santa Fe Institute Working Paper #01-07-035

http://eclectic.ss.uci.edu/~drwhite/working.pdf

130 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This algorithm1 finds node independents paths between two nodes by computing their shortest path using BFS,
marking the nodes of the path found as ‘used’ and then searching other shortest paths excluding the nodes
marked as used until no more paths exist. It is not exact because a shortest path could use nodes that, if the path
were longer, may belong to two different node independent paths. Thus it only guarantees an strict lower bound
on node connectivity.
Note that the authors propose a further refinement, losing accuracy and gaining speed, which is not implemented
yet.
See also:
all_pairs_node_connectivity(), node_connectivity()

References

networkx.algorithms.approximation.connectivity.node_connectivity

node_connectivity(G, s=None, t=None)


Returns an approximation for node connectivity for a graph or digraph G.
Node connectivity is equal to the minimum number of nodes that must be removed to disconnect G or render it
trivial. By Menger’s theorem, this is equal to the number of node independent paths (paths that share no nodes
other than source and target).
If source and target nodes are provided, this function returns the local node connectivity: the minimum number
of nodes that must be removed to break all paths from source to target in G.
This algorithm is based on a fast approximation that gives an strict lower bound on the actual number of node
independent paths between two nodes1 . It works for both directed and undirected graphs.
Parameters
• G (NetworkX graph) – Undirected graph
• s (node) – Source node. Optional. Default value: None.
• t (node) – Target node. Optional. Default value: None.
Returns K – Node connectivity of G, or local node connectivity if source and target are provided.
Return type integer

Examples

>>> # Platonic octahedral graph is 4-node-connected


>>> from networkx.algorithms import approximation as approx
>>> G = nx.octahedral_graph()
>>> approx.node_connectivity(G)
4

1 White, Douglas R., and Mark Newman. 2001 A Fast Algorithm for Node-Independent Paths. Santa Fe Institute Working Paper #01-07-035

http://eclectic.ss.uci.edu/~drwhite/working.pdf

3.1. Approximations and Heuristics 131


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This algorithm1 finds node independents paths between two nodes by computing their shortest path using BFS,
marking the nodes of the path found as ‘used’ and then searching other shortest paths excluding the nodes
marked as used until no more paths exist. It is not exact because a shortest path could use nodes that, if the path
were longer, may belong to two different node independent paths. Thus it only guarantees an strict lower bound
on node connectivity.
See also:
all_pairs_node_connectivity(), local_node_connectivity()

References

3.1.2 K-components

Fast approximation for k-component structure

k_components(G[, min_density]) Returns the approximate k-component structure of a


graph G.

networkx.algorithms.approximation.kcomponents.k_components

k_components(G, min_density=0.95)
Returns the approximate k-component structure of a graph G.
A k-component is a maximal subgraph of a graph G that has, at least, node connectivity k: we need to remove at
least k nodes to break it into more components. k-components have an inherent hierarchical structure because
they are nested in terms of connectivity: a connected graph can contain several 2-components, each of which
can contain one or more 3-components, and so forth.
This implementation is based on the fast heuristics to approximate the k-component structure of a graph1 .
Which, in turn, it is based on a fast approximation algorithm for finding good lower bounds of the number of
node independent paths between two nodes2 .
Parameters
• G (NetworkX graph) – Undirected graph
• min_density (Float) – Density relaxation threshold. Default value 0.95
Returns k_components – Dictionary with connectivity level k as key and a list of sets of nodes that
form a k-component of level k as values.
Return type dict

Examples

>>> # Petersen graph has 10 nodes and it is triconnected, thus all


>>> # nodes are in a single component on all three connectivity levels
>>> from networkx.algorithms import approximation as apxa
(continues on next page)
1 Torrents, J. and F. Ferraro (2015) Structural Cohesion: Visualization and Heuristics for Fast Computation. https://arxiv.org/pdf/1503.04476v1
2 White, Douglas R., and Mark Newman (2001) A Fast Algorithm for Node-Independent Paths. Santa Fe Institute Working Paper #01-07-035
http://eclectic.ss.uci.edu/~drwhite/working.pdf

132 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> G = nx.petersen_graph()
>>> k_components = apxa.k_components(G)

Notes

The logic of the approximation algorithm for computing the k-component structure1 is based on repeatedly
applying simple and fast algorithms for k-cores and biconnected components in order to narrow down the
number of pairs of nodes over which we have to compute White and Newman’s approximation algorithm for
finding node independent paths2 . More formally, this algorithm is based on Whitney’s theorem, which states
an inclusion relation among node connectivity, edge connectivity, and minimum degree for any graph G. This
theorem implies that every k-component is nested inside a k-edge-component, which in turn, is contained in a
k-core. Thus, this algorithm computes node independent paths among pairs of nodes in each biconnected part
of each k-core, and repeats this procedure for each k from 3 to the maximal core number of a node in the input
graph.
Because, in practice, many nodes of the core of level k inside a bicomponent actually are part of a component
of level k, the auxiliary graph needed for the algorithm is likely to be very dense. Thus, we use a complement
graph data structure (see AntiGraph) to save memory. AntiGraph only stores information of the edges that
are not present in the actual auxiliary graph. When applying algorithms to this complement graph data structure,
it behaves as if it were the dense version.
See also:
k_components()

References

3.1.3 Clique

Functions for computing large cliques.

max_clique(G) Find the Maximum Clique


clique_removal(G) Repeatedly remove cliques from the graph.
large_clique_size(G) Find the size of a large clique in a graph.

networkx.algorithms.approximation.clique.max_clique

max_clique(G)
Find the Maximum Clique
Finds the 𝑂(|𝑉 |/(𝑙𝑜𝑔|𝑉 |)2 ) apx of maximum clique/independent set in the worst case.
Parameters G (NetworkX graph) – Undirected graph
Returns clique – The apx-maximum clique of the graph
Return type set

Notes

A clique in an undirected graph G = (V, E) is a subset of the vertex set C subseteq V such that for every two
vertices in C there exists an edge connecting the two. This is equivalent to saying that the subgraph induced by

3.1. Approximations and Heuristics 133


NetworkX Reference, Release 2.3rc1.dev20190222214247

C is complete (in some cases, the term clique may also refer to the subgraph).
A maximum clique is a clique of the largest possible size in a given graph. The clique number omega(G) of
a graph G is the number of vertices in a maximum clique in G. The intersection number of G is the smallest
number of cliques that together cover all edges of G.
https://en.wikipedia.org/wiki/Maximum_clique

References

networkx.algorithms.approximation.clique.clique_removal

clique_removal(G)
Repeatedly remove cliques from the graph.
Results in a 𝑂(|𝑉 |/(log |𝑉 |)2 ) approximation of maximum clique and independent set. Returns the largest
independent set found, along with found maximal cliques.
Parameters G (NetworkX graph) – Undirected graph
Returns max_ind_cliques – 2-tuple of Maximal Independent Set and list of maximal cliques (sets).
Return type (set, list) tuple

References

networkx.algorithms.approximation.clique.large_clique_size

large_clique_size(G)
Find the size of a large clique in a graph.
A clique is a subset of nodes in which each pair of nodes is adjacent. This function is a heuristic for finding the
size of a large clique in the graph.
Parameters G (NetworkX graph)
Returns The size of a large clique in the graph.
Return type int

Notes

This implementation is from1 . Its worst case time complexity is 𝑂(𝑛𝑑2 ), where n is the number of nodes in the
graph and d is the maximum degree.
This function is a heuristic, which means it may work well in practice, but there is no rigorous mathematical
guarantee on the ratio between the returned number and the actual largest clique size in the graph.

References

See also:

networkx.algorithms.approximation.clique.max_clique() A function that returns an ap-


proximate maximum clique with a guarantee on the approximation ratio.
1 Pattabiraman, Bharath, et al. “Fast Algorithms for the Maximum Clique Problem on Massive Graphs with Applications to Overlapping

Community Detection.” Internet Mathematics 11.4-5 (2015): 421–448. <https://doi.org/10.1080/15427951.2014.986778>

134 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.clique Functions for finding the exact maximum clique in a graph.

3.1.4 Clustering

average_clustering(G[, trials, seed]) Estimates the average clustering coefficient of G.

networkx.algorithms.approximation.clustering_coefficient.average_clustering

average_clustering(G, trials=1000, seed=None)


Estimates the average clustering coefficient of G.
The local clustering of each node in G is the fraction of triangles that actually exist over all possible triangles in
its neighborhood. The average clustering coefficient of a graph G is the mean of local clusterings.
This function finds an approximate average clustering coefficient for G by repeating n times (defined in
trials) the following experiment: choose a node at random, choose two of its neighbors at random, and
check if they are connected. The approximate coefficient is the fraction of triangles found over the number of
trials1 .
Parameters
• G (NetworkX graph)
• trials (integer) – Number of trials to perform (default 1000).
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns c – Approximated average clustering coefficient.
Return type float

References

3.1.5 Dominating Set

Functions for finding node and edge dominating sets.


A dominating set for an undirected graph G with vertex set V and edge set E is a subset D of V such that every vertex
not in D is adjacent to at least one member of D. An edge dominating set is a subset F of E such that every edge not
in F is incident to an endpoint of at least one edge in F.

min_weighted_dominating_set(G[, weight]) Returns a dominating set that approximates the mini-


mum weight node dominating set.
min_edge_dominating_set(G) Returns minimum cardinality edge dominating set.

networkx.algorithms.approximation.dominating_set.min_weighted_dominating_set

min_weighted_dominating_set(G, weight=None)
Returns a dominating set that approximates the minimum weight node dominating set.
Parameters
1Schank, Thomas, and Dorothea Wagner. Approximating clustering coefficient and transitivity. Universität Karlsruhe, Fakultät für Informatik,
2004. http://www.emis.ams.org/journals/JGAA/accepted/2005/SchankWagner2005.9.2.pdf

3.1. Approximations and Heuristics 135


NetworkX Reference, Release 2.3rc1.dev20190222214247

• G (NetworkX graph) – Undirected graph.


• weight (string) – The node attribute storing the weight of an node. If provided, the node
attribute with this key must be a number for each node. If not provided, each node is assumed
to have weight one.
Returns min_weight_dominating_set – A set of nodes, the sum of whose weights is no more than
(log w(V)) w(V^*), where w(V) denotes the sum of the weights of each node in the graph
and w(V^*) denotes the sum of the weights of each node in the minimum weight dominating
set.
Return type set

Notes

This algorithm computes an approximate minimum weighted dominating set for the graph G. The returned
solution has weight (log w(V)) w(V^*), where w(V) denotes the sum of the weights of each node in the
graph and w(V^*) denotes the sum of the weights of each node in the minimum weight dominating set for the
graph.
This implementation of the algorithm runs in 𝑂(𝑚) time, where 𝑚 is the number of edges in the graph.

References

networkx.algorithms.approximation.dominating_set.min_edge_dominating_set

min_edge_dominating_set(G)
Returns minimum cardinality edge dominating set.
Parameters G (NetworkX graph) – Undirected graph
Returns min_edge_dominating_set – Returns a set of dominating edges whose size is no more
than 2 * OPT.
Return type set

Notes

The algorithm computes an approximate solution to the edge dominating set problem. The result is no more
than 2 * OPT in terms of size of the set. Runtime of the algorithm is 𝑂(|𝐸|).

3.1.6 Independent Set

Independent Set
Independent set or stable set is a set of vertices in a graph, no two of which are adjacent. That is, it is a set I of vertices
such that for every two vertices in I, there is no edge connecting the two. Equivalently, each edge in the graph has at
most one endpoint in I. The size of an independent set is the number of vertices it contains.
A maximum independent set is a largest independent set for a given graph G and its size is denoted 𝛼(G). The problem
of finding such a set is called the maximum independent set problem and is an NP-hard optimization problem. As
such, it is unlikely that there exists an efficient algorithm for finding a maximum independent set of a graph.
Wikipedia: Independent set
Independent set algorithm is based on the following paper:

136 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

𝑂(|𝑉 |/(𝑙𝑜𝑔|𝑉 |)2 ) apx of maximum clique/independent set.


Boppana, R., & Halldórsson, M. M. (1992). Approximating maximum independent sets by excluding subgraphs. BIT
Numerical Mathematics, 32(2), 180–196. Springer. doi:10.1007/BF01994876

maximum_independent_set(G) Returns an approximate maximum independent set.

networkx.algorithms.approximation.independent_set.maximum_independent_set

maximum_independent_set(G)
Returns an approximate maximum independent set.
Parameters G (NetworkX graph) – Undirected graph
Returns iset – The apx-maximum independent set
Return type Set

Notes

Finds the 𝑂(|𝑉 |/(𝑙𝑜𝑔|𝑉 |)2 ) apx of independent set in the worst case.

References

3.1.7 Matching

Graph Matching

Given a graph G = (V,E), a matching M in G is a set of pairwise non-adjacent edges; that is, no two edges share a
common vertex.
Wikipedia: Matching

min_maximal_matching(G) Returns the minimum maximal matching of G.

networkx.algorithms.approximation.matching.min_maximal_matching

min_maximal_matching(G)
Returns the minimum maximal matching of G. That is, out of all maximal matchings of the graph G, the smallest
is returned.
Parameters G (NetworkX graph) – Undirected graph
Returns min_maximal_matching – Returns a set of edges such that no two edges share a common
endpoint and every edge not in the set shares some common endpoint in the set. Cardinality will
be 2*OPT in the worst case.
Return type set

Notes

The algorithm computes an approximate solution fo the minimum maximal cardinality matching problem. The
solution is no more than 2 * OPT in size. Runtime is 𝑂(|𝐸|).

3.1. Approximations and Heuristics 137


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.1.8 Ramsey

Ramsey numbers.

ramsey_R2(G) Approximately computes the Ramsey number R(2;s,


t) for graph.

networkx.algorithms.approximation.ramsey.ramsey_R2

ramsey_R2(G)
Approximately computes the Ramsey number R(2;s,t) for graph.
Parameters G (NetworkX graph) – Undirected graph
Returns max_pair – Maximum clique, Maximum independent set.
Return type (set, set) tuple

3.1.9 Steiner Tree

metric_closure(G[, weight]) Return the metric closure of a graph.


steiner_tree(G, terminal_nodes[, weight]) Return an approximation to the minimum Steiner tree of
a graph.

networkx.algorithms.approximation.steinertree.metric_closure

metric_closure(G, weight=’weight’)
Return the metric closure of a graph.
The metric closure of a graph G is the complete graph in which each edge is weighted by the shortest path
distance between the nodes in G .
Parameters G (NetworkX graph)
Returns Metric closure of the graph G.
Return type NetworkX graph

networkx.algorithms.approximation.steinertree.steiner_tree

steiner_tree(G, terminal_nodes, weight=’weight’)


Return an approximation to the minimum Steiner tree of a graph.
Parameters
• G (NetworkX graph)
• terminal_nodes (list) – A list of terminal nodes for which minimum steiner tree is to be
found.
Returns Approximation to the minimum steiner tree of G induced by terminal_nodes .
Return type NetworkX graph

138 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Steiner tree can be approximated by computing the minimum spanning tree of the subgraph of the metric closure
of the graph induced by the terminal nodes, where the metric closure of G is the complete graph in which each
edge is weighted by the shortest path distance between the nodes in G . This algorithm produces a tree whose
weight is within a (2 - (2 / t)) factor of the weight of the optimal Steiner tree where t is number of terminal
nodes.

3.1.10 Treewidth

Functions for computing treewidth decomposition.


Treewidth of an undirected graph is a number associated with the graph. It can be defined as the size of the largest
vertex set (bag) in a tree decomposition of the graph minus one.
Wikipedia: Treewidth
The notions of treewidth and tree decomposition have gained their attractiveness partly because many graph and
network problems that are intractable (e.g., NP-hard) on arbitrary graphs become efficiently solvable (e.g., with a
linear time algorithm) when the treewidth of the input graphs is bounded by a constant12 .
There are two different functions for computing a tree decomposition: treewidth_min_degree() and
treewidth_min_fill_in().

treewidth_min_degree(G) Returns a treewidth decomposition using the Minimum


Degree heuristic.
treewidth_min_fill_in(G) Returns a treewidth decomposition using the Minimum
Fill-in heuristic.

networkx.algorithms.approximation.treewidth.treewidth_min_degree

treewidth_min_degree(G)
Returns a treewidth decomposition using the Minimum Degree heuristic.
The heuristic chooses the nodes according to their degree, i.e., first the node with the lowest degree is chosen,
then the graph is updated and the corresponding node is removed. Next, a new node with the lowest degree is
chosen, and so on.
Parameters G (NetworkX graph)
Returns Treewidth decomposition – 2-tuple with treewidth and the corresponding decomposed
tree.
Return type (int, Graph) tuple

networkx.algorithms.approximation.treewidth.treewidth_min_fill_in

treewidth_min_fill_in(G)
Returns a treewidth decomposition using the Minimum Fill-in heuristic.
The heuristic chooses a node from the graph, where the number of edges added turning the neighbourhood of
the chosen node into clique is as small as possible.
1 Hans L. Bodlaender and Arie M. C. A. Koster. 2010. “Treewidth computations I.Upper bounds”. Inf. Comput. 208, 3 (March 2010),259-275.
http://dx.doi.org/10.1016/j.ic.2009.03.008
2 Hand L. Bodlaender. “Discovering Treewidth”. Institute of Information and Computing Sciences, Utrecht University. Technical Report

UU-CS-2005-018. http://www.cs.uu.nl

3.1. Approximations and Heuristics 139


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters G (NetworkX graph)


Returns Treewidth decomposition – 2-tuple with treewidth and the corresponding decomposed
tree.
Return type (int, Graph) tuple

3.1.11 Vertex Cover

Functions for computing an approximate minimum weight vertex cover.


A vertex cover is a subset of nodes such that each edge in the graph is incident to at least one node in the subset.

min_weighted_vertex_cover(G[, weight]) Returns an approximate minimum weighted vertex


cover.

networkx.algorithms.approximation.vertex_cover.min_weighted_vertex_cover

min_weighted_vertex_cover(G, weight=None)
Returns an approximate minimum weighted vertex cover.
The set of nodes returned by this function is guaranteed to be a vertex cover, and the total weight of the set is
guaranteed to be at most twice the total weight of the minimum weight vertex cover. In other words,
𝑤(𝑆) ≤ 2 * 𝑤(𝑆 * ),
where 𝑆 is the vertex cover returned by this function, 𝑆 * is the vertex cover of minimum weight out of all vertex
covers of the graph, and 𝑤 is the function that computes the sum of the weights of each node in that given set.
Parameters
• G (NetworkX graph)
• weight (string, optional (default = None)) – If None, every node has weight 1. If a string,
use this node attribute as the node weight. A node without this attribute is assumed to have
weight 1.
Returns min_weighted_cover – Returns a set of nodes whose weight sum is no more than twice
the weight sum of the minimum weight vertex cover.
Return type set

Notes

For a directed graph, a vertex cover has the same definition: a set of nodes such that each edge in the graph is
incident to at least one node in the set. Whether the node is the head or tail of the directed edge is ignored.
This is the local-ratio algorithm for computing an approximate vertex cover. The algorithm greedily reduces
the costs over edges, iteratively building a cover. The worst-case runtime of this implementation is 𝑂(𝑚 log 𝑛),
where 𝑛 is the number of nodes and 𝑚 the number of edges in the graph.

References

3.2 Assortativity

140 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.2.1 Assortativity

degree_assortativity_coefficient(G[, x, Compute degree assortativity of graph.


y, . . . ])
attribute_assortativity_coefficient(G, Compute assortativity for node attributes.
attribute)
numeric_assortativity_coefficient(G, Compute assortativity for numerical node attributes.
attribute)
Compute degree assortativity of graph.
degree_pearson_correlation_coefficient(G[,
. . . ])

networkx.algorithms.assortativity.degree_assortativity_coefficient

degree_assortativity_coefficient(G, x=’out’, y=’in’, weight=None, nodes=None)


Compute degree assortativity of graph.
Assortativity measures the similarity of connections in the graph with respect to the node degree.
Parameters
• G (NetworkX graph)
• x (string (‘in’,’out’)) – The degree type for source node (directed graphs only).
• y (string (‘in’,’out’)) – The degree type for target node (directed graphs only).
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used as a weight. If None, then each edge has weight 1. The degree is the sum of
the edge weights adjacent to the node.
• nodes (list or iterable (optional)) – Compute degree assortativity only for nodes in container.
The default is all nodes.
Returns r – Assortativity of graph by degree.
Return type float

Examples

>>> G=nx.path_graph(4)
>>> r=nx.degree_assortativity_coefficient(G)
>>> print("%3.1f"%r)
-0.5

See also:
attribute_assortativity_coefficient(), numeric_assortativity_coefficient(),
neighbor_connectivity(), degree_mixing_dict(), degree_mixing_matrix()

Notes

This computes Eq. (21) in Ref.1 , where e is the joint probability distribution (mixing matrix) of the degrees. If
G is directed than the matrix e is the joint probability of the user-specified degree type for the source and target.
1 M. E. J. Newman, Mixing patterns in networks, Physical Review E, 67 026126, 2003

3.2. Assortativity 141


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

networkx.algorithms.assortativity.attribute_assortativity_coefficient

attribute_assortativity_coefficient(G, attribute, nodes=None)


Compute assortativity for node attributes.
Assortativity measures the similarity of connections in the graph with respect to the given attribute.
Parameters
• G (NetworkX graph)
• attribute (string) – Node attribute key
• nodes (list or iterable (optional)) – Compute attribute assortativity for nodes in container.
The default is all nodes.
Returns r – Assortativity of graph for given attribute
Return type float

Examples

>>> G=nx.Graph()
>>> G.add_nodes_from([0,1],color='red')
>>> G.add_nodes_from([2,3],color='blue')
>>> G.add_edges_from([(0,1),(2,3)])
>>> print(nx.attribute_assortativity_coefficient(G,'color'))
1.0

Notes

This computes Eq. (2) in Ref.1 , trace(M)-sum(M))/(1-sum(M), where M is the joint probability distribution
(mixing matrix) of the specified attribute.

References

networkx.algorithms.assortativity.numeric_assortativity_coefficient

numeric_assortativity_coefficient(G, attribute, nodes=None)


Compute assortativity for numerical node attributes.
Assortativity measures the similarity of connections in the graph with respect to the given numeric attribute.
The numeric attribute must be an integer.
Parameters
• G (NetworkX graph)
• attribute (string) – Node attribute key. The corresponding attribute value must be an integer.
• nodes (list or iterable (optional)) – Compute numeric assortativity only for attributes of
nodes in container. The default is all nodes.
Returns r – Assortativity of graph for given attribute
1 M. E. J. Newman, Mixing patterns in networks, Physical Review E, 67 026126, 2003

142 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type float

Examples

>>> G=nx.Graph()
>>> G.add_nodes_from([0,1],size=2)
>>> G.add_nodes_from([2,3],size=3)
>>> G.add_edges_from([(0,1),(2,3)])
>>> print(nx.numeric_assortativity_coefficient(G,'size'))
1.0

Notes

This computes Eq. (21) in Ref.1 , for the mixing matrix of of the specified attribute.

References

networkx.algorithms.assortativity.degree_pearson_correlation_coefficient

degree_pearson_correlation_coefficient(G, x=’out’, y=’in’, weight=None, nodes=None)


Compute degree assortativity of graph.
Assortativity measures the similarity of connections in the graph with respect to the node degree.
This is the same as degree_assortativity_coefficient but uses the potentially faster scipy.stats.pearsonr function.
Parameters
• G (NetworkX graph)
• x (string (‘in’,’out’)) – The degree type for source node (directed graphs only).
• y (string (‘in’,’out’)) – The degree type for target node (directed graphs only).
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used as a weight. If None, then each edge has weight 1. The degree is the sum of
the edge weights adjacent to the node.
• nodes (list or iterable (optional)) – Compute pearson correlation of degrees only for speci-
fied nodes. The default is all nodes.
Returns r – Assortativity of graph by degree.
Return type float

Examples

>>> G=nx.path_graph(4)
>>> r=nx.degree_pearson_correlation_coefficient(G)
>>> print("%3.1f"%r)
-0.5

1 M. E. J. Newman, Mixing patterns in networks Physical Review E, 67 026126, 2003

3.2. Assortativity 143


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This calls scipy.stats.pearsonr.

References

3.2.2 Average neighbor degree

average_neighbor_degree(G[, source, target, Returns the average degree of the neighborhood of each
. . . ]) node.

networkx.algorithms.assortativity.average_neighbor_degree

average_neighbor_degree(G, source=’out’, target=’out’, nodes=None, weight=None)


Returns the average degree of the neighborhood of each node.
The average neighborhood degree of a node i is
1 ∑︁
𝑘𝑛𝑛,𝑖 = 𝑘𝑗
|𝑁 (𝑖)|
𝑗∈𝑁 (𝑖)

where N(i) are the neighbors of node i and k_j is the degree of node j which belongs to N(i). For weighted
graphs, an analogous measure can be defined1 ,

𝑤 1 ∑︁
𝑘𝑛𝑛,𝑖 = 𝑤𝑖𝑗 𝑘𝑗
𝑠𝑖
𝑗∈𝑁 (𝑖)

where s_i is the weighted degree of node i, w_{ij} is the weight of the edge that links i and j and N(i)
are the neighbors of node i.
Parameters
• G (NetworkX graph)
• source (string (“in”|”out”)) – Directed graphs only. Use “in”- or “out”-degree for source
node.
• target (string (“in”|”out”)) – Directed graphs only. Use “in”- or “out”-degree for target
node.
• nodes (list or iterable, optional) – Compute neighbor degree for specified nodes. The default
is all nodes in the graph.
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used as a weight. If None, then each edge has weight 1.
Returns d – A dictionary keyed by node with average neighbors degree value.
Return type dict

Examples

1 A. Barrat, M. Barthélemy, R. Pastor-Satorras, and A. Vespignani, “The architecture of complex weighted networks”. PNAS 101 (11):

3747–3752 (2004).

144 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G=nx.path_graph(4)
>>> G.edges[0, 1]['weight'] = 5
>>> G.edges[2, 3]['weight'] = 3

>>> nx.average_neighbor_degree(G)
{0: 2.0, 1: 1.5, 2: 1.5, 3: 2.0}
>>> nx.average_neighbor_degree(G, weight='weight')
{0: 2.0, 1: 1.1666666666666667, 2: 1.25, 3: 2.0}

>>> G=nx.DiGraph()
>>> nx.add_path(G, [0, 1, 2, 3])
>>> nx.average_neighbor_degree(G, source='in', target='in')
{0: 1.0, 1: 1.0, 2: 1.0, 3: 0.0}

>>> nx.average_neighbor_degree(G, source='out', target='out')


{0: 1.0, 1: 1.0, 2: 0.0, 3: 0.0}

Notes

For directed graphs you can also specify in-degree or out-degree by passing keyword arguments.
See also:
average_degree_connectivity()

References

3.2.3 Average degree connectivity

average_degree_connectivity(G[, source, Compute the average degree connectivity of graph.


. . . ])
k_nearest_neighbors(G[, source, target, . . . ]) Compute the average degree connectivity of graph.

networkx.algorithms.assortativity.average_degree_connectivity

average_degree_connectivity(G, source=’in+out’, target=’in+out’, nodes=None, weight=None)


Compute the average degree connectivity of graph.
The average degree connectivity is the average nearest neighbor degree of nodes with degree k. For weighted
graphs, an analogous measure can be computed using the weighted average neighbors degree defined in1 , for a
node i, as

𝑤 1 ∑︁
𝑘𝑛𝑛,𝑖 = 𝑤𝑖𝑗 𝑘𝑗
𝑠𝑖
𝑗∈𝑁 (𝑖)

where s_i is the weighted degree of node i, w_{ij} is the weight of the edge that links i and j, and N(i)
are the neighbors of node i.
Parameters
• G (NetworkX graph)
1A. Barrat, M. Barthélemy, R. Pastor-Satorras, and A. Vespignani, “The architecture of complex weighted networks”. PNAS 101 (11):
3747–3752 (2004).

3.2. Assortativity 145


NetworkX Reference, Release 2.3rc1.dev20190222214247

• source (“in”|”out”|”in+out” (default:”in+out”)) – Directed graphs only. Use “in”- or


“out”-degree for source node.
• target (“in”|”out”|”in+out” (default:”in+out”) – Directed graphs only. Use “in”- or
“out”-degree for target node.
• nodes (list or iterable (optional)) – Compute neighbor connectivity for these nodes. The
default is all nodes.
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used as a weight. If None, then each edge has weight 1.
Returns d – A dictionary keyed by degree k with the value of average connectivity.
Return type dict
Raises ValueError – If either source or target are not one of ‘in’, ‘out’, or ‘in+out’.

Examples

>>> G=nx.path_graph(4)
>>> G.edges[1, 2]['weight'] = 3
>>> nx.k_nearest_neighbors(G)
{1: 2.0, 2: 1.5}
>>> nx.k_nearest_neighbors(G, weight='weight')
{1: 2.0, 2: 1.75}

See also:
neighbors_average_degree()

Notes

This algorithm is sometimes called “k nearest neighbors” and is also available as k_nearest_neighbors.

References

networkx.algorithms.assortativity.k_nearest_neighbors

k_nearest_neighbors(G, source=’in+out’, target=’in+out’, nodes=None, weight=None)


Compute the average degree connectivity of graph.
The average degree connectivity is the average nearest neighbor degree of nodes with degree k. For weighted
graphs, an analogous measure can be computed using the weighted average neighbors degree defined in1 , for a
node i, as

𝑤 1 ∑︁
𝑘𝑛𝑛,𝑖 = 𝑤𝑖𝑗 𝑘𝑗
𝑠𝑖
𝑗∈𝑁 (𝑖)

where s_i is the weighted degree of node i, w_{ij} is the weight of the edge that links i and j, and N(i)
are the neighbors of node i.
Parameters
• G (NetworkX graph)
1 A. Barrat, M. Barthélemy, R. Pastor-Satorras, and A. Vespignani, “The architecture of complex weighted networks”. PNAS 101 (11):

3747–3752 (2004).

146 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• source (“in”|”out”|”in+out” (default:”in+out”)) – Directed graphs only. Use “in”- or


“out”-degree for source node.
• target (“in”|”out”|”in+out” (default:”in+out”) – Directed graphs only. Use “in”- or
“out”-degree for target node.
• nodes (list or iterable (optional)) – Compute neighbor connectivity for these nodes. The
default is all nodes.
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used as a weight. If None, then each edge has weight 1.
Returns d – A dictionary keyed by degree k with the value of average connectivity.
Return type dict
Raises ValueError – If either source or target are not one of ‘in’, ‘out’, or ‘in+out’.

Examples

>>> G=nx.path_graph(4)
>>> G.edges[1, 2]['weight'] = 3
>>> nx.k_nearest_neighbors(G)
{1: 2.0, 2: 1.5}
>>> nx.k_nearest_neighbors(G, weight='weight')
{1: 2.0, 2: 1.75}

See also:
neighbors_average_degree()

Notes

This algorithm is sometimes called “k nearest neighbors” and is also available as k_nearest_neighbors.

References

3.2.4 Mixing

attribute_mixing_matrix(G, attribute[, . . . ]) Returns mixing matrix for attribute.


degree_mixing_matrix(G[, x, y, weight, . . . ]) Returns mixing matrix for attribute.
numeric_mixing_matrix(G, attribute[, nodes, Returns numeric mixing matrix for attribute.
. . . ])
attribute_mixing_dict(G, attribute[, nodes, Returns dictionary representation of mixing matrix for
. . . ]) attribute.
degree_mixing_dict(G[, x, y, weight, nodes, Returns dictionary representation of mixing matrix for
. . . ]) degree.
mixing_dict(xy[, normalized]) Returns a dictionary representation of mixing matrix.

networkx.algorithms.assortativity.attribute_mixing_matrix

attribute_mixing_matrix(G, attribute, nodes=None, mapping=None, normalized=True)


Returns mixing matrix for attribute.
Parameters

3.2. Assortativity 147


NetworkX Reference, Release 2.3rc1.dev20190222214247

• G (graph) – NetworkX graph object.


• attribute (string) – Node attribute key.
• nodes (list or iterable (optional)) – Use only nodes in container to build the matrix. The
default is all nodes.
• mapping (dictionary, optional) – Mapping from node attribute to integer index in matrix. If
not specified, an arbitrary ordering will be used.
• normalized (bool (default=True)) – Return counts if False or probabilities if True.
Returns m – Counts or joint probability of occurrence of attribute pairs.
Return type numpy array

networkx.algorithms.assortativity.degree_mixing_matrix

degree_mixing_matrix(G, x=’out’, y=’in’, weight=None, nodes=None, normalized=True)


Returns mixing matrix for attribute.
Parameters
• G (graph) – NetworkX graph object.
• x (string (‘in’,’out’)) – The degree type for source node (directed graphs only).
• y (string (‘in’,’out’)) – The degree type for target node (directed graphs only).
• nodes (list or iterable (optional)) – Build the matrix using only nodes in container. The
default is all nodes.
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used as a weight. If None, then each edge has weight 1. The degree is the sum of
the edge weights adjacent to the node.
• normalized (bool (default=True)) – Return counts if False or probabilities if True.
Returns m – Counts, or joint probability, of occurrence of node degree.
Return type numpy array

networkx.algorithms.assortativity.numeric_mixing_matrix

numeric_mixing_matrix(G, attribute, nodes=None, normalized=True)


Returns numeric mixing matrix for attribute.
The attribute must be an integer.
Parameters
• G (graph) – NetworkX graph object.
• attribute (string) – Node attribute key. The corresponding attribute must be an integer.
• nodes (list or iterable (optional)) – Build the matrix only with nodes in container. The
default is all nodes.
• normalized (bool (default=True)) – Return counts if False or probabilities if True.
Returns m – Counts, or joint, probability of occurrence of node attribute pairs.
Return type numpy array

148 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.assortativity.attribute_mixing_dict

attribute_mixing_dict(G, attribute, nodes=None, normalized=False)


Returns dictionary representation of mixing matrix for attribute.
Parameters
• G (graph) – NetworkX graph object.
• attribute (string) – Node attribute key.
• nodes (list or iterable (optional)) – Unse nodes in container to build the dict. The default is
all nodes.
• normalized (bool (default=False)) – Return counts if False or probabilities if True.

Examples

>>> G=nx.Graph()
>>> G.add_nodes_from([0,1],color='red')
>>> G.add_nodes_from([2,3],color='blue')
>>> G.add_edge(1,3)
>>> d=nx.attribute_mixing_dict(G,'color')
>>> print(d['red']['blue'])
1
>>> print(d['blue']['red']) # d symmetric for undirected graphs
1

Returns d – Counts or joint probability of occurrence of attribute pairs.


Return type dictionary

networkx.algorithms.assortativity.degree_mixing_dict

degree_mixing_dict(G, x=’out’, y=’in’, weight=None, nodes=None, normalized=False)


Returns dictionary representation of mixing matrix for degree.
Parameters
• G (graph) – NetworkX graph object.
• x (string (‘in’,’out’)) – The degree type for source node (directed graphs only).
• y (string (‘in’,’out’)) – The degree type for target node (directed graphs only).
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used as a weight. If None, then each edge has weight 1. The degree is the sum of
the edge weights adjacent to the node.
• normalized (bool (default=False)) – Return counts if False or probabilities if True.
Returns d – Counts or joint probability of occurrence of degree pairs.
Return type dictionary

3.2. Assortativity 149


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.assortativity.mixing_dict

mixing_dict(xy, normalized=False)
Returns a dictionary representation of mixing matrix.
Parameters
• xy (list or container of two-tuples) – Pairs of (x,y) items.
• attribute (string) – Node attribute key
• normalized (bool (default=False)) – Return counts if False or probabilities if True.
Returns d – Counts or Joint probability of occurrence of values in xy.
Return type dictionary

3.3 Bipartite

This module provides functions and operations for bipartite graphs. Bipartite graphs B = (U, V, E) have two
node sets U,V and edges in E that only connect nodes from opposite sets. It is common in the literature to use an
spatial analogy referring to the two node sets as top and bottom nodes.
The bipartite algorithms are not imported into the networkx namespace at the top level so the easiest way to use them
is with:

>>> import networkx as nx


>>> from networkx.algorithms import bipartite

NetworkX does not have a custom bipartite graph class but the Graph() or DiGraph() classes can be used to represent
bipartite graphs. However, you have to keep track of which set each node belongs to, and make sure that there is no
edge between nodes of the same set. The convention used in NetworkX is to use a node attribute named bipartite
with values 0 or 1 to identify the sets each node belongs to. This convention is not enforced in the source code of
bipartite functions, it’s only a recommendation.
For example:

>>> B = nx.Graph()
>>> # Add nodes with the node attribute "bipartite"
>>> B.add_nodes_from([1, 2, 3, 4], bipartite=0)
>>> B.add_nodes_from(['a', 'b', 'c'], bipartite=1)
>>> # Add edges only between nodes of opposite node sets
>>> B.add_edges_from([(1, 'a'), (1, 'b'), (2, 'b'), (2, 'c'), (3, 'c'), (4, 'a')])

Many algorithms of the bipartite module of NetworkX require, as an argument, a container with all the nodes that
belong to one set, in addition to the bipartite graph B. The functions in the bipartite package do not check that the node
set is actually correct nor that the input graph is actually bipartite. If B is connected, you can find the two node sets
using a two-coloring algorithm:

>>> nx.is_connected(B)
True
>>> bottom_nodes, top_nodes = bipartite.sets(B)

However, if the input graph is not connected, there are more than one possible colorations. This is the reason why we
require the user to pass a container with all nodes of one bipartite node set as an argument to most bipartite functions.
In the face of ambiguity, we refuse the temptation to guess and raise an AmbiguousSolution Exception if the
input graph for bipartite.sets is disconnected.

150 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Using the bipartite node attribute, you can easily get the two node sets:

>>> top_nodes = {n for n, d in B.nodes(data=True) if d['bipartite']==0}


>>> bottom_nodes = set(B) - top_nodes

So you can easily use the bipartite algorithms that require, as an argument, a container with all nodes that belong to
one node set:

>>> print(round(bipartite.density(B, bottom_nodes), 2))


0.5
>>> G = bipartite.projected_graph(B, top_nodes)

All bipartite graph generators in NetworkX build bipartite graphs with the bipartite node attribute. Thus, you can
use the same approach:

>>> RB = bipartite.random_graph(5, 7, 0.2)


>>> RB_top = {n for n, d in RB.nodes(data=True) if d['bipartite']==0}
>>> RB_bottom = set(RB) - RB_top
>>> list(RB_top)
[0, 1, 2, 3, 4]
>>> list(RB_bottom)
[5, 6, 7, 8, 9, 10, 11]

For other bipartite graph generators see Generators.

3.3.1 Basic functions

Bipartite Graph Algorithms

is_bipartite(G) Returns True if graph G is bipartite, False if not.


is_bipartite_node_set(G, nodes) Returns True if nodes and G/nodes are a bipartition of
G.
sets(G[, top_nodes]) Returns bipartite node sets of graph G.
color(G) Returns a two-coloring of the graph.
density(B, nodes) Returns density of bipartite graph B.
degrees(B, nodes[, weight]) Returns the degrees of the two node sets in the bipartite
graph B.

networkx.algorithms.bipartite.basic.is_bipartite

is_bipartite(G)
Returns True if graph G is bipartite, False if not.
Parameters G (NetworkX graph)

Examples

>>> from networkx.algorithms import bipartite


>>> G = nx.path_graph(4)
>>> print(bipartite.is_bipartite(G))
True

See also:

3.3. Bipartite 151


NetworkX Reference, Release 2.3rc1.dev20190222214247

color(), is_bipartite_node_set()

networkx.algorithms.bipartite.basic.is_bipartite_node_set

is_bipartite_node_set(G, nodes)
Returns True if nodes and G/nodes are a bipartition of G.
Parameters
• G (NetworkX graph)
• nodes (list or container) – Check if nodes are a one of a bipartite set.

Examples

>>> from networkx.algorithms import bipartite


>>> G = nx.path_graph(4)
>>> X = set([1,3])
>>> bipartite.is_bipartite_node_set(G,X)
True

Notes

For connected graphs the bipartite sets are unique. This function handles disconnected graphs.

networkx.algorithms.bipartite.basic.sets

sets(G, top_nodes=None)
Returns bipartite node sets of graph G.
Raises an exception if the graph is not bipartite or if the input graph is disconnected and thus more than one valid
solution exists. See bipartite documentation for further details on how bipartite graphs are handled in
NetworkX.
Parameters
• G (NetworkX graph)
• top_nodes (container) – Container with all nodes in one bipartite node set. If not supplied
it will be computed. But if more than one solution exists an exception will be raised.
Returns (X,Y) – One set of nodes for each part of the bipartite graph.
Return type two-tuple of sets
Raises
• AmbiguousSolution : Exception – Raised if the input bipartite graph is disconnected and no
container with all nodes in one bipartite set is provided. When determining the nodes in
each bipartite set more than one valid solution is possible if the input graph is disconnected.
• NetworkXError: Exception – Raised if the input graph is not bipartite.

152 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> from networkx.algorithms import bipartite


>>> G = nx.path_graph(4)
>>> X, Y = bipartite.sets(G)
>>> list(X)
[0, 2]
>>> list(Y)
[1, 3]

See also:
color()

networkx.algorithms.bipartite.basic.color

color(G)
Returns a two-coloring of the graph.
Raises an exception if the graph is not bipartite.
Parameters G (NetworkX graph)
Returns color – A dictionary keyed by node with a 1 or 0 as data for each node color.
Return type dictionary
Raises exc:NetworkXError if the graph is not two-colorable.

Examples

>>> from networkx.algorithms import bipartite


>>> G = nx.path_graph(4)
>>> c = bipartite.color(G)
>>> print(c)
{0: 1, 1: 0, 2: 1, 3: 0}

You can use this to set a node attribute indicating the biparite set:

>>> nx.set_node_attributes(G, c, 'bipartite')


>>> print(G.nodes[0]['bipartite'])
1
>>> print(G.nodes[1]['bipartite'])
0

networkx.algorithms.bipartite.basic.density

density(B, nodes)
Returns density of bipartite graph B.
Parameters
• G (NetworkX graph)
• nodes (list or container) – Nodes in one node set of the bipartite graph.
Returns d – The bipartite density

3.3. Bipartite 153


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type float

Examples

>>> from networkx.algorithms import bipartite


>>> G = nx.complete_bipartite_graph(3,2)
>>> X=set([0,1,2])
>>> bipartite.density(G,X)
1.0
>>> Y=set([3,4])
>>> bipartite.density(G,Y)
1.0

Notes

The container of nodes passed as argument must contain all nodes in one of the two bipartite node sets to avoid
ambiguity in the case of disconnected graphs. See bipartite documentation for further details on how
bipartite graphs are handled in NetworkX.
See also:
color()

networkx.algorithms.bipartite.basic.degrees

degrees(B, nodes, weight=None)


Returns the degrees of the two node sets in the bipartite graph B.
Parameters
• G (NetworkX graph)
• nodes (list or container) – Nodes in one node set of the bipartite graph.
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used as a weight. If None, then each edge has weight 1. The degree is the sum of
the edge weights adjacent to the node.
Returns (degX,degY) – The degrees of the two bipartite sets as dictionaries keyed by node.
Return type tuple of dictionaries

Examples

>>> from networkx.algorithms import bipartite


>>> G = nx.complete_bipartite_graph(3,2)
>>> Y=set([3,4])
>>> degX,degY=bipartite.degrees(G,Y)
>>> dict(degX)
{0: 2, 1: 2, 2: 2}

154 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The container of nodes passed as argument must contain all nodes in one of the two bipartite node sets to avoid
ambiguity in the case of disconnected graphs. See bipartite documentation for further details on how
bipartite graphs are handled in NetworkX.
See also:
color(), density()

3.3.2 Matching

Provides functions for computing a maximum cardinality matching in a bipartite graph.


If you don’t care about the particular implementation of the maximum matching algorithm, simply use the
maximum_matching(). If you do care, you can import one of the named maximum matching algorithms directly.
For example, to find a maximum matching in the complete bipartite graph with two vertices on the left and three
vertices on the right:

>>> import networkx as nx


>>> G = nx.complete_bipartite_graph(2, 3)
>>> left, right = nx.bipartite.sets(G)
>>> list(left)
[0, 1]
>>> list(right)
[2, 3, 4]
>>> nx.bipartite.maximum_matching(G)
{0: 2, 1: 3, 2: 0, 3: 1}

The dictionary returned by maximum_matching() includes a mapping for vertices in both the left and right vertex
sets.

eppstein_matching(G[, top_nodes]) Returns the maximum cardinality matching of the bipar-


tite graph G.
hopcroft_karp_matching(G[, top_nodes]) Returns the maximum cardinality matching of the bipar-
tite graph G.
to_vertex_cover(G, matching[, top_nodes]) Returns the minimum vertex cover corresponding to the
given maximum matching of the bipartite graph G.

networkx.algorithms.bipartite.matching.eppstein_matching

eppstein_matching(G, top_nodes=None)
Returns the maximum cardinality matching of the bipartite graph G.
Parameters
• G (NetworkX graph) – Undirected bipartite graph
• top_nodes (container) – Container with all nodes in one bipartite node set. If not supplied
it will be computed. But if more than one solution exists an exception will be raised.
Returns matches – The matching is returned as a dictionary, matching, such that matching[v]
== w if node v is matched to node w. Unmatched nodes do not occur as a key in mate.
Return type dictionary

3.3. Bipartite 155


NetworkX Reference, Release 2.3rc1.dev20190222214247

Raises AmbiguousSolution : Exception – Raised if the input bipartite graph is disconnected and no
container with all nodes in one bipartite set is provided. When determining the nodes in each
bipartite set more than one valid solution is possible if the input graph is disconnected.

Notes

This function is implemented with David Eppstein’s version of the algorithm Hopcroft–Karp algorithm (see
hopcroft_karp_matching()), which originally appeared in the Python Algorithms and Data Structures
library (PADS).
See bipartite documentation for further details on how bipartite graphs are handled in NetworkX.
See also:
hopcroft_karp_matching()

networkx.algorithms.bipartite.matching.hopcroft_karp_matching

hopcroft_karp_matching(G, top_nodes=None)
Returns the maximum cardinality matching of the bipartite graph G.
Parameters
• G (NetworkX graph) – Undirected bipartite graph
• top_nodes (container) – Container with all nodes in one bipartite node set. If not supplied
it will be computed. But if more than one solution exists an exception will be raised.
Returns matches – The matching is returned as a dictionary, matches, such that matches[v]
== w if node v is matched to node w. Unmatched nodes do not occur as a key in mate.
Return type dictionary
Raises AmbiguousSolution : Exception – Raised if the input bipartite graph is disconnected and no
container with all nodes in one bipartite set is provided. When determining the nodes in each
bipartite set more than one valid solution is possible if the input graph is disconnected.

Notes

This function is implemented with the Hopcroft–Karp matching algorithm for bipartite graphs.
See bipartite documentation for further details on how bipartite graphs are handled in NetworkX.
See also:
eppstein_matching()

References

networkx.algorithms.bipartite.matching.to_vertex_cover

to_vertex_cover(G, matching, top_nodes=None)


Returns the minimum vertex cover corresponding to the given maximum matching of the bipartite graph G.
Parameters
• G (NetworkX graph) – Undirected bipartite graph

156 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• matching (dictionary) – A dictionary whose keys are vertices in G and whose values are the
distinct neighbors comprising the maximum matching for G, as returned by, for example,
maximum_matching(). The dictionary must represent the maximum matching.
• top_nodes (container) – Container with all nodes in one bipartite node set. If not supplied
it will be computed. But if more than one solution exists an exception will be raised.
Returns vertex_cover – The minimum vertex cover in G.
Return type set
Raises AmbiguousSolution : Exception – Raised if the input bipartite graph is disconnected and no
container with all nodes in one bipartite set is provided. When determining the nodes in each
bipartite set more than one valid solution is possible if the input graph is disconnected.

Notes

This function is implemented using the procedure guaranteed by Konig’s theorem, which proves an equivalence
between a maximum matching and a minimum vertex cover in bipartite graphs.
Since a minimum vertex cover is the complement of a maximum independent set for any graph, one can compute
the maximum independent set of a bipartite graph this way:

>>> import networkx as nx


>>> G = nx.complete_bipartite_graph(2, 3)
>>> matching = nx.bipartite.maximum_matching(G)
>>> vertex_cover = nx.bipartite.to_vertex_cover(G, matching)
>>> independent_set = set(G) - vertex_cover
>>> print(list(independent_set))
[2, 3, 4]

See bipartite documentation for further details on how bipartite graphs are handled in NetworkX.

3.3.3 Matrix

Biadjacency matrices

biadjacency_matrix(G, row_order[, . . . ]) Returns the biadjacency matrix of the bipartite graph G.


from_biadjacency_matrix(A[, create_using, Creates a new bipartite graph from a biadjacency matrix
. . . ]) given as a SciPy sparse matrix.

networkx.algorithms.bipartite.matrix.biadjacency_matrix

biadjacency_matrix(G, row_order, column_order=None, dtype=None, weight=’weight’, format=’csr’)


Returns the biadjacency matrix of the bipartite graph G.
Let G = (U, V, E) be a bipartite graph with node sets U = u_{1},...,u_{r} and V = v_{1},..
.,v_{s}. The biadjacency matrix1 is the r x s matrix B in which b_{i,j} = 1 if, and only if, (u_i,
v_j) in E. If the parameter weight is not None and matches the name of an edge attribute, its value is
used instead of 1.
Parameters
• G (graph) – A NetworkX graph
1 https://en.wikipedia.org/wiki/Adjacency_matrix#Adjacency_matrix_of_a_bipartite_graph

3.3. Bipartite 157


NetworkX Reference, Release 2.3rc1.dev20190222214247

• row_order (list of nodes) – The rows of the matrix are ordered according to the list of nodes.
• column_order (list, optional) – The columns of the matrix are ordered according to the list
of nodes. If column_order is None, then the ordering of columns is arbitrary.
• dtype (NumPy data-type, optional) – A valid NumPy dtype used to initialize the array. If
None, then the NumPy default is used.
• weight (string or None, optional (default=’weight’)) – The edge data key used to provide
each value in the matrix. If None, then each edge has weight 1.
• format (str in {‘bsr’, ‘csr’, ‘csc’, ‘coo’, ‘lil’, ‘dia’, ‘dok’}) – The type of the matrix to be
returned (default ‘csr’). For some algorithms different implementations of sparse matrices
can perform better. See2 for details.
Returns M – Biadjacency matrix representation of the bipartite graph G.
Return type SciPy sparse matrix

Notes

No attempt is made to check that the input graph is bipartite.


For directed bipartite graphs only successors are considered as neighbors. To obtain an adjacency matrix with
ones (or weight values) for both predecessors and successors you have to generate two biadjacency matrices
where the rows of one of them are the columns of the other, and then add one to the transpose of the other.
See also:
adjacency_matrix(), from_biadjacency_matrix()

References

networkx.algorithms.bipartite.matrix.from_biadjacency_matrix

from_biadjacency_matrix(A, create_using=None, edge_attribute=’weight’)


Creates a new bipartite graph from a biadjacency matrix given as a SciPy sparse matrix.
Parameters
• A (scipy sparse matrix) – A biadjacency matrix representation of a graph
• create_using (NetworkX graph) – Use specified graph for result. The default is Graph()
• edge_attribute (string) – Name of edge attribute to store matrix numeric value. The data
will have the same type as the matrix entry (int, float, (real,imag)).

Notes

The nodes are labeled with the attribute bipartite set to an integer 0 or 1 representing membership in part 0
or part 1 of the bipartite graph.
If create_using is an instance of networkx.MultiGraph or networkx.MultiDiGraph and the
entries of A are of type int, then this function returns a multigraph (of the same type as create_using)
with parallel edges. In this case, edge_attribute will be ignored.
See also:
biadjacency_matrix(), from_numpy_matrix()
2 Scipy Dev. References, “Sparse Matrices”, https://docs.scipy.org/doc/scipy/reference/sparse.html

158 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

[1] https://en.wikipedia.org/wiki/Adjacency_matrix#Adjacency_matrix_of_a_bipartite_graph

3.3.4 Projections

One-mode (unipartite) projections of bipartite graphs.

projected_graph(B, nodes[, multigraph])Returns the projection of B onto one of its node sets.
weighted_projected_graph(B, nodes[, ratio])
Returns a weighted projection of B onto one of its node
sets.
Newman’s weighted projection of B onto one of its node
collaboration_weighted_projected_graph(B,
nodes) sets.
overlap_weighted_projected_graph(B, Overlap weighted projection of B onto one of its node
nodes[, . . . ]) sets.
generic_weighted_projected_graph(B, Weighted projection of B with a user-specified weight
nodes[, . . . ]) function.

networkx.algorithms.bipartite.projection.projected_graph

projected_graph(B, nodes, multigraph=False)


Returns the projection of B onto one of its node sets.
Returns the graph G that is the projection of the bipartite graph B onto the specified nodes. They retain their
attributes and are connected in G if they have a common neighbor in B.
Parameters
• B (NetworkX graph) – The input graph should be bipartite.
• nodes (list or iterable) – Nodes to project onto (the “bottom” nodes).
• multigraph (bool (default=False)) – If True return a multigraph where the multiple edges
represent multiple shared neighbors. They edge key in the multigraph is assigned to the
label of the neighbor.
Returns Graph – A graph that is the projection onto the given nodes.
Return type NetworkX graph or multigraph

Examples

>>> from networkx.algorithms import bipartite


>>> B = nx.path_graph(4)
>>> G = bipartite.projected_graph(B, [1, 3])
>>> list(G)
[1, 3]
>>> list(G.edges())
[(1, 3)]

If nodes a, and b are connected through both nodes 1 and 2 then building a multigraph results in two edges in
the projection onto [a, b]:

3.3. Bipartite 159


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> B = nx.Graph()
>>> B.add_edges_from([('a', 1), ('b', 1), ('a', 2), ('b', 2)])
>>> G = bipartite.projected_graph(B, ['a', 'b'], multigraph=True)
>>> print([sorted((u, v)) for u, v in G.edges()])
[['a', 'b'], ['a', 'b']]

Notes

No attempt is made to verify that the input graph B is bipartite. Returns a simple graph that is the projection of
the bipartite graph B onto the set of nodes given in list nodes. If multigraph=True then a multigraph is returned
with an edge for every shared neighbor.
Directed graphs are allowed as input. The output will also then be a directed graph with edges if there is a
directed path between the nodes.
The graph and node properties are (shallow) copied to the projected graph.
See bipartite documentation for further details on how bipartite graphs are handled in NetworkX.
See also:
is_bipartite(), is_bipartite_node_set(), sets(), weighted_projected_graph(),
collaboration_weighted_projected_graph(), overlap_weighted_projected_graph(),
generic_weighted_projected_graph()

networkx.algorithms.bipartite.projection.weighted_projected_graph

weighted_projected_graph(B, nodes, ratio=False)


Returns a weighted projection of B onto one of its node sets.
The weighted projected graph is the projection of the bipartite network B onto the specified nodes with weights
representing the number of shared neighbors or the ratio between actual shared neighbors and possible shared
neighbors if ratio is True1 . The nodes retain their attributes and are connected in the resulting graph if
they have an edge to a common node in the original graph.
Parameters
• B (NetworkX graph) – The input graph should be bipartite.
• nodes (list or iterable) – Nodes to project onto (the “bottom” nodes).
• ratio (Bool (default=False)) – If True, edge weight is the ratio between actual shared neigh-
bors and maximum possible shared neighbors (i.e., the size of the other node set). If False,
edges weight is the number of shared neighbors.
Returns Graph – A graph that is the projection onto the given nodes.
Return type NetworkX graph

Examples

1 Borgatti, S.P. and Halgin, D. In press. “Analyzing Affiliation Networks”. In Carrington, P. and Scott, J. (eds) The Sage Handbook of Social

Network Analysis. Sage Publications.

160 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> from networkx.algorithms import bipartite


>>> B = nx.path_graph(4)
>>> G = bipartite.weighted_projected_graph(B, [1, 3])
>>> list(G)
[1, 3]
>>> list(G.edges(data=True))
[(1, 3, {'weight': 1})]
>>> G = bipartite.weighted_projected_graph(B, [1, 3], ratio=True)
>>> list(G.edges(data=True))
[(1, 3, {'weight': 0.5})]

Notes

No attempt is made to verify that the input graph B is bipartite. The graph and node properties are (shallow)
copied to the projected graph.
See bipartite documentation for further details on how bipartite graphs are handled in NetworkX.
See also:
is_bipartite(), is_bipartite_node_set(), sets(), collaboration_weighted_projected_graph(),
overlap_weighted_projected_graph(), generic_weighted_projected_graph(),
projected_graph()

References

networkx.algorithms.bipartite.projection.collaboration_weighted_projected_graph

collaboration_weighted_projected_graph(B, nodes)
Newman’s weighted projection of B onto one of its node sets.
The collaboration weighted projection is the projection of the bipartite network B onto the specified nodes with
weights assigned using Newman’s collaboration model1 :
∑︁ 𝛿 𝑘 𝛿 𝑘
𝑢 𝑣
𝑤𝑢,𝑣 =
𝑑𝑘 − 1
𝑘

where u and v are nodes from the bottom bipartite node set, and k is a node of the top node set. The value d_k
is the degree of node k in the bipartite network and delta_{u}^{k} is 1 if node u is linked to node k in the
original bipartite graph or 0 otherwise.
The nodes retain their attributes and are connected in the resulting graph if have an edge to a common node in
the original bipartite graph.
Parameters
• B (NetworkX graph) – The input graph should be bipartite.
• nodes (list or iterable) – Nodes to project onto (the “bottom” nodes).
Returns Graph – A graph that is the projection onto the given nodes.
Return type NetworkX graph
1 Scientific collaboration networks: II. Shortest paths, weighted networks, and centrality, M. E. J. Newman, Phys. Rev. E 64, 016132 (2001).

3.3. Bipartite 161


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> from networkx.algorithms import bipartite


>>> B = nx.path_graph(5)
>>> B.add_edge(1, 5)
>>> G = bipartite.collaboration_weighted_projected_graph(B, [0, 2, 4, 5])
>>> list(G)
[0, 2, 4, 5]
>>> for edge in G.edges(data=True): print(edge)
...
(0, 2, {'weight': 0.5})
(0, 5, {'weight': 0.5})
(2, 4, {'weight': 1.0})
(2, 5, {'weight': 0.5})

Notes

No attempt is made to verify that the input graph B is bipartite. The graph and node properties are (shallow)
copied to the projected graph.
See bipartite documentation for further details on how bipartite graphs are handled in NetworkX.
See also:
is_bipartite(), is_bipartite_node_set(), sets(), weighted_projected_graph(),
overlap_weighted_projected_graph(), generic_weighted_projected_graph(),
projected_graph()

References

networkx.algorithms.bipartite.projection.overlap_weighted_projected_graph

overlap_weighted_projected_graph(B, nodes, jaccard=True)


Overlap weighted projection of B onto one of its node sets.
The overlap weighted projection is the projection of the bipartite network B onto the specified nodes with
weights representing the Jaccard index between the neighborhoods of the two nodes in the original bipartite
network1 :
|𝑁 (𝑢) ∩ 𝑁 (𝑣)|
𝑤𝑣,𝑢 =
|𝑁 (𝑢) ∪ 𝑁 (𝑣)|

or if the parameter ‘jaccard’ is False, the fraction of common neighbors by minimum of both nodes degree in
the original bipartite graph1 :

|𝑁 (𝑢) ∩ 𝑁 (𝑣)|
𝑤𝑣,𝑢 =
𝑚𝑖𝑛(|𝑁 (𝑢)|, |𝑁 (𝑣)|)

The nodes retain their attributes and are connected in the resulting graph if have an edge to a common node in
the original bipartite graph.
Parameters
• B (NetworkX graph) – The input graph should be bipartite.
1 Borgatti, S.P. and Halgin, D. In press. Analyzing Affiliation Networks. In Carrington, P. and Scott, J. (eds) The Sage Handbook of Social

Network Analysis. Sage Publications.

162 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• nodes (list or iterable) – Nodes to project onto (the “bottom” nodes).


• jaccard (Bool (default=True))
Returns Graph – A graph that is the projection onto the given nodes.
Return type NetworkX graph

Examples

>>> from networkx.algorithms import bipartite


>>> B = nx.path_graph(5)
>>> nodes = [0, 2, 4]
>>> G = bipartite.overlap_weighted_projected_graph(B, nodes)
>>> list(G)
[0, 2, 4]
>>> list(G.edges(data=True))
[(0, 2, {'weight': 0.5}), (2, 4, {'weight': 0.5})]
>>> G = bipartite.overlap_weighted_projected_graph(B, nodes, jaccard=False)
>>> list(G.edges(data=True))
[(0, 2, {'weight': 1.0}), (2, 4, {'weight': 1.0})]

Notes

No attempt is made to verify that the input graph B is bipartite. The graph and node properties are (shallow)
copied to the projected graph.
See bipartite documentation for further details on how bipartite graphs are handled in NetworkX.
See also:
is_bipartite(), is_bipartite_node_set(), sets(), weighted_projected_graph(),
collaboration_weighted_projected_graph(), generic_weighted_projected_graph(),
projected_graph()

References

networkx.algorithms.bipartite.projection.generic_weighted_projected_graph

generic_weighted_projected_graph(B, nodes, weight_function=None)


Weighted projection of B with a user-specified weight function.
The bipartite network B is projected on to the specified nodes with weights computed by a user-specified func-
tion. This function must accept as a parameter the neighborhood sets of two nodes and return an integer or a
float.
The nodes retain their attributes and are connected in the resulting graph if they have an edge to a common node
in the original graph.
Parameters
• B (NetworkX graph) – The input graph should be bipartite.
• nodes (list or iterable) – Nodes to project onto (the “bottom” nodes).
• weight_function (function) – This function must accept as parameters the same input graph
that this function, and two nodes; and return an integer or a float. The default function
computes the number of shared neighbors.

3.3. Bipartite 163


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns Graph – A graph that is the projection onto the given nodes.
Return type NetworkX graph

Examples

>>> from networkx.algorithms import bipartite


>>> # Define some custom weight functions
>>> def jaccard(G, u, v):
... unbrs = set(G[u])
... vnbrs = set(G[v])
... return float(len(unbrs & vnbrs)) / len(unbrs | vnbrs)
...
>>> def my_weight(G, u, v, weight='weight'):
... w = 0
... for nbr in set(G[u]) & set(G[v]):
... w += G[u][nbr].get(weight, 1) + G[v][nbr].get(weight, 1)
... return w
...
>>> # A complete bipartite graph with 4 nodes and 4 edges
>>> B = nx.complete_bipartite_graph(2, 2)
>>> # Add some arbitrary weight to the edges
>>> for i,(u,v) in enumerate(B.edges()):
... B.edges[u, v]['weight'] = i + 1
...
>>> for edge in B.edges(data=True):
... print(edge)
...
(0, 2, {'weight': 1})
(0, 3, {'weight': 2})
(1, 2, {'weight': 3})
(1, 3, {'weight': 4})
>>> # By default, the weight is the number of shared neighbors
>>> G = bipartite.generic_weighted_projected_graph(B, [0, 1])
>>> print(list(G.edges(data=True)))
[(0, 1, {'weight': 2})]
>>> # To specify a custom weight function use the weight_function parameter
>>> G = bipartite.generic_weighted_projected_graph(B, [0, 1], weight_
˓→function=jaccard)

>>> print(list(G.edges(data=True)))
[(0, 1, {'weight': 1.0})]
>>> G = bipartite.generic_weighted_projected_graph(B, [0, 1], weight_function=my_
˓→weight)

>>> print(list(G.edges(data=True)))
[(0, 1, {'weight': 10})]

Notes

No attempt is made to verify that the input graph B is bipartite. The graph and node properties are (shallow)
copied to the projected graph.
See bipartite documentation for further details on how bipartite graphs are handled in NetworkX.
See also:
is_bipartite(), is_bipartite_node_set(), sets(), weighted_projected_graph(),
collaboration_weighted_projected_graph(), overlap_weighted_projected_graph(),

164 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

projected_graph()

3.3.5 Spectral

Spectral bipartivity measure.

spectral_bipartivity(G[, nodes, weight]) Returns the spectral bipartivity.

networkx.algorithms.bipartite.spectral.spectral_bipartivity

spectral_bipartivity(G, nodes=None, weight=’weight’)


Returns the spectral bipartivity.
Parameters
• G (NetworkX graph)
• nodes (list or container optional(default is all nodes)) – Nodes to return value of spectral
bipartivity contribution.
• weight (string or None optional (default = ‘weight’)) – Edge data key to use for edge
weights. If None, weights set to 1.
Returns sb – A single number if the keyword nodes is not specified, or a dictionary keyed by node
with the spectral bipartivity contribution of that node as the value.
Return type float or dict

Examples

>>> from networkx.algorithms import bipartite


>>> G = nx.path_graph(4)
>>> bipartite.spectral_bipartivity(G)
1.0

Notes

This implementation uses Numpy (dense) matrices which are not efficient for storing large sparse graphs.
See also:
color()

References

3.3.6 Clustering

clustering(G[, nodes, mode]) Compute a bipartite clustering coefficient for nodes.


average_clustering(G[, nodes, mode]) Compute the average bipartite clustering coefficient.
latapy_clustering(G[, nodes, mode]) Compute a bipartite clustering coefficient for nodes.
robins_alexander_clustering(G) Compute the bipartite clustering of G.

3.3. Bipartite 165


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.bipartite.cluster.clustering

clustering(G, nodes=None, mode=’dot’)


Compute a bipartite clustering coefficient for nodes.
The bipartie clustering coefficient is a measure of local density of connections defined as1 :
∑︀
𝑣∈𝑁 (𝑁 (𝑢)) 𝑐𝑢𝑣
𝑐𝑢 =
|𝑁 (𝑁 (𝑢))|

where N(N(u)) are the second order neighbors of u in G excluding u, and c_{uv} is the pairwise clustering
coefficient between nodes u and v.
The mode selects the function for c_{uv} which can be:
dot:
|𝑁 (𝑢) ∩ 𝑁 (𝑣)|
𝑐𝑢𝑣 =
|𝑁 (𝑢) ∪ 𝑁 (𝑣)|
min:
|𝑁 (𝑢) ∩ 𝑁 (𝑣)|
𝑐𝑢𝑣 =
𝑚𝑖𝑛(|𝑁 (𝑢)|, |𝑁 (𝑣)|)
max:
|𝑁 (𝑢) ∩ 𝑁 (𝑣)|
𝑐𝑢𝑣 =
𝑚𝑎𝑥(|𝑁 (𝑢)|, |𝑁 (𝑣)|)

Parameters
• G (graph) – A bipartite graph
• nodes (list or iterable (optional)) – Compute bipartite clustering for these nodes. The default
is all nodes in G.
• mode (string) – The pariwise bipartite clustering method to be used in the computation. It
must be “dot”, “max”, or “min”.
Returns clustering – A dictionary keyed by node with the clustering coefficient value.
Return type dictionary

Examples

>>> from networkx.algorithms import bipartite


>>> G = nx.path_graph(4) # path graphs are bipartite
>>> c = bipartite.clustering(G)
>>> c[0]
0.5
>>> c = bipartite.clustering(G,mode='min')
>>> c[0]
1.0

See also:
robins_alexander_clustering(), square_clustering(), average_clustering()
1 Latapy, Matthieu, Clémence Magnien, and Nathalie Del Vecchio (2008). Basic notions for the analysis of large two-mode networks. Social

Networks 30(1), 31–48.

166 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

networkx.algorithms.bipartite.cluster.average_clustering

average_clustering(G, nodes=None, mode=’dot’)


Compute the average bipartite clustering coefficient.
A clustering coefficient for the whole graph is the average,
1 ∑︁
𝐶= 𝑐𝑣 ,
𝑛
𝑣∈𝐺

where n is the number of nodes in G.


Similar measures for the two bipartite sets can be defined1
1 ∑︁
𝐶𝑋 = 𝑐𝑣 ,
|𝑋|
𝑣∈𝑋

where X is a bipartite set of G.


Parameters
• G (graph) – a bipartite graph
• nodes (list or iterable, optional) – A container of nodes to use in computing the average.
The nodes should be either the entire graph (the default) or one of the bipartite sets.
• mode (string) – The pariwise bipartite clustering method. It must be “dot”, “max”, or “min”
Returns clustering – The average bipartite clustering for the given set of nodes or the entire graph
if no nodes are specified.
Return type float

Examples

>>> from networkx.algorithms import bipartite


>>> G=nx.star_graph(3) # star graphs are bipartite
>>> bipartite.average_clustering(G)
0.75
>>> X,Y=bipartite.sets(G)
>>> bipartite.average_clustering(G,X)
0.0
>>> bipartite.average_clustering(G,Y)
1.0

See also:
clustering()

Notes

The container of nodes passed to this function must contain all of the nodes in one of the bipartite sets
(“top” or “bottom”) in order to compute the correct average bipartite clustering coefficients. See bipartite
documentation for further details on how bipartite graphs are handled in NetworkX.
1Latapy, Matthieu, Clémence Magnien, and Nathalie Del Vecchio (2008). Basic notions for the analysis of large two-mode networks. Social
Networks 30(1), 31–48.

3.3. Bipartite 167


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

networkx.algorithms.bipartite.cluster.latapy_clustering

latapy_clustering(G, nodes=None, mode=’dot’)


Compute a bipartite clustering coefficient for nodes.
The bipartie clustering coefficient is a measure of local density of connections defined as1 :
∑︀
𝑣∈𝑁 (𝑁 (𝑢)) 𝑐𝑢𝑣
𝑐𝑢 =
|𝑁 (𝑁 (𝑢))|
where N(N(u)) are the second order neighbors of u in G excluding u, and c_{uv} is the pairwise clustering
coefficient between nodes u and v.
The mode selects the function for c_{uv} which can be:
dot:
|𝑁 (𝑢) ∩ 𝑁 (𝑣)|
𝑐𝑢𝑣 =
|𝑁 (𝑢) ∪ 𝑁 (𝑣)|
min:
|𝑁 (𝑢) ∩ 𝑁 (𝑣)|
𝑐𝑢𝑣 =
𝑚𝑖𝑛(|𝑁 (𝑢)|, |𝑁 (𝑣)|)
max:
|𝑁 (𝑢) ∩ 𝑁 (𝑣)|
𝑐𝑢𝑣 =
𝑚𝑎𝑥(|𝑁 (𝑢)|, |𝑁 (𝑣)|)
Parameters
• G (graph) – A bipartite graph
• nodes (list or iterable (optional)) – Compute bipartite clustering for these nodes. The default
is all nodes in G.
• mode (string) – The pariwise bipartite clustering method to be used in the computation. It
must be “dot”, “max”, or “min”.
Returns clustering – A dictionary keyed by node with the clustering coefficient value.
Return type dictionary

Examples

>>> from networkx.algorithms import bipartite


>>> G = nx.path_graph(4) # path graphs are bipartite
>>> c = bipartite.clustering(G)
>>> c[0]
0.5
>>> c = bipartite.clustering(G,mode='min')
>>> c[0]
1.0

See also:
robins_alexander_clustering(), square_clustering(), average_clustering()
1Latapy, Matthieu, Clémence Magnien, and Nathalie Del Vecchio (2008). Basic notions for the analysis of large two-mode networks. Social
Networks 30(1), 31–48.

168 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

networkx.algorithms.bipartite.cluster.robins_alexander_clustering

robins_alexander_clustering(G)
Compute the bipartite clustering of G.
Robins and Alexander1 defined bipartite clustering coefficient as four times the number of four cycles C_4
divided by the number of three paths L_3 in a bipartite graph:
4 * 𝐶4
𝐶𝐶4 =
𝐿3
Parameters G (graph) – a bipartite graph
Returns clustering – The Robins and Alexander bipartite clustering for the input graph.
Return type float

Examples

>>> from networkx.algorithms import bipartite


>>> G = nx.davis_southern_women_graph()
>>> print(round(bipartite.robins_alexander_clustering(G), 3))
0.468

See also:
latapy_clustering(), square_clustering()

References

3.3.7 Redundancy

Node redundancy for bipartite graphs.

node_redundancy(G[, nodes]) Computes the node redundancy coefficients for the


nodes in the bipartite graph G.

networkx.algorithms.bipartite.redundancy.node_redundancy

node_redundancy(G, nodes=None)
Computes the node redundancy coefficients for the nodes in the bipartite graph G.
The redundancy coefficient of a node v is the fraction of pairs of neighbors of v that are both linked to other
nodes. In a one-mode projection these nodes would be linked together even if v were not there.
More formally, for any vertex v, the redundancy coefficient of ‘v‘ is defined by

|{{𝑢, 𝑤} ⊆ 𝑁 (𝑣), ∃𝑣 ′ ̸= 𝑣, (𝑣 ′ , 𝑢) ∈ 𝐸 and (𝑣 ′ , 𝑤) ∈ 𝐸}|


𝑟𝑐(𝑣) = |𝑁 (𝑣)|(|𝑁 (𝑣)|−1)
,
2

where N(v) is the set of neighbors of v in G.


1 Robins, G. and M. Alexander (2004). Small worlds among interlocking directors: Network structure and distance in bipartite graphs. Compu-
tational & Mathematical Organization Theory 10(1), 69–94.

3.3. Bipartite 169


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (graph) – A bipartite graph
• nodes (list or iterable (optional)) – Compute redundancy for these nodes. The default is all
nodes in G.
Returns redundancy – A dictionary keyed by node with the node redundancy value.
Return type dictionary

Examples

Compute the redundancy coefficient of each node in a graph:

>>> import networkx as nx


>>> from networkx.algorithms import bipartite
>>> G = nx.cycle_graph(4)
>>> rc = bipartite.node_redundancy(G)
>>> rc[0]
1.0

Compute the average redundancy for the graph:

>>> import networkx as nx


>>> from networkx.algorithms import bipartite
>>> G = nx.cycle_graph(4)
>>> rc = bipartite.node_redundancy(G)
>>> sum(rc.values()) / len(G)
1.0

Compute the average redundancy for a set of nodes:

>>> import networkx as nx


>>> from networkx.algorithms import bipartite
>>> G = nx.cycle_graph(4)
>>> rc = bipartite.node_redundancy(G)
>>> nodes = [0, 2]
>>> sum(rc[n] for n in nodes) / len(nodes)
1.0

Raises NetworkXError – If any of the nodes in the graph (or in nodes, if specified) has (out-
)degree less than two (which would result in division by zero, according to the definition of the
redundancy coefficient).

References

3.3.8 Centrality

closeness_centrality(G, nodes[, normalized]) Compute the closeness centrality for nodes in a bipartite
network.
degree_centrality(G, nodes) Compute the degree centrality for nodes in a bipartite
network.
Continued on next page

170 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 24 – continued from previous page


betweenness_centrality(G, nodes) Compute betweenness centrality for nodes in a bipartite
network.

networkx.algorithms.bipartite.centrality.closeness_centrality

closeness_centrality(G, nodes, normalized=True)


Compute the closeness centrality for nodes in a bipartite network.
The closeness of a node is the distance to all other nodes in the graph or in the case that the graph is not connected
to all other nodes in the connected component containing that node.
Parameters
• G (graph) – A bipartite network
• nodes (list or container) – Container with all nodes in one bipartite node set.
• normalized (bool, optional) – If True (default) normalize by connected component size.
Returns closeness – Dictionary keyed by node with bipartite closeness centrality as the value.
Return type dictionary
See also:
betweenness_centrality(), degree_centrality(), sets(), is_bipartite()

Notes

The nodes input parameter must contain all nodes in one bipartite node set, but the dictionary returned contains
all nodes from both node sets. See bipartite documentation for further details on how bipartite graphs
are handled in NetworkX.
Closeness centrality is normalized by the minimum distance possible. In the bipartite case the minimum distance
for a node in one bipartite node set is 1 from all nodes in the other node set and 2 from all other nodes in its own
set1 . Thus the closeness centrality for node v in the two bipartite sets U with n nodes and V with m nodes is

𝑚 + 2(𝑛 − 1)
𝑐𝑣 = , for𝑣 ∈ 𝑈,
𝑑
𝑛 + 2(𝑚 − 1)
𝑐𝑣 = , for𝑣 ∈ 𝑉,
𝑑
where d is the sum of the distances from v to all other nodes.
Higher values of closeness indicate higher centrality.
As in the unipartite case, setting normalized=True causes the values to normalized further to n-1 / size(G)-1
where n is the number of nodes in the connected part of graph containing the node. If the graph is not completely
connected, this algorithm computes the closeness centrality for each connected part separately.

References

networkx.algorithms.bipartite.centrality.degree_centrality

degree_centrality(G, nodes)
Compute the degree centrality for nodes in a bipartite network.
1Borgatti, S.P. and Halgin, D. In press. “Analyzing Affiliation Networks”. In Carrington, P. and Scott, J. (eds) The Sage Handbook of Social
Network Analysis. Sage Publications. http://www.steveborgatti.com/research/publications/bhaffiliations.pdf

3.3. Bipartite 171


NetworkX Reference, Release 2.3rc1.dev20190222214247

The degree centrality for a node v is the fraction of nodes connected to it.
Parameters
• G (graph) – A bipartite network
• nodes (list or container) – Container with all nodes in one bipartite node set.
Returns centrality – Dictionary keyed by node with bipartite degree centrality as the value.
Return type dictionary
See also:
betweenness_centrality(), closeness_centrality(), sets(), is_bipartite()

Notes

The nodes input parameter must contain all nodes in one bipartite node set, but the dictionary returned con-
tains all nodes from both bipartite node sets. See bipartite documentation for further details on how
bipartite graphs are handled in NetworkX.
For unipartite networks, the degree centrality values are normalized by dividing by the maximum possible degree
(which is n-1 where n is the number of nodes in G).
In the bipartite case, the maximum possible degree of a node in a bipartite node set is the number of nodes in the
opposite node set1 . The degree centrality for a node v in the bipartite sets U with n nodes and V with m nodes is

𝑑𝑒𝑔(𝑣)
𝑑𝑣 = , for𝑣 ∈ 𝑈,
𝑚
𝑑𝑒𝑔(𝑣)
𝑑𝑣 = , for𝑣 ∈ 𝑉,
𝑛
where deg(v) is the degree of node v.

References

networkx.algorithms.bipartite.centrality.betweenness_centrality

betweenness_centrality(G, nodes)
Compute betweenness centrality for nodes in a bipartite network.
Betweenness centrality of a node v is the sum of the fraction of all-pairs shortest paths that pass through v.
Values of betweenness are normalized by the maximum possible value which for bipartite graphs is limited by
the relative size of the two node sets1 .
Let n be the number of nodes in the node set U and m be the number of nodes in the node set V, then nodes in U
are normalized by dividing by
1 2
[𝑚 (𝑠 + 1)2 + 𝑚(𝑠 + 1)(2𝑡 − 𝑠 − 1) − 𝑡(2𝑠 − 𝑡 + 3)],
2
where

𝑠 = (𝑛 − 1) ÷ 𝑚, 𝑡 = (𝑛 − 1) mod 𝑚,
1 Borgatti, S.P. and Halgin, D. In press. “Analyzing Affiliation Networks”. In Carrington, P. and Scott, J. (eds) The Sage Handbook of Social

Network Analysis. Sage Publications. http://www.steveborgatti.com/research/publications/bhaffiliations.pdf


1 Borgatti, S.P. and Halgin, D. In press. “Analyzing Affiliation Networks”. In Carrington, P. and Scott, J. (eds) The Sage Handbook of Social

Network Analysis. Sage Publications. http://www.steveborgatti.com/research/publications/bhaffiliations.pdf

172 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

and nodes in V are normalized by dividing by


1 2
[𝑛 (𝑝 + 1)2 + 𝑛(𝑝 + 1)(2𝑟 − 𝑝 − 1) − 𝑟(2𝑝 − 𝑟 + 3)],
2
where,

𝑝 = (𝑚 − 1) ÷ 𝑛, 𝑟 = (𝑚 − 1) mod 𝑛.

Parameters
• G (graph) – A bipartite graph
• nodes (list or container) – Container with all nodes in one bipartite node set.
Returns betweenness – Dictionary keyed by node with bipartite betweenness centrality as the value.
Return type dictionary
See also:
degree_centrality(), closeness_centrality(), sets(), is_bipartite()

Notes

The nodes input parameter must contain all nodes in one bipartite node set, but the dictionary returned contains
all nodes from both node sets. See bipartite documentation for further details on how bipartite graphs
are handled in NetworkX.

References

3.3.9 Generators

Generators and functions for bipartite graphs.

complete_bipartite_graph(n1, n2[, cre- Returns the complete bipartite graph K_{n_1,n_2}.


ate_using])
configuration_model(aseq, bseq[, . . . ]) Returns a random bipartite graph from two given degree
sequences.
havel_hakimi_graph(aseq, bseq[, create_using]) Returns a bipartite graph from two given degree se-
quences using a Havel-Hakimi style construction.
reverse_havel_hakimi_graph(aseq, bseq[, Returns a bipartite graph from two given degree se-
. . . ]) quences using a Havel-Hakimi style construction.
alternating_havel_hakimi_graph(aseq, Returns a bipartite graph from two given degree se-
bseq[, . . . ]) quences using an alternating Havel-Hakimi style con-
struction.
preferential_attachment_graph(aseq, p[, Create a bipartite graph with a preferential attachment
. . . ]) model from a given single degree sequence.
random_graph(n, m, p[, seed, directed]) Returns a bipartite random graph.
gnmk_random_graph(n, m, k[, seed, directed]) Returns a random bipartite graph G_{n,m,k}.

3.3. Bipartite 173


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.bipartite.generators.complete_bipartite_graph

complete_bipartite_graph(n1, n2, create_using=None)


Returns the complete bipartite graph K_{n_1,n_2}.
Composed of two partitions with n_1 nodes in the first and n_2 nodes in the second. Each node in the first is
connected to each node in the second.
Parameters
• n1 (integer) – Number of nodes for node set A.
• n2 (integer) – Number of nodes for node set B.
• create_using (NetworkX graph instance, optional) – Return graph of this type.

Notes

Node labels are the integers 0 to n_1 + n_2 - 1.


The nodes are assigned the attribute ‘bipartite’ with the value 0 or 1 to indicate which bipartite set the node
belongs to.

networkx.algorithms.bipartite.generators.configuration_model

configuration_model(aseq, bseq, create_using=None, seed=None)


Returns a random bipartite graph from two given degree sequences.
Parameters
• aseq (list) – Degree sequence for node set A.
• bseq (list) – Degree sequence for node set B.
• create_using (NetworkX graph instance, optional) – Return graph of this type.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• Nodes from the set A are connected to nodes in the set B by
• choosing randomly from the possible free stubs, one in A and
• one in B.

Notes

The sum of the two sequences must be equal: sum(aseq)=sum(bseq) If no graph type is specified use MultiGraph
with parallel edges. If you want a graph with no parallel edges use create_using=Graph() but then the resulting
degree sequences might not be exact.
The nodes are assigned the attribute ‘bipartite’ with the value 0 or 1 to indicate which bipartite set the node
belongs to.
This function is not imported in the main namespace. To use it you have to explicitly import the bipartite
package.

174 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.bipartite.generators.havel_hakimi_graph

havel_hakimi_graph(aseq, bseq, create_using=None)


Returns a bipartite graph from two given degree sequences using a Havel-Hakimi style construction.
Nodes from the set A are connected to nodes in the set B by connecting the highest degree nodes in set A to the
highest degree nodes in set B until all stubs are connected.
Parameters
• aseq (list) – Degree sequence for node set A.
• bseq (list) – Degree sequence for node set B.
• create_using (NetworkX graph instance, optional) – Return graph of this type.

Notes

This function is not imported in the main namespace. To use it you have to explicitly import the bipartite
package.
The sum of the two sequences must be equal: sum(aseq)=sum(bseq) If no graph type is specified use MultiGraph
with parallel edges. If you want a graph with no parallel edges use create_using=Graph() but then the resulting
degree sequences might not be exact.
The nodes are assigned the attribute ‘bipartite’ with the value 0 or 1 to indicate which bipartite set the node
belongs to.

networkx.algorithms.bipartite.generators.reverse_havel_hakimi_graph

reverse_havel_hakimi_graph(aseq, bseq, create_using=None)


Returns a bipartite graph from two given degree sequences using a Havel-Hakimi style construction.
Nodes from set A are connected to nodes in the set B by connecting the highest degree nodes in set A to the
lowest degree nodes in set B until all stubs are connected.
Parameters
• aseq (list) – Degree sequence for node set A.
• bseq (list) – Degree sequence for node set B.
• create_using (NetworkX graph instance, optional) – Return graph of this type.

Notes

This function is not imported in the main namespace. To use it you have to explicitly import the bipartite
package.
The sum of the two sequences must be equal: sum(aseq)=sum(bseq) If no graph type is specified use MultiGraph
with parallel edges. If you want a graph with no parallel edges use create_using=Graph() but then the resulting
degree sequences might not be exact.
The nodes are assigned the attribute ‘bipartite’ with the value 0 or 1 to indicate which bipartite set the node
belongs to.

3.3. Bipartite 175


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.bipartite.generators.alternating_havel_hakimi_graph

alternating_havel_hakimi_graph(aseq, bseq, create_using=None)


Returns a bipartite graph from two given degree sequences using an alternating Havel-Hakimi style construction.
Nodes from the set A are connected to nodes in the set B by connecting the highest degree nodes in set A to
alternatively the highest and the lowest degree nodes in set B until all stubs are connected.
Parameters
• aseq (list) – Degree sequence for node set A.
• bseq (list) – Degree sequence for node set B.
• create_using (NetworkX graph instance, optional) – Return graph of this type.

Notes

This function is not imported in the main namespace. To use it you have to explicitly import the bipartite
package.
The sum of the two sequences must be equal: sum(aseq)=sum(bseq) If no graph type is specified use MultiGraph
with parallel edges. If you want a graph with no parallel edges use create_using=Graph() but then the resulting
degree sequences might not be exact.
The nodes are assigned the attribute ‘bipartite’ with the value 0 or 1 to indicate which bipartite set the node
belongs to.

networkx.algorithms.bipartite.generators.preferential_attachment_graph

preferential_attachment_graph(aseq, p, create_using=None, seed=None)


Create a bipartite graph with a preferential attachment model from a given single degree sequence.
Parameters
• aseq (list) – Degree sequence for node set A.
• p (float) – Probability that a new bottom node is added.
• create_using (NetworkX graph instance, optional) – Return graph of this type.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

References

Notes

This function is not imported in the main namespace. To use it you have to explicitly import the bipartite
package.

networkx.algorithms.bipartite.generators.random_graph

random_graph(n, m, p, seed=None, directed=False)


Returns a bipartite random graph.
This is a bipartite version of the binomial (Erdős-Rényi) graph.

176 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• n (int) – The number of nodes in the first bipartite set.
• m (int) – The number of nodes in the second bipartite set.
• p (float) – Probability for edge creation.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• directed (bool, optional (default=False)) – If True return a directed graph

Notes

This function is not imported in the main namespace. To use it you have to explicitly import the bipartite
package.
The bipartite random graph algorithm chooses each of the n*m (undirected) or 2*nm (directed) possible edges
with probability p.
This algorithm is 𝑂(𝑛 + 𝑚) where 𝑚 is the expected number of edges.
The nodes are assigned the attribute ‘bipartite’ with the value 0 or 1 to indicate which bipartite set the node
belongs to.
See also:
gnp_random_graph(), configuration_model()

References

networkx.algorithms.bipartite.generators.gnmk_random_graph

gnmk_random_graph(n, m, k, seed=None, directed=False)


Returns a random bipartite graph G_{n,m,k}.
Produces a bipartite graph chosen randomly out of the set of all graphs with n top nodes, m bottom nodes, and
k edges.
Parameters
• n (int) – The number of nodes in the first bipartite set.
• m (int) – The number of nodes in the second bipartite set.
• k (int) – The number of edges
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• directed (bool, optional (default=False)) – If True return a directed graph

Examples

from nx.algorithms import bipartite G = bipartite.gnmk_random_graph(10,20,50)


See also:
gnm_random_graph()

3.3. Bipartite 177


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This function is not imported in the main namespace. To use it you have to explicitly import the bipartite
package.
If k > m * n then a complete bipartite graph is returned.
This graph is a bipartite version of the G_{nm} random graph model.

3.3.10 Covering

Functions related to graph covers.

min_edge_cover(G[, matching_algorithm]) Returns a set of edges which constitutes the minimum


edge cover of the graph.

networkx.algorithms.bipartite.covering.min_edge_cover

min_edge_cover(G, matching_algorithm=None)
Returns a set of edges which constitutes the minimum edge cover of the graph.
The smallest edge cover can be found in polynomial time by finding a maximum matching and extending it
greedily so that all nodes are covered.
Parameters
• G (NetworkX graph) – An undirected bipartite graph.
• matching_algorithm (function) – A function that returns a maximum cardinality matching
in a given bipartite graph. The function must take one input, the graph G, and return a dic-
tionary mapping each node to its mate. If not specified, hopcroft_karp_matching()
will be used. Other possibilities include eppstein_matching(),
Returns A set of the edges in a minimum edge cover of the graph, given as pairs of nodes. It contains
both the edges (u, v) and (v, u) for given nodes u and v among the edges of minimum
edge cover.
Return type set

Notes

An edge cover of a graph is a set of edges such that every node of the graph is incident to at least one edge of
the set. A minimum edge cover is an edge covering of smallest cardinality.
Due to its implementation, the worst-case running time of this algorithm is bounded by the worst-case running
time of the function matching_algorithm.

3.4 Boundary

Routines to find the boundary of a set of nodes.


An edge boundary is a set of edges, each of which has exactly one endpoint in a given set of nodes (or, in the case of
directed graphs, the set of edges whose source node is in the set).
A node boundary of a set S of nodes is the set of (out-)neighbors of nodes in S that are outside S.

178 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

edge_boundary(G, nbunch1[, nbunch2, data, . . . ]) Returns the edge boundary of nbunch1.


node_boundary(G, nbunch1[, nbunch2]) Returns the node boundary of nbunch1.

3.4.1 networkx.algorithms.boundary.edge_boundary

edge_boundary(G, nbunch1, nbunch2=None, data=False, keys=False, default=None)


Returns the edge boundary of nbunch1.
The edge boundary of a set S with respect to a set T is the set of edges (u, v) such that u is in S and v is in T. If
T is not specified, it is assumed to be the set of all nodes not in S.
Parameters
• G (NetworkX graph)
• nbunch1 (iterable) – Iterable of nodes in the graph representing the set of nodes whose edge
boundary will be returned. (This is the set S from the definition above.)
• nbunch2 (iterable) – Iterable of nodes representing the target (or “exterior”) set of nodes.
(This is the set T from the definition above.) If not specified, this is assumed to be the set of
all nodes in G not in nbunch1.
• keys (bool) – This parameter has the same meaning as in MultiGraph.edges().
• data (bool or object) – This parameter has the same meaning as in MultiGraph.
edges().
• default (object) – This parameter has the same meaning as in MultiGraph.edges().
Returns An iterator over the edges in the boundary of nbunch1 with respect to nbunch2. If
keys, data, or default are specified and G is a multigraph, then edges are returned with
keys and/or data, as in MultiGraph.edges().
Return type iterator

Notes

Any element of nbunch that is not in the graph G will be ignored.


nbunch1 and nbunch2 are usually meant to be disjoint, but in the interest of speed and generality, that is not
required here.

3.4.2 networkx.algorithms.boundary.node_boundary

node_boundary(G, nbunch1, nbunch2=None)


Returns the node boundary of nbunch1.
The node boundary of a set S with respect to a set T is the set of nodes v in T such that for some u in S, there is
an edge joining u to v. If T is not specified, it is assumed to be the set of all nodes not in S.
Parameters
• G (NetworkX graph)
• nbunch1 (iterable) – Iterable of nodes in the graph representing the set of nodes whose node
boundary will be returned. (This is the set S from the definition above.)

3.4. Boundary 179


NetworkX Reference, Release 2.3rc1.dev20190222214247

• nbunch2 (iterable) – Iterable of nodes representing the target (or “exterior”) set of nodes.
(This is the set T from the definition above.) If not specified, this is assumed to be the set of
all nodes in G not in nbunch1.
Returns The node boundary of nbunch1 with respect to nbunch2.
Return type set

Notes

Any element of nbunch that is not in the graph G will be ignored.


nbunch1 and nbunch2 are usually meant to be disjoint, but in the interest of speed and generality, that is not
required here.

3.5 Bridges

Bridge-finding algorithms.

bridges(G[, root]) Generate all bridges in a graph.


has_bridges(G[, root]) Decide whether a graph has any bridges.
local_bridges(G[, with_span, weight]) Iterate over local bridges of G optionally computing the
span

3.5.1 networkx.algorithms.bridges.bridges

bridges(G, root=None)
Generate all bridges in a graph.
A bridge in a graph is an edge whose removal causes the number of connected components of the graph to
increase. Equivalently, a bridge is an edge that does not belong to any cycle.
Parameters
• G (undirected graph)
• root (node (optional)) – A node in the graph G. If specified, only the bridges in the connected
component containing this node will be returned.
Yields e (edge) – An edge in the graph whose removal disconnects the graph (or causes the number
of connected components to increase).
Raises NodeNotFound – If root is not in the graph G.

Examples

The barbell graph with parameter zero has a single bridge:

>>> G = nx.barbell_graph(10, 0)
>>> list(nx.bridges(G))
[(9, 10)]

180 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This is an implementation of the algorithm described in _[1]. An edge is a bridge if and only if it is not contained
in any chain. Chains are found using the networkx.chain_decomposition() function.
Ignoring polylogarithmic factors, the worst-case time complexity is the same as the networkx.
chain_decomposition() function, 𝑂(𝑚 + 𝑛), where 𝑛 is the number of nodes in the graph and 𝑚 is
the number of edges.

References

3.5.2 networkx.algorithms.bridges.has_bridges

has_bridges(G, root=None)
Decide whether a graph has any bridges.
A bridge in a graph is an edge whose removal causes the number of connected components of the graph to
increase.
Parameters
• G (undirected graph)
• root (node (optional)) – A node in the graph G. If specified, only the bridges in the connected
component containing this node will be considered.
Returns Whether the graph (or the connected component containing root) has any bridges.
Return type bool
Raises NodeNotFound – If root is not in the graph G.

Examples

The barbell graph with parameter zero has a single bridge:

>>> G = nx.barbell_graph(10, 0)
>>> nx.has_bridges(G)
True

On the other hand, the cycle graph has no bridges:

>>> G = nx.cycle_graph(5)
>>> nx.has_bridges(G)
False

Notes

This implementation uses the networkx.bridges() function, so it shares its worst-case time complexity,
𝑂(𝑚 + 𝑛), ignoring polylogarithmic factors, where 𝑛 is the number of nodes in the graph and 𝑚 is the number
of edges.

3.5. Bridges 181


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.5.3 networkx.algorithms.bridges.local_bridges

local_bridges(G, with_span=True, weight=None)


Iterate over local bridges of G optionally computing the span
A local bridge is an edge whose endpoints have no common neighbors. That is, the edge is not part of a triangle
in the graph.
The span of a local bridge is the shortest path length between the endpoints if the local bridge is removed.
Parameters
• G (undirected graph)
• with_span (bool) – If True, yield a 3-tuple (u, v, span)
• weight (function, string or None (default: None)) – If function, used to compute edge
weights for the span. If string, the edge data attribute used in calculating span. If None,
all edges have weight 1.
Yields e (edge) – The local bridges as an edge 2-tuple of nodes (u, v) or as a 3-tuple (u, v,
span) when with_span is True.

Examples

A cycle graph has every edge a local bridge with span N-1.

>>> G = nx.cycle_graph(9)
>>> (0, 8, 8) in set(nx.local_bridges(G))
True

3.6 Centrality

3.6.1 Degree

degree_centrality(G) Compute the degree centrality for nodes.


in_degree_centrality(G) Compute the in-degree centrality for nodes.
out_degree_centrality(G) Compute the out-degree centrality for nodes.

networkx.algorithms.centrality.degree_centrality

degree_centrality(G)
Compute the degree centrality for nodes.
The degree centrality for a node v is the fraction of nodes it is connected to.
Parameters G (graph) – A networkx graph
Returns nodes – Dictionary of nodes with degree centrality as the value.
Return type dictionary
See also:
betweenness_centrality(), load_centrality(), eigenvector_centrality()

182 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The degree centrality values are normalized by dividing by the maximum possible degree in a simple graph n-1
where n is the number of nodes in G.
For multigraphs or graphs with self loops the maximum degree might be higher than n-1 and values of degree
centrality greater than 1 are possible.

networkx.algorithms.centrality.in_degree_centrality

in_degree_centrality(G)
Compute the in-degree centrality for nodes.
The in-degree centrality for a node v is the fraction of nodes its incoming edges are connected to.
Parameters G (graph) – A NetworkX graph
Returns nodes – Dictionary of nodes with in-degree centrality as values.
Return type dictionary
Raises NetworkXNotImplemented: – If G is undirected.
See also:
degree_centrality(), out_degree_centrality()

Notes

The degree centrality values are normalized by dividing by the maximum possible degree in a simple graph n-1
where n is the number of nodes in G.
For multigraphs or graphs with self loops the maximum degree might be higher than n-1 and values of degree
centrality greater than 1 are possible.

networkx.algorithms.centrality.out_degree_centrality

out_degree_centrality(G)
Compute the out-degree centrality for nodes.
The out-degree centrality for a node v is the fraction of nodes its outgoing edges are connected to.
Parameters G (graph) – A NetworkX graph
Returns nodes – Dictionary of nodes with out-degree centrality as values.
Return type dictionary
Raises NetworkXNotImplemented: – If G is undirected.
See also:
degree_centrality(), in_degree_centrality()

3.6. Centrality 183


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The degree centrality values are normalized by dividing by the maximum possible degree in a simple graph n-1
where n is the number of nodes in G.
For multigraphs or graphs with self loops the maximum degree might be higher than n-1 and values of degree
centrality greater than 1 are possible.

3.6.2 Eigenvector

eigenvector_centrality(G[, max_iter, tol, Compute the eigenvector centrality for the graph G.
. . . ])
eigenvector_centrality_numpy(G[, weight, Compute the eigenvector centrality for the graph G.
. . . ])
katz_centrality(G[, alpha, beta, max_iter, . . . ]) Compute the Katz centrality for the nodes of the graph
G.
katz_centrality_numpy(G[, alpha, beta, . . . ]) Compute the Katz centrality for the graph G.

networkx.algorithms.centrality.eigenvector_centrality

eigenvector_centrality(G, max_iter=100, tol=1e-06, nstart=None, weight=None)


Compute the eigenvector centrality for the graph G.
Eigenvector centrality computes the centrality for a node based on the centrality of its neighbors. The eigenvec-
tor centrality for node 𝑖 is the 𝑖-th element of the vector 𝑥 defined by the equation

𝐴𝑥 = 𝜆𝑥

where 𝐴 is the adjacency matrix of the graph G with eigenvalue 𝜆. By virtue of the Perron–Frobenius theorem,
there is a unique solution 𝑥, all of whose entries are positive, if 𝜆 is the largest eigenvalue of the adjacency
matrix 𝐴 (2 ).
Parameters
• G (graph) – A networkx graph
• max_iter (integer, optional (default=100)) – Maximum number of iterations in power
method.
• tol (float, optional (default=1.0e-6)) – Error tolerance used to check convergence in power
method iteration.
• nstart (dictionary, optional (default=None)) – Starting value of eigenvector iteration for
each node.
• weight (None or string, optional (default=None)) – If None, all edge weights are considered
equal. Otherwise holds the name of the edge attribute used as weight.
Returns nodes – Dictionary of nodes with eigenvector centrality as the value.
Return type dictionary
2 Mark E. J. Newman. Networks: An Introduction. Oxford University Press, USA, 2010, pp. 169.

184 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(4)
>>> centrality = nx.eigenvector_centrality(G)
>>> sorted((v, '{:0.2f}'.format(c)) for v, c in centrality.items())
[(0, '0.37'), (1, '0.60'), (2, '0.60'), (3, '0.37')]

Raises
• NetworkXPointlessConcept – If the graph G is the null graph.
• NetworkXError – If each value in nstart is zero.
• PowerIterationFailedConvergence – If the algorithm fails to converge to the
specified tolerance within the specified number of iterations of the power iteration method.

See also:
eigenvector_centrality_numpy(), pagerank(), hits()

Notes

The measure was introduced by1 and is discussed in2 .


The power iteration method is used to compute the eigenvector and convergence is not guaranteed. Our method
stops after max_iter iterations or when the change in the computed vector between two iterations is smaller
than an error tolerance of G.number_of_nodes() * tol. This implementation uses (𝐴 + 𝐼) rather than
the adjacency matrix 𝐴 because it shifts the spectrum to enable discerning the correct eigenvector even for
networks with multiple dominant eigenvalues.
For directed graphs this is “left” eigenvector centrality which corresponds to the in-edges in the graph. For
out-edges eigenvector centrality first reverse the graph with G.reverse().

References

networkx.algorithms.centrality.eigenvector_centrality_numpy

eigenvector_centrality_numpy(G, weight=None, max_iter=50, tol=0)


Compute the eigenvector centrality for the graph G.
Eigenvector centrality computes the centrality for a node based on the centrality of its neighbors. The eigenvec-
tor centrality for node 𝑖 is

𝐴𝑥 = 𝜆𝑥

where 𝐴 is the adjacency matrix of the graph G with eigenvalue 𝜆. By virtue of the Perron–Frobenius theorem,
there is a unique and positive solution if 𝜆 is the largest eigenvalue associated with the eigenvector of the
adjacency matrix 𝐴 (2 ).
Parameters
• G (graph) – A networkx graph
1 Phillip Bonacich. “Power and Centrality: A Family of Measures.” American Journal of Sociology 92(5):1170–1182, 1986 <http://www.
leonidzhukov.net/hse/2014/socialnetworks/papers/Bonacich-Centrality.pdf>
2 Mark E. J. Newman: Networks: An Introduction. Oxford University Press, USA, 2010, pp. 169.

3.6. Centrality 185


NetworkX Reference, Release 2.3rc1.dev20190222214247

• weight (None or string, optional (default=None)) – The name of the edge attribute used as
weight. If None, all edge weights are considered equal.
• max_iter (integer, optional (default=100)) – Maximum number of iterations in power
method.
• tol (float, optional (default=1.0e-6)) – Relative accuracy for eigenvalues (stopping crite-
rion). The default value of 0 implies machine precision.
Returns nodes – Dictionary of nodes with eigenvector centrality as the value.
Return type dictionary

Examples

>>> G = nx.path_graph(4)
>>> centrality = nx.eigenvector_centrality_numpy(G)
>>> print(['{} {:0.2f}'.format(node, centrality[node]) for node in centrality])
['0 0.37', '1 0.60', '2 0.60', '3 0.37']

See also:
eigenvector_centrality(), pagerank(), hits()

Notes

The measure was introduced by1 .


This algorithm uses the SciPy sparse eigenvalue solver (ARPACK) to find the largest eigenvalue/eigenvector
pair.
For directed graphs this is “left” eigenvector centrality which corresponds to the in-edges in the graph. For
out-edges eigenvector centrality first reverse the graph with G.reverse().
Raises NetworkXPointlessConcept – If the graph G is the null graph.

References

networkx.algorithms.centrality.katz_centrality

katz_centrality(G, alpha=0.1, beta=1.0, max_iter=1000, tol=1e-06, nstart=None, normalized=True,


weight=None)
Compute the Katz centrality for the nodes of the graph G.
Katz centrality computes the centrality for a node based on the centrality of its neighbors. It is a generalization
of the eigenvector centrality. The Katz centrality for node 𝑖 is
∑︁
𝑥𝑖 = 𝛼 𝐴𝑖𝑗 𝑥𝑗 + 𝛽,
𝑗

where 𝐴 is the adjacency matrix of graph G with eigenvalues 𝜆.


The parameter 𝛽 controls the initial centrality and
1
𝛼< .
𝜆max
1 Phillip Bonacich: Power and Centrality: A Family of Measures. American Journal of Sociology 92(5):1170–1182, 1986 http://www.
leonidzhukov.net/hse/2014/socialnetworks/papers/Bonacich-Centrality.pdf

186 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Katz centrality computes the relative influence of a node within a network by measuring the number of the
immediate neighbors (first degree nodes) and also all other nodes in the network that connect to the node under
consideration through these immediate neighbors.
Extra weight can be provided to immediate neighbors through the parameter 𝛽. Connections made with distant
neighbors are, however, penalized by an attenuation factor 𝛼 which should be strictly less than the inverse largest
eigenvalue of the adjacency matrix in order for the Katz centrality to be computed correctly. More information
is provided in1 .
Parameters
• G (graph) – A NetworkX graph.
• alpha (float) – Attenuation factor
• beta (scalar or dictionary, optional (default=1.0)) – Weight attributed to the immediate
neighborhood. If not a scalar, the dictionary must have an value for every node.
• max_iter (integer, optional (default=1000)) – Maximum number of iterations in power
method.
• tol (float, optional (default=1.0e-6)) – Error tolerance used to check convergence in power
method iteration.
• nstart (dictionary, optional) – Starting value of Katz iteration for each node.
• normalized (bool, optional (default=True)) – If True normalize the resulting values.
• weight (None or string, optional (default=None)) – If None, all edge weights are considered
equal. Otherwise holds the name of the edge attribute used as weight.
Returns nodes – Dictionary of nodes with Katz centrality as the value.
Return type dictionary
Raises
• NetworkXError – If the parameter beta is not a scalar but lacks a value for at least one
node
• PowerIterationFailedConvergence – If the algorithm fails to converge to the
specified tolerance within the specified number of iterations of the power iteration method.

Examples

>>> import math


>>> G = nx.path_graph(4)
>>> phi = (1 + math.sqrt(5)) / 2.0 # largest eigenvalue of adj matrix
>>> centrality = nx.katz_centrality(G, 1/phi - 0.01)
>>> for n, c in sorted(centrality.items()):
... print("%d %0.2f" % (n, c))
0 0.37
1 0.60
2 0.60
3 0.37

See also:
katz_centrality_numpy(), eigenvector_centrality(), eigenvector_centrality_numpy(),
pagerank(), hits()
1 Mark E. J. Newman: Networks: An Introduction. Oxford University Press, USA, 2010, p. 720.

3.6. Centrality 187


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Katz centrality was introduced by2 .


This algorithm it uses the power method to find the eigenvector corresponding to the largest eigenvalue of the
adjacency matrix of G. The parameter alpha should be strictly less than the inverse of largest eigenvalue of the
adjacency matrix for the algorithm to converge. You can use max(nx.adjacency_spectrum(G)) to get
𝜆max the largest eigenvalue of the adjacency matrix. The iteration will stop after max_iter iterations or an
error tolerance of number_of_nodes(G) * tol has been reached.
When 𝛼 = 1/𝜆max and 𝛽 = 0, Katz centrality is the same as eigenvector centrality.
For directed graphs this finds “left” eigenvectors which corresponds to the in-edges in the graph. For out-edges
Katz centrality first reverse the graph with G.reverse().

References

networkx.algorithms.centrality.katz_centrality_numpy

katz_centrality_numpy(G, alpha=0.1, beta=1.0, normalized=True, weight=None)


Compute the Katz centrality for the graph G.
Katz centrality computes the centrality for a node based on the centrality of its neighbors. It is a generalization
of the eigenvector centrality. The Katz centrality for node 𝑖 is
∑︁
𝑥𝑖 = 𝛼 𝐴𝑖𝑗 𝑥𝑗 + 𝛽,
𝑗

where 𝐴 is the adjacency matrix of graph G with eigenvalues 𝜆.


The parameter 𝛽 controls the initial centrality and
1
𝛼< .
𝜆max
Katz centrality computes the relative influence of a node within a network by measuring the number of the
immediate neighbors (first degree nodes) and also all other nodes in the network that connect to the node under
consideration through these immediate neighbors.
Extra weight can be provided to immediate neighbors through the parameter 𝛽. Connections made with distant
neighbors are, however, penalized by an attenuation factor 𝛼 which should be strictly less than the inverse largest
eigenvalue of the adjacency matrix in order for the Katz centrality to be computed correctly. More information
is provided in1 .
Parameters
• G (graph) – A NetworkX graph
• alpha (float) – Attenuation factor
• beta (scalar or dictionary, optional (default=1.0)) – Weight attributed to the immediate
neighborhood. If not a scalar the dictionary must have an value for every node.
• normalized (bool) – If True normalize the resulting values.
• weight (None or string, optional) – If None, all edge weights are considered equal. Other-
wise holds the name of the edge attribute used as weight.
2 Leo Katz: A New Status Index Derived from Sociometric Index. Psychometrika 18(1):39–43, 1953 http://phya.snu.ac.kr/~dkim/
PRL87278701.pdf
1 Mark E. J. Newman: Networks: An Introduction. Oxford University Press, USA, 2010, p. 720.

188 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns nodes – Dictionary of nodes with Katz centrality as the value.


Return type dictionary
Raises NetworkXError – If the parameter beta is not a scalar but lacks a value for at least one
node

Examples

>>> import math


>>> G = nx.path_graph(4)
>>> phi = (1 + math.sqrt(5)) / 2.0 # largest eigenvalue of adj matrix
>>> centrality = nx.katz_centrality_numpy(G, 1/phi)
>>> for n, c in sorted(centrality.items()):
... print("%d %0.2f" % (n, c))
0 0.37
1 0.60
2 0.60
3 0.37

See also:
katz_centrality(), eigenvector_centrality_numpy(), eigenvector_centrality(),
pagerank(), hits()

Notes

Katz centrality was introduced by2 .


This algorithm uses a direct linear solver to solve the above equation. The parameter alpha should be strictly
less than the inverse of largest eigenvalue of the adjacency matrix for there to be a solution. You can use
max(nx.adjacency_spectrum(G)) to get 𝜆max the largest eigenvalue of the adjacency matrix.
When 𝛼 = 1/𝜆max and 𝛽 = 0, Katz centrality is the same as eigenvector centrality.
For directed graphs this finds “left” eigenvectors which corresponds to the in-edges in the graph. For out-edges
Katz centrality first reverse the graph with G.reverse().

References

3.6.3 Closeness

closeness_centrality(G[, u, distance, . . . ]) Compute closeness centrality for nodes.

networkx.algorithms.centrality.closeness_centrality

closeness_centrality(G, u=None, distance=None, wf_improved=True, reverse=False)


Compute closeness centrality for nodes.
Closeness centrality1 of a node u is the reciprocal of the average shortest path distance to u over all n-1
2 Leo Katz: A New Status Index Derived from Sociometric Index. Psychometrika 18(1):39–43, 1953 http://phya.snu.ac.kr/~dkim/
PRL87278701.pdf
1 Linton C. Freeman: Centrality in networks: I. Conceptual clarification. Social Networks 1:215-239, 1979. http://leonidzhukov.ru/hse/2013/

socialnetworks/papers/freeman79-centrality.pdf

3.6. Centrality 189


NetworkX Reference, Release 2.3rc1.dev20190222214247

reachable nodes.
𝑛−1
𝐶(𝑢) = ∑︀𝑛−1 ,
𝑣=1 𝑑(𝑣, 𝑢)

where d(v, u) is the shortest-path distance between v and u, and n is the number of nodes that can reach u.
Notice that higher values of closeness indicate higher centrality.
Wasserman and Faust propose an improved formula for graphs with more than one connected component. The
result is “a ratio of the fraction of actors in the group who are reachable, to the average distance” from the
reachable actors2 . You might think this scale factor is inverted but it is not. As is, nodes from small components
receive a smaller closeness value. Letting N denote the number of nodes in the graph,
𝑛−1 𝑛−1
𝐶𝑊 𝐹 (𝑢) = ∑︀𝑛−1 ,
𝑁 − 1 𝑣=1 𝑑(𝑣, 𝑢)

Parameters
• G (graph) – A NetworkX graph
• u (node, optional) – Return only the value for node u
• distance (edge attribute key, optional (default=None)) – Use the specified edge attribute as
the edge distance in shortest path calculations
• wf_improved (bool, optional (default=True)) – If True, scale by the fraction of nodes reach-
able. This gives the Wasserman and Faust improved formula. For single component graphs
it is the same as the original formula.
• reverse (bool, optional (default=False)) – If True and G is a digraph, reverse the edges of
G, using successors instead of predecessors.
Returns nodes – Dictionary of nodes with closeness centrality as the value.
Return type dictionary
See also:
betweenness_centrality(), load_centrality(), eigenvector_centrality(),
degree_centrality()

Notes

The closeness centrality is normalized to (n-1)/(|G|-1) where n is the number of nodes in the connected
part of graph containing the node. If the graph is not completely connected, this algorithm computes the close-
ness centrality for each connected part separately scaled by that parts size.
If the ‘distance’ keyword is set to an edge attribute key then the shortest-path length will be computed using
Dijkstra’s algorithm with that edge attribute as the edge weight.

References

3.6.4 Current Flow Closeness

2 pg. 201 of Wasserman, S. and Faust, K., Social Network Analysis: Methods and Applications, 1994, Cambridge University Press.

190 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

current_flow_closeness_centrality(G[, Compute current-flow closeness centrality for nodes.


. . . ])
information_centrality(G[, weight, dtype, Compute current-flow closeness centrality for nodes.
. . . ])

networkx.algorithms.centrality.current_flow_closeness_centrality

current_flow_closeness_centrality(G, weight=None, dtype=<class ’float’>, solver=’lu’)


Compute current-flow closeness centrality for nodes.
Current-flow closeness centrality is variant of closeness centrality based on effective resistance between nodes
in a network. This metric is also known as information centrality.
Parameters
• G (graph) – A NetworkX graph.
• weight (None or string, optional (default=None)) – If None, all edge weights are considered
equal. Otherwise holds the name of the edge attribute used as weight.
• dtype (data type (default=float)) – Default data type for internal matrices. Set to np.float32
for lower memory consumption.
• solver (string (default=’lu’)) – Type of linear solver to use for computing the flow matrix.
Options are “full” (uses most memory), “lu” (recommended), and “cg” (uses least memory).
Returns nodes – Dictionary of nodes with current flow closeness centrality as the value.
Return type dictionary
See also:
closeness_centrality()

Notes

The algorithm is from Brandes1 .


See also2 for the original definition of information centrality.

References

networkx.algorithms.centrality.information_centrality

information_centrality(G, weight=None, dtype=<class ’float’>, solver=’lu’)


Compute current-flow closeness centrality for nodes.
Current-flow closeness centrality is variant of closeness centrality based on effective resistance between nodes
in a network. This metric is also known as information centrality.
Parameters
• G (graph) – A NetworkX graph.
1 Ulrik Brandes and Daniel Fleischer, Centrality Measures Based on Current Flow. Proc. 22nd Symp. Theoretical Aspects of Computer Science

(STACS ‘05). LNCS 3404, pp. 533-544. Springer-Verlag, 2005. http://algo.uni-konstanz.de/publications/bf-cmbcf-05.pdf


2 Karen Stephenson and Marvin Zelen: Rethinking centrality: Methods and examples. Social Networks 11(1):1-37, 1989. https://doi.org/10.

1016/0378-8733(89)90016-6

3.6. Centrality 191


NetworkX Reference, Release 2.3rc1.dev20190222214247

• weight (None or string, optional (default=None)) – If None, all edge weights are considered
equal. Otherwise holds the name of the edge attribute used as weight.
• dtype (data type (default=float)) – Default data type for internal matrices. Set to np.float32
for lower memory consumption.
• solver (string (default=’lu’)) – Type of linear solver to use for computing the flow matrix.
Options are “full” (uses most memory), “lu” (recommended), and “cg” (uses least memory).
Returns nodes – Dictionary of nodes with current flow closeness centrality as the value.
Return type dictionary
See also:
closeness_centrality()

Notes

The algorithm is from Brandes1 .


See also2 for the original definition of information centrality.

References

3.6.5 (Shortest Path) Betweenness

betweenness_centrality(G[, k, normalized, Compute the shortest-path betweenness centrality for


. . . ]) nodes.
edge_betweenness_centrality(G[, k, . . . ]) Compute betweenness centrality for edges.
betweenness_centrality_subset(G, sources, Compute betweenness centrality for a subset of nodes.
...)
edge_betweenness_centrality_subset(G, Compute betweenness centrality for edges for a subset
. . . [, . . . ]) of nodes.

networkx.algorithms.centrality.betweenness_centrality

betweenness_centrality(G, k=None, normalized=True, weight=None, endpoints=False, seed=None)


Compute the shortest-path betweenness centrality for nodes.
Betweenness centrality of a node 𝑣 is the sum of the fraction of all-pairs shortest paths that pass through 𝑣
∑︁ 𝜎(𝑠, 𝑡|𝑣)
𝑐𝐵 (𝑣) =
𝜎(𝑠, 𝑡)
𝑠,𝑡∈𝑉

where 𝑉 is the set of nodes, 𝜎(𝑠, 𝑡) is the number of shortest (𝑠, 𝑡)-paths, and 𝜎(𝑠, 𝑡|𝑣) is the number of those
paths passing through some node 𝑣 other than 𝑠, 𝑡. If 𝑠 = 𝑡, 𝜎(𝑠, 𝑡) = 1, and if 𝑣 ∈ 𝑠, 𝑡, 𝜎(𝑠, 𝑡|𝑣) = 02 .
Parameters
• G (graph) – A NetworkX graph.
1 Ulrik Brandes and Daniel Fleischer, Centrality Measures Based on Current Flow. Proc. 22nd Symp. Theoretical Aspects of Computer Science

(STACS ‘05). LNCS 3404, pp. 533-544. Springer-Verlag, 2005. http://algo.uni-konstanz.de/publications/bf-cmbcf-05.pdf


2 Karen Stephenson and Marvin Zelen: Rethinking centrality: Methods and examples. Social Networks 11(1):1-37, 1989. https://doi.org/10.

1016/0378-8733(89)90016-6
2 Ulrik Brandes: On Variants of Shortest-Path Betweenness Centrality and their Generic Computation. Social Networks 30(2):136-145, 2008.

http://www.inf.uni-konstanz.de/algo/publications/b-vspbc-08.pdf

192 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• k (int, optional (default=None)) – If k is not None use k node samples to estimate between-
ness. The value of k <= n where n is the number of nodes in the graph. Higher values give
better approximation.
• normalized (bool, optional) – If True the betweenness values are normalized by 2/
((n-1)(n-2)) for graphs, and 1/((n-1)(n-2)) for directed graphs where n is the
number of nodes in G.
• weight (None or string, optional (default=None)) – If None, all edge weights are considered
equal. Otherwise holds the name of the edge attribute used as weight.
• endpoints (bool, optional) – If True include the endpoints in the shortest path counts.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness. Note that this is only used if k is not None.
Returns nodes – Dictionary of nodes with betweenness centrality as the value.
Return type dictionary
See also:
edge_betweenness_centrality(), load_centrality()

Notes

The algorithm is from Ulrik Brandes1 . See4 for the original first published version and2 for details on algorithms
for variations and related metrics.
For approximate betweenness calculations set k=#samples to use k nodes (“pivots”) to estimate the betweenness
values. For an estimate of the number of pivots needed see3 .
For weighted graphs the edge weights must be greater than zero. Zero edge weights can produce an infinite
number of equal length paths between pairs of nodes.

References

networkx.algorithms.centrality.edge_betweenness_centrality

edge_betweenness_centrality(G, k=None, normalized=True, weight=None, seed=None)


Compute betweenness centrality for edges.
Betweenness centrality of an edge 𝑒 is the sum of the fraction of all-pairs shortest paths that pass through 𝑒
∑︁ 𝜎(𝑠, 𝑡|𝑒)
𝑐𝐵 (𝑒) =
𝜎(𝑠, 𝑡)
𝑠,𝑡∈𝑉

where 𝑉 is the set of nodes, 𝜎(𝑠, 𝑡) is the number of shortest (𝑠, 𝑡)-paths, and 𝜎(𝑠, 𝑡|𝑒) is the number of those
paths passing through edge 𝑒2 .
Parameters
• G (graph) – A NetworkX graph.
1 Ulrik Brandes: A Faster Algorithm for Betweenness Centrality. Journal of Mathematical Sociology 25(2):163-177, 2001. http://www.inf.

uni-konstanz.de/algo/publications/b-fabc-01.pdf
4 Linton C. Freeman: A set of measures of centrality based on betweenness. Sociometry 40: 35–41, 1977 http://moreno.ss.uci.edu/23.pdf
3 Ulrik Brandes and Christian Pich: Centrality Estimation in Large Networks. International Journal of Bifurcation and Chaos 17(7):2303-2318,

2007. http://www.inf.uni-konstanz.de/algo/publications/bp-celn-06.pdf
2 Ulrik Brandes: On Variants of Shortest-Path Betweenness Centrality and their Generic Computation. Social Networks 30(2):136-145, 2008.

http://www.inf.uni-konstanz.de/algo/publications/b-vspbc-08.pdf

3.6. Centrality 193


NetworkX Reference, Release 2.3rc1.dev20190222214247

• k (int, optional (default=None)) – If k is not None use k node samples to estimate between-
ness. The value of k <= n where n is the number of nodes in the graph. Higher values give
better approximation.
• normalized (bool, optional) – If True the betweenness values are normalized by 2/(𝑛(𝑛 −
1)) for graphs, and 1/(𝑛(𝑛 − 1)) for directed graphs where 𝑛 is the number of nodes in G.
• weight (None or string, optional (default=None)) – If None, all edge weights are considered
equal. Otherwise holds the name of the edge attribute used as weight.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness. Note that this is only used if k is not None.
Returns edges – Dictionary of edges with betweenness centrality as the value.
Return type dictionary
See also:
betweenness_centrality(), edge_load()

Notes

The algorithm is from Ulrik Brandes1 .


For weighted graphs the edge weights must be greater than zero. Zero edge weights can produce an infinite
number of equal length paths between pairs of nodes.

References

networkx.algorithms.centrality.betweenness_centrality_subset

betweenness_centrality_subset(G, sources, targets, normalized=False, weight=None)


Compute betweenness centrality for a subset of nodes.
∑︁ 𝜎(𝑠, 𝑡|𝑣)
𝑐𝐵 (𝑣) =
𝜎(𝑠, 𝑡)
𝑠∈𝑆,𝑡∈𝑇

where 𝑆 is the set of sources, 𝑇 is the set of targets, 𝜎(𝑠, 𝑡) is the number of shortest (𝑠, 𝑡)-paths, and 𝜎(𝑠, 𝑡|𝑣)
is the number of those paths passing through some node 𝑣 other than 𝑠, 𝑡. If 𝑠 = 𝑡, 𝜎(𝑠, 𝑡) = 1, and if 𝑣 ∈ 𝑠, 𝑡,
𝜎(𝑠, 𝑡|𝑣) = 02 .
Parameters
• G (graph) – A NetworkX graph.
• sources (list of nodes) – Nodes to use as sources for shortest paths in betweenness
• targets (list of nodes) – Nodes to use as targets for shortest paths in betweenness
• normalized (bool, optional) – If True the betweenness values are normalized by 2/((𝑛 −
1)(𝑛 − 2)) for graphs, and 1/((𝑛 − 1)(𝑛 − 2)) for directed graphs where 𝑛 is the number of
nodes in G.
• weight (None or string, optional (default=None)) – If None, all edge weights are considered
equal. Otherwise holds the name of the edge attribute used as weight.
1 A Faster Algorithm for Betweenness Centrality. Ulrik Brandes, Journal of Mathematical Sociology 25(2):163-177, 2001. http://www.inf.

uni-konstanz.de/algo/publications/b-fabc-01.pdf
2 Ulrik Brandes: On Variants of Shortest-Path Betweenness Centrality and their Generic Computation. Social Networks 30(2):136-145, 2008.

http://www.inf.uni-konstanz.de/algo/publications/b-vspbc-08.pdf

194 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns nodes – Dictionary of nodes with betweenness centrality as the value.


Return type dictionary
See also:
edge_betweenness_centrality(), load_centrality()

Notes

The basic algorithm is from1 .


For weighted graphs the edge weights must be greater than zero. Zero edge weights can produce an infinite
number of equal length paths between pairs of nodes.
The normalization might seem a little strange but it is the same as in between-
ness_centrality() and is designed to make betweenness_centrality(G) be the same as between-
ness_centrality_subset(G,sources=G.nodes(),targets=G.nodes()).

References

networkx.algorithms.centrality.edge_betweenness_centrality_subset

edge_betweenness_centrality_subset(G, sources, targets, normalized=False, weight=None)


Compute betweenness centrality for edges for a subset of nodes.
∑︁ 𝜎(𝑠, 𝑡|𝑒)
𝑐𝐵 (𝑣) =
𝜎(𝑠, 𝑡)
𝑠∈𝑆,𝑡∈𝑇

where 𝑆 is the set of sources, 𝑇 is the set of targets, 𝜎(𝑠, 𝑡) is the number of shortest (𝑠, 𝑡)-paths, and 𝜎(𝑠, 𝑡|𝑒)
is the number of those paths passing through edge 𝑒2 .
Parameters
• G (graph) – A networkx graph.
• sources (list of nodes) – Nodes to use as sources for shortest paths in betweenness
• targets (list of nodes) – Nodes to use as targets for shortest paths in betweenness
• normalized (bool, optional) – If True the betweenness values are normalized by 2/
(n(n-1)) for graphs, and 1/(n(n-1)) for directed graphs where n is the number of
nodes in G.
• weight (None or string, optional (default=None)) – If None, all edge weights are considered
equal. Otherwise holds the name of the edge attribute used as weight.
Returns edges – Dictionary of edges with Betweenness centrality as the value.
Return type dictionary
See also:
betweenness_centrality(), edge_load()
1 Ulrik Brandes, A Faster Algorithm for Betweenness Centrality. Journal of Mathematical Sociology 25(2):163-177, 2001. http://www.inf.

uni-konstanz.de/algo/publications/b-fabc-01.pdf
2 Ulrik Brandes: On Variants of Shortest-Path Betweenness Centrality and their Generic Computation. Social Networks 30(2):136-145, 2008.

http://www.inf.uni-konstanz.de/algo/publications/b-vspbc-08.pdf

3.6. Centrality 195


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The basic algorithm is from1 .


For weighted graphs the edge weights must be greater than zero. Zero edge weights can produce an infinite
number of equal length paths between pairs of nodes.
The normalization might seem a little strange but it is the same as in
edge_betweenness_centrality() and is designed to make edge_betweenness_centrality(G) be the same as
edge_betweenness_centrality_subset(G,sources=G.nodes(),targets=G.nodes()).

References

3.6.6 Current Flow Betweenness

current_flow_betweenness_centrality(G[, Compute current-flow betweenness centrality for nodes.


. . . ])
Compute current-flow betweenness centrality for edges.
edge_current_flow_betweenness_centrality(G)
Compute the approximate current-flow betweenness
approximate_current_flow_betweenness_centrality(G)
centrality for nodes.
Compute current-flow betweenness centrality for sub-
current_flow_betweenness_centrality_subset(G,
...) sets of nodes.
Compute current-flow betweenness centrality for edges
edge_current_flow_betweenness_centrality_subset(G,
...) using subsets of nodes.

networkx.algorithms.centrality.current_flow_betweenness_centrality

current_flow_betweenness_centrality(G, normalized=True, weight=None, dtype=<class


’float’>, solver=’full’)
Compute current-flow betweenness centrality for nodes.
Current-flow betweenness centrality uses an electrical current model for information spreading in contrast to
betweenness centrality which uses shortest paths.
Current-flow betweenness centrality is also known as random-walk betweenness centrality2 .
Parameters
• G (graph) – A NetworkX graph
• normalized (bool, optional (default=True)) – If True the betweenness values are normalized
by 2/[(n-1)(n-2)] where n is the number of nodes in G.
• weight (string or None, optional (default=None)) – Key for edge data used as the edge
weight. If None, then use 1 as each edge weight.
• dtype (data type (float)) – Default data type for internal matrices. Set to np.float32 for lower
memory consumption.
• solver (string (default=’lu’)) – Type of linear solver to use for computing the flow matrix.
Options are “full” (uses most memory), “lu” (recommended), and “cg” (uses least memory).
Returns nodes – Dictionary of nodes with betweenness centrality as the value.
Return type dictionary
1 Ulrik Brandes, A Faster Algorithm for Betweenness Centrality. Journal of Mathematical Sociology 25(2):163-177, 2001. http://www.inf.

uni-konstanz.de/algo/publications/b-fabc-01.pdf
2 A measure of betweenness centrality based on random walks, M. E. J. Newman, Social Networks 27, 39-54 (2005).

196 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
approximate_current_flow_betweenness_centrality(), betweenness_centrality(),
edge_betweenness_centrality(), edge_current_flow_betweenness_centrality()

Notes

Current-flow betweenness can be computed in 𝑂(𝐼(𝑛 − 1) + 𝑚𝑛 log 𝑛) time1 , where 𝐼(𝑛 − 1) is the time
3
√ the inverse Laplacian. For a full matrix this is 𝑂(𝑛 ) but using sparse methods you can
needed to compute
achieve 𝑂(𝑛𝑚 𝑘) where 𝑘 is the Laplacian matrix condition number.
The space required is 𝑂(𝑛𝑤) where 𝑤 is the width of the sparse Laplacian matrix. Worse case is 𝑤 = 𝑛 for
𝑂(𝑛2 ).
If the edges have a ‘weight’ attribute they will be used as weights in this algorithm. Unspecified weights are set
to 1.

References

networkx.algorithms.centrality.edge_current_flow_betweenness_centrality

edge_current_flow_betweenness_centrality(G, normalized=True, weight=None,


dtype=<class ’float’>, solver=’full’)
Compute current-flow betweenness centrality for edges.
Current-flow betweenness centrality uses an electrical current model for information spreading in contrast to
betweenness centrality which uses shortest paths.
Current-flow betweenness centrality is also known as random-walk betweenness centrality2 .
Parameters
• G (graph) – A NetworkX graph
• normalized (bool, optional (default=True)) – If True the betweenness values are normalized
by 2/[(n-1)(n-2)] where n is the number of nodes in G.
• weight (string or None, optional (default=None)) – Key for edge data used as the edge
weight. If None, then use 1 as each edge weight.
• dtype (data type (default=float)) – Default data type for internal matrices. Set to np.float32
for lower memory consumption.
• solver (string (default=’lu’)) – Type of linear solver to use for computing the flow matrix.
Options are “full” (uses most memory), “lu” (recommended), and “cg” (uses least memory).
Returns nodes – Dictionary of edge tuples with betweenness centrality as the value.
Return type dictionary
Raises NetworkXError – The algorithm does not support DiGraphs. If the input graph is an
instance of DiGraph class, NetworkXError is raised.
See also:
betweenness_centrality(), edge_betweenness_centrality(),
current_flow_betweenness_centrality()
1 Centrality Measures Based on Current Flow. Ulrik Brandes and Daniel Fleischer, Proc. 22nd Symp. Theoretical Aspects of Computer Science

(STACS ‘05). LNCS 3404, pp. 533-544. Springer-Verlag, 2005. http://algo.uni-konstanz.de/publications/bf-cmbcf-05.pdf


2 A measure of betweenness centrality based on random walks, M. E. J. Newman, Social Networks 27, 39-54 (2005).

3.6. Centrality 197


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Current-flow betweenness can be computed in 𝑂(𝐼(𝑛 − 1) + 𝑚𝑛 log 𝑛) time1 , where 𝐼(𝑛 − 1) is the time
3
√ the inverse Laplacian. For a full matrix this is 𝑂(𝑛 ) but using sparse methods you can
needed to compute
achieve 𝑂(𝑛𝑚 𝑘) where 𝑘 is the Laplacian matrix condition number.
The space required is 𝑂(𝑛𝑤) where 𝑤 is the width of the sparse Laplacian matrix. Worse case is 𝑤 = 𝑛 for
𝑂(𝑛2 ).
If the edges have a ‘weight’ attribute they will be used as weights in this algorithm. Unspecified weights are set
to 1.

References

networkx.algorithms.centrality.approximate_current_flow_betweenness_centrality

approximate_current_flow_betweenness_centrality(G, normalized=True, weight=None,


dtype=<class ’float’>, solver=’full’,
epsilon=0.5, kmax=10000,
seed=None)
Compute the approximate current-flow betweenness centrality for nodes.
Approximates the current-flow betweenness centrality within absolute error of epsilon with high probability1 .
Parameters
• G (graph) – A NetworkX graph
• normalized (bool, optional (default=True)) – If True the betweenness values are normalized
by 2/[(n-1)(n-2)] where n is the number of nodes in G.
• weight (string or None, optional (default=None)) – Key for edge data used as the edge
weight. If None, then use 1 as each edge weight.
• dtype (data type (float)) – Default data type for internal matrices. Set to np.float32 for lower
memory consumption.
• solver (string (default=’lu’)) – Type of linear solver to use for computing the flow matrix.
Options are “full” (uses most memory), “lu” (recommended), and “cg” (uses least memory).
• epsilon (float) – Absolute error tolerance.
• kmax (int) – Maximum number of sample node pairs to use for approximation.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns nodes – Dictionary of nodes with betweenness centrality as the value.
Return type dictionary
See also:
current_flow_betweenness_centrality()
1 Centrality Measures Based on Current Flow. Ulrik Brandes and Daniel Fleischer, Proc. 22nd Symp. Theoretical Aspects of Computer Science

(STACS ‘05). LNCS 3404, pp. 533-544. Springer-Verlag, 2005. http://algo.uni-konstanz.de/publications/bf-cmbcf-05.pdf


1 Ulrik Brandes and Daniel Fleischer: Centrality Measures Based on Current Flow. Proc. 22nd Symp. Theoretical Aspects of Computer Science

(STACS ‘05). LNCS 3404, pp. 533-544. Springer-Verlag, 2005. http://algo.uni-konstanz.de/publications/bf-cmbcf-05.pdf

198 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The running time is 𝑂((1/𝜖2 )𝑚 𝑘 log 𝑛) and the space required is 𝑂(𝑚) for 𝑛 nodes and 𝑚 edges.
If the edges have a ‘weight’ attribute they will be used as weights in this algorithm. Unspecified weights are set
to 1.

References

networkx.algorithms.centrality.current_flow_betweenness_centrality_subset

current_flow_betweenness_centrality_subset(G, sources, targets, normalized=True,


weight=None, dtype=<class ’float’>,
solver=’lu’)
Compute current-flow betweenness centrality for subsets of nodes.
Current-flow betweenness centrality uses an electrical current model for information spreading in contrast to
betweenness centrality which uses shortest paths.
Current-flow betweenness centrality is also known as random-walk betweenness centrality2 .
Parameters
• G (graph) – A NetworkX graph
• sources (list of nodes) – Nodes to use as sources for current
• targets (list of nodes) – Nodes to use as sinks for current
• normalized (bool, optional (default=True)) – If True the betweenness values are normalized
by b=b/(n-1)(n-2) where n is the number of nodes in G.
• weight (string or None, optional (default=None)) – Key for edge data used as the edge
weight. If None, then use 1 as each edge weight.
• dtype (data type (float)) – Default data type for internal matrices. Set to np.float32 for lower
memory consumption.
• solver (string (default=’lu’)) – Type of linear solver to use for computing the flow matrix.
Options are “full” (uses most memory), “lu” (recommended), and “cg” (uses least memory).
Returns nodes – Dictionary of nodes with betweenness centrality as the value.
Return type dictionary
See also:
approximate_current_flow_betweenness_centrality(), betweenness_centrality(),
edge_betweenness_centrality(), edge_current_flow_betweenness_centrality()

Notes

Current-flow betweenness can be computed in 𝑂(𝐼(𝑛 − 1) + 𝑚𝑛 log 𝑛) time1 , where 𝐼(𝑛 − 1) is the time
3
√ the inverse Laplacian. For a full matrix this is 𝑂(𝑛 ) but using sparse methods you can
needed to compute
achieve 𝑂(𝑛𝑚 𝑘) where 𝑘 is the Laplacian matrix condition number.
2A measure of betweenness centrality based on random walks, M. E. J. Newman, Social Networks 27, 39-54 (2005).
1Centrality Measures Based on Current Flow. Ulrik Brandes and Daniel Fleischer, Proc. 22nd Symp. Theoretical Aspects of Computer Science
(STACS ‘05). LNCS 3404, pp. 533-544. Springer-Verlag, 2005. http://algo.uni-konstanz.de/publications/bf-cmbcf-05.pdf

3.6. Centrality 199


NetworkX Reference, Release 2.3rc1.dev20190222214247

The space required is 𝑂(𝑛𝑤) where 𝑤 is the width of the sparse Laplacian matrix. Worse case is 𝑤 = 𝑛 for
𝑂(𝑛2 ).
If the edges have a ‘weight’ attribute they will be used as weights in this algorithm. Unspecified weights are set
to 1.

References

networkx.algorithms.centrality.edge_current_flow_betweenness_centrality_subset

edge_current_flow_betweenness_centrality_subset(G, sources, targets, normalized=True,


weight=None, dtype=<class ’float’>,
solver=’lu’)
Compute current-flow betweenness centrality for edges using subsets of nodes.
Current-flow betweenness centrality uses an electrical current model for information spreading in contrast to
betweenness centrality which uses shortest paths.
Current-flow betweenness centrality is also known as random-walk betweenness centrality2 .
Parameters
• G (graph) – A NetworkX graph
• sources (list of nodes) – Nodes to use as sources for current
• targets (list of nodes) – Nodes to use as sinks for current
• normalized (bool, optional (default=True)) – If True the betweenness values are normalized
by b=b/(n-1)(n-2) where n is the number of nodes in G.
• weight (string or None, optional (default=None)) – Key for edge data used as the edge
weight. If None, then use 1 as each edge weight.
• dtype (data type (float)) – Default data type for internal matrices. Set to np.float32 for lower
memory consumption.
• solver (string (default=’lu’)) – Type of linear solver to use for computing the flow matrix.
Options are “full” (uses most memory), “lu” (recommended), and “cg” (uses least memory).
Returns nodes – Dictionary of edge tuples with betweenness centrality as the value.
Return type dict
See also:
betweenness_centrality(), edge_betweenness_centrality(),
current_flow_betweenness_centrality()

Notes

Current-flow betweenness can be computed in 𝑂(𝐼(𝑛 − 1) + 𝑚𝑛 log 𝑛) time1 , where 𝐼(𝑛 − 1) is the time
3
√ the inverse Laplacian. For a full matrix this is 𝑂(𝑛 ) but using sparse methods you can
needed to compute
achieve 𝑂(𝑛𝑚 𝑘) where 𝑘 is the Laplacian matrix condition number.
The space required is 𝑂(𝑛𝑤) where 𝑤 is the width of the sparse Laplacian matrix. Worse case is 𝑤 = 𝑛 for
𝑂(𝑛2 ).
2A measure of betweenness centrality based on random walks, M. E. J. Newman, Social Networks 27, 39-54 (2005).
1Centrality Measures Based on Current Flow. Ulrik Brandes and Daniel Fleischer, Proc. 22nd Symp. Theoretical Aspects of Computer Science
(STACS ‘05). LNCS 3404, pp. 533-544. Springer-Verlag, 2005. http://algo.uni-konstanz.de/publications/bf-cmbcf-05.pdf

200 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

If the edges have a ‘weight’ attribute they will be used as weights in this algorithm. Unspecified weights are set
to 1.

References

3.6.7 Communicability Betweenness

Returns subgraph communicability for all pairs of nodes


communicability_betweenness_centrality(G[,
. . . ]) in G.

networkx.algorithms.centrality.communicability_betweenness_centrality

communicability_betweenness_centrality(G, normalized=True)
Returns subgraph communicability for all pairs of nodes in G.
Communicability betweenness measure makes use of the number of walks connecting every pair of nodes as the
basis of a betweenness centrality measure.
Parameters G (graph)
Returns nodes – Dictionary of nodes with communicability betweenness as the value.
Return type dictionary
Raises NetworkXError – If the graph is not undirected and simple.

Notes

Let G=(V,E) be a simple undirected graph with n nodes and m edges, and A denote the adjacency matrix of G.
Let G(r)=(V,E(r)) be the graph resulting from removing all edges connected to node r but not the node
itself.
The adjacency matrix for G(r) is A+E(r), where E(r) has nonzeros only in row and column r.
The subraph betweenness of a node r is1
1 ∑︁ ∑︁ 𝐺𝑝𝑟𝑞
𝜔𝑟 = , 𝑝 ̸= 𝑞, 𝑞 ̸= 𝑟,
𝐶 𝑝 𝑞 𝐺𝑝𝑞

where G_{prq}=(e^{A}_{pq} - (e^{A+E(r)})_{pq} is the number of walks involving node r,


G_{pq}=(e^{A})_{pq} is the number of closed walks starting at node p and ending at node q, and
C=(n-1)^{2}-(n-1) is a normalization factor equal to the number of terms in the sum.
The resulting omega_{r} takes values between zero and one. The lower bound cannot be attained for a
connected graph, and the upper bound is attained in the star graph.

References

Examples

1 Ernesto Estrada, Desmond J. Higham, Naomichi Hatano, “Communicability Betweenness in Complex Networks” Physica A 388 (2009) 764-

774. https://arxiv.org/abs/0905.4102

3.6. Centrality 201


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.Graph([(0,1),(1,2),(1,5),(5,4),(2,4),(2,3),(4,3),(3,6)])
>>> cbc = nx.communicability_betweenness_centrality(G)

3.6.8 Load

load_centrality(G[, v, cutoff, normalized, . . . ]) Compute load centrality for nodes.


edge_load_centrality(G[, cutoff]) Compute edge load.

networkx.algorithms.centrality.load_centrality

load_centrality(G, v=None, cutoff=None, normalized=True, weight=None)


Compute load centrality for nodes.
The load centrality of a node is the fraction of all shortest paths that pass through that node.
Parameters
• G (graph) – A networkx graph.
• normalized (bool, optional (default=True)) – If True the betweenness values are normalized
by b=b/(n-1)(n-2) where n is the number of nodes in G.
• weight (None or string, optional (default=None)) – If None, edge weights are ignored.
Otherwise holds the name of the edge attribute used as weight.
• cutoff (bool, optional (default=None)) – If specified, only consider paths of length <= cut-
off.
Returns nodes – Dictionary of nodes with centrality as the value.
Return type dictionary
See also:
betweenness_centrality()

Notes

Load centrality is slightly different than betweenness. It was originally introduced by2 . For this load algorithm
see1 .

References

networkx.algorithms.centrality.edge_load_centrality

edge_load_centrality(G, cutoff=False)
Compute edge load.
WARNING: This concept of edge load has not been analysed or discussed outside of NetworkX that we know
of. It is based loosely on load_centrality in the sense that it counts the number of shortest paths which cross
each edge. This function is for demonstration and testing purposes.
2 Kwang-Il Goh, Byungnam Kahng and Doochul Kim Universal behavior of Load Distribution in Scale-Free Networks. Physical Review Letters

87(27):1–4, 2001. http://phya.snu.ac.kr/~dkim/PRL87278701.pdf


1 Mark E. J. Newman: Scientific collaboration networks. II. Shortest paths, weighted networks, and centrality. Physical Review E 64, 016132,

2001. http://journals.aps.org/pre/abstract/10.1103/PhysRevE.64.016132

202 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (graph) – A networkx graph
• cutoff (bool, optional (default=False)) – If specified, only consider paths of length <= cut-
off.
Returns
• A dict keyed by edge 2-tuple to the number of shortest paths
• which use that edge. Where more than one path is shortest
• the count is divided equally among paths.

3.6.9 Subgraph

subgraph_centrality(G) Returns subgraph centrality for each node in G.


subgraph_centrality_exp(G) Returns the subgraph centrality for each node of G.
estrada_index(G) Returns the Estrada index of a the graph G.

networkx.algorithms.centrality.subgraph_centrality

subgraph_centrality(G)
Returns subgraph centrality for each node in G.
Subgraph centrality of a node n is the sum of weighted closed walks of all lengths starting and ending at node
n. The weights decrease with path length. Each closed walk is associated with a connected subgraph (1 ).
Parameters G (graph)
Returns nodes – Dictionary of nodes with subgraph centrality as the value.
Return type dictionary
Raises NetworkXError – If the graph is not undirected and simple.
See also:

subgraph_centrality_exp() Alternative algorithm of the subgraph centrality for each node of G.

Notes

This version of the algorithm computes eigenvalues and eigenvectors of the adjacency matrix.
Subgraph centrality of a node u in G can be found using a spectral decomposition of the adjacency matrix1 ,
𝑁
∑︁
𝑆𝐶(𝑢) = (𝑣𝑗𝑢 )2 𝑒𝜆𝑗 ,
𝑗=1

where v_j is an eigenvector of the adjacency matrix A of G corresponding corresponding to the eigenvalue
lambda_j.
1 Ernesto Estrada, Juan A. Rodriguez-Velazquez, “Subgraph centrality in complex networks”, Physical Review E 71, 056103 (2005). https:

//arxiv.org/abs/cond-mat/0504730

3.6. Centrality 203


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

(Example from1 ) >>> G = nx.Graph([(1,2),(1,5),(1,8),(2,3),(2,8),(3,4),(3,6),(4,5),(4,7),(5,6),(6,7),(7,8)]) >>> sc


= nx.subgraph_centrality(G) >>> print([‘%s %0.2f’%(node,sc[node]) for node in sorted(sc)]) [‘1 3.90’, ‘2 3.90’,
‘3 3.64’, ‘4 3.71’, ‘5 3.64’, ‘6 3.71’, ‘7 3.64’, ‘8 3.90’]

References

networkx.algorithms.centrality.subgraph_centrality_exp

subgraph_centrality_exp(G)
Returns the subgraph centrality for each node of G.
Subgraph centrality of a node n is the sum of weighted closed walks of all lengths starting and ending at node
n. The weights decrease with path length. Each closed walk is associated with a connected subgraph (1 ).
Parameters G (graph)
Returns nodes – Dictionary of nodes with subgraph centrality as the value.
Return type dictionary
Raises NetworkXError – If the graph is not undirected and simple.
See also:

subgraph_centrality() Alternative algorithm of the subgraph centrality for each node of G.

Notes

This version of the algorithm exponentiates the adjacency matrix.


The subgraph centrality of a node u in G can be found using the matrix exponential of the adjacency matrix of
G1 ,

𝑆𝐶(𝑢) = (𝑒𝐴 )𝑢𝑢 .

References

Examples

(Example from1 ) >>> G = nx.Graph([(1,2),(1,5),(1,8),(2,3),(2,8),(3,4),(3,6),(4,5),(4,7),(5,6),(6,7),(7,8)]) >>> sc


= nx.subgraph_centrality_exp(G) >>> print([‘%s %0.2f’%(node,sc[node]) for node in sorted(sc)]) [‘1 3.90’, ‘2
3.90’, ‘3 3.64’, ‘4 3.71’, ‘5 3.64’, ‘6 3.71’, ‘7 3.64’, ‘8 3.90’]

networkx.algorithms.centrality.estrada_index

estrada_index(G)
Returns the Estrada index of a the graph G.
The Estrada Index is a topological index of folding or 3D “compactness” (1 ).
1 Ernesto Estrada, Juan A. Rodriguez-Velazquez, “Subgraph centrality in complex networks”, Physical Review E 71, 056103 (2005). https:

//arxiv.org/abs/cond-mat/0504730
1 E. Estrada, “Characterization of 3D molecular structure”, Chem. Phys. Lett. 319, 713 (2000). https://doi.org/10.1016/S0009-2614(00)00158-5

204 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters G (graph)
Returns estrada index
Return type float
Raises NetworkXError – If the graph is not undirected and simple.

Notes

Let G=(V,E) be a simple undirected graph with n nodes and let


lambda_{1}leqlambda_{2}leqcdotslambda_{n} be a non-increasing ordering of the eigen-
values of its adjacency matrix A. The Estrada index is (1 ,2 )
𝑛
∑︁
𝐸𝐸(𝐺) = 𝑒𝜆𝑗 .
𝑗=1

References

Examples

>>> G=nx.Graph([(0,1),(1,2),(1,5),(5,4),(2,4),(2,3),(4,3),(3,6)])
>>> ei=nx.estrada_index(G)

3.6.10 Harmonic Centrality

harmonic_centrality(G[, nbunch, distance]) Compute harmonic centrality for nodes.

networkx.algorithms.centrality.harmonic_centrality

harmonic_centrality(G, nbunch=None, distance=None)


Compute harmonic centrality for nodes.
Harmonic centrality1 of a node u is the sum of the reciprocal of the shortest path distances from all other nodes
to u
∑︁ 1
𝐶(𝑢) =
𝑑(𝑣, 𝑢)
𝑣̸=𝑢

where d(v, u) is the shortest-path distance between v and u.


Notice that higher values indicate higher centrality.
Parameters
• G (graph) – A NetworkX graph
• nbunch (container) – Container of nodes. If provided harmonic centrality will be computed
only over the nodes in nbunch.
2 José Antonio de la Peñaa, Ivan Gutman, Juan Rada, “Estimating the Estrada index”, Linear Algebra and its Applications. 427, 1 (2007).

https://doi.org/10.1016/j.laa.2007.06.020
1 Boldi, Paolo, and Sebastiano Vigna. “Axioms for centrality.” Internet Mathematics 10.3-4 (2014): 222-262.

3.6. Centrality 205


NetworkX Reference, Release 2.3rc1.dev20190222214247

• distance (edge attribute key, optional (default=None)) – Use the specified edge attribute as
the edge distance in shortest path calculations. If None, then each edge will have distance
equal to 1.
Returns nodes – Dictionary of nodes with harmonic centrality as the value.
Return type dictionary
See also:
betweenness_centrality(), load_centrality(), eigenvector_centrality(),
degree_centrality(), closeness_centrality()

Notes

If the ‘distance’ keyword is set to an edge attribute key then the shortest-path length will be computed using
Dijkstra’s algorithm with that edge attribute as the edge weight.

References

3.6.11 Reaching

local_reaching_centrality(G, v[, paths, Returns the local reaching centrality of a node in a di-
. . . ]) rected graph.
global_reaching_centrality(G[, weight, Returns the global reaching centrality of a directed
. . . ]) graph.

networkx.algorithms.centrality.local_reaching_centrality

local_reaching_centrality(G, v, paths=None, weight=None, normalized=True)


Returns the local reaching centrality of a node in a directed graph.
The local reaching centrality of a node in a directed graph is the proportion of other nodes reachable from that
node1 .
Parameters
• G (DiGraph) – A NetworkX DiGraph.
• v (node) – A node in the directed graph G.
• paths (dictionary (default=None)) – If this is not None it must be a dictionary rep-
resentation of single-source shortest paths, as computed by, for example, networkx.
shortest_path() with source node v. Use this keyword argument if you intend to
invoke this function many times but don’t want the paths to be recomputed each time.
• weight (None or string, optional (default=None)) – Attribute to use for edge weights. If
None, each edge weight is assumed to be one. A higher weight implies a stronger connec-
tion between nodes and a shorter path length.
• normalized (bool, optional (default=True)) – Whether to normalize the edge weights by
the total sum of edge weights.
Returns h – The local reaching centrality of the node v in the graph G.
1 Mones, Enys, Lilla Vicsek, and Tamás Vicsek. “Hierarchy Measure for Complex Networks.” PLoS ONE 7.3 (2012): e33799. https://doi.org/
10.1371/journal.pone.0033799

206 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type float

Examples

>>> import networkx as nx


>>> G = nx.DiGraph()
>>> G.add_edges_from([(1, 2), (1, 3)])
>>> nx.local_reaching_centrality(G, 3)
0.0
>>> G.add_edge(3, 2)
>>> nx.local_reaching_centrality(G, 3)
0.5

See also:
global_reaching_centrality()

References

networkx.algorithms.centrality.global_reaching_centrality

global_reaching_centrality(G, weight=None, normalized=True)


Returns the global reaching centrality of a directed graph.
The global reaching centrality of a weighted directed graph is the average over all nodes of the difference
between the local reaching centrality of the node and the greatest local reaching centrality of any node in the
graph1 . For more information on the local reaching centrality, see local_reaching_centrality().
Informally, the local reaching centrality is the proportion of the graph that is reachable from the neighbors of
the node.
Parameters
• G (DiGraph) – A networkx DiGraph.
• weight (None or string, optional (default=None)) – Attribute to use for edge weights. If
None, each edge weight is assumed to be one. A higher weight implies a stronger connec-
tion between nodes and a shorter path length.
• normalized (bool, optional (default=True)) – Whether to normalize the edge weights by
the total sum of edge weights.
Returns h – The global reaching centrality of the graph.
Return type float

Examples

>>> import networkx as nx


>>> G = nx.DiGraph()
>>> G.add_edge(1, 2)
>>> G.add_edge(1, 3)
>>> nx.global_reaching_centrality(G)
1.0
(continues on next page)
1 Mones, Enys, Lilla Vicsek, and Tamás Vicsek. “Hierarchy Measure for Complex Networks.” PLoS ONE 7.3 (2012): e33799. https://doi.org/

10.1371/journal.pone.0033799

3.6. Centrality 207


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> G.add_edge(3, 2)
>>> nx.global_reaching_centrality(G)
0.75

See also:
local_reaching_centrality()

References

3.6.12 Percolation

percolation_centrality(G[, attribute, . . . ]) Compute the percolation centrality for nodes.

networkx.algorithms.centrality.percolation_centrality

percolation_centrality(G, attribute=’percolation’, states=None, weight=None)


Compute the percolation centrality for nodes.
Percolation centrality of a node 𝑣, at a given time, is defined as the proportion of ‘percolated paths’ that go
through that node.
This measure quantifies relative impact of nodes based on their topological connectivity, as well as their perco-
lation states.
Percolation states of nodes are used to depict network percolation scenarios (such as during infection transmis-
sion in a social network of individuals, spreading of computer viruses on computer networks, or transmission
of disease over a network of towns) over time. In this measure usually the percolation state is expressed as a
decimal between 0.0 and 1.0.
When all nodes are in the same percolated state this measure is equivalent to betweenness centrality.
Parameters
• G (graph) – A NetworkX graph.
• attribute (None or string, optional (default=’percolation’)) – Name of the node attribute to
use for percolation state, used if states is None.
• states (None or dict, optional (default=None)) – Specify percolation states for the nodes,
nodes as keys states as values.
• weight (None or string, optional (default=None)) – If None, all edge weights are considered
equal. Otherwise holds the name of the edge attribute used as weight.
Returns nodes – Dictionary of nodes with percolation centrality as the value.
Return type dictionary
See also:
betweenness_centrality()

208 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The algorithm is from Mahendra Piraveenan, Mikhail Prokopenko, and Liaquat Hossain1 Pair dependecies are
calculated and accumulated using2
For weighted graphs the edge weights must be greater than zero. Zero edge weights can produce an infinite
number of equal length paths between pairs of nodes.

References

3.6.13 Second Order Centrality

second_order_centrality(G) Compute the second order centrality for nodes of G.

networkx.algorithms.centrality.second_order_centrality

second_order_centrality(G)
Compute the second order centrality for nodes of G.
The second order centrality of a given node is the standard deviation of the return times to that node of a
perpetual random walk on G:
Parameters G (graph) – A NetworkX connected and undirected graph.
Returns nodes – Dictionary keyed by node with second order centrality as the value.
Return type dictionary

Examples

>>> G = nx.star_graph(10)
>>> soc = nx.second_order_centrality(G)
>>> print(sorted(soc.items(), key=lambda x:x[1])[0][0]) # pick first id
0

Raises NetworkXException – If the graph G is empty, non connected or has negative weights.

See also:
betweenness_centrality()

Notes

Lower values of second order centrality indicate higher centrality.


The algorithm is from Kermarrec, Le Merrer, Sericola and Trédan1 .
1 Mahendra Piraveenan, Mikhail Prokopenko, Liaquat Hossain Percolation Centrality: Quantifying Graph-Theoretic Impact of Nodes during

Percolation in Networks http://journals.plos.org/plosone/article?id=10.1371/journal.pone.0053095


2 Ulrik Brandes: A Faster Algorithm for Betweenness Centrality. Journal of Mathematical Sociology 25(2):163-177, 2001. http://www.inf.

uni-konstanz.de/algo/publications/b-fabc-01.pdf
1 Anne-Marie Kermarrec, Erwan Le Merrer, Bruno Sericola, Gilles Trédan “Second order centrality: Distributed assessment of nodes criticity

in complex networks”, Elsevier Computer Communications 34(5):619-628, 2011.

3.6. Centrality 209


NetworkX Reference, Release 2.3rc1.dev20190222214247

This code implements the analytical version of the algorithm, i.e., there is no simulation of a random walk
process involved. The random walk is here unbiased (corresponding to eq 6 of the paper1 ), thus the centrality
values are the standard deviations for random walk return times on the transformed input graph G (equal in-
degree at each nodes by adding self-loops).
Complexity of this implementation, made to run locally on a single machine, is O(n^3), with n the size of G,
which makes it viable only for small graphs.

References

3.7 Chains

Functions for finding chains in a graph.

chain_decomposition(G[, root]) Returns the chain decomposition of a graph.

3.7.1 networkx.algorithms.chains.chain_decomposition

chain_decomposition(G, root=None)
Returns the chain decomposition of a graph.
The chain decomposition of a graph with respect a depth-first search tree is a set of cycles or paths derived from
the set of fundamental cycles of the tree in the following manner. Consider each fundamental cycle with respect
to the given tree, represented as a list of edges beginning with the nontree edge oriented away from the root of
the tree. For each fundamental cycle, if it overlaps with any previous fundamental cycle, just take the initial
non-overlapping segment, which is a path instead of a cycle. Each cycle or path is called a chain. For more
information, see1 .
Parameters
• G (undirected graph)
• root (node (optional)) – A node in the graph G. If specified, only the chain decomposition
for the connected component containing this node will be returned. This node indicates the
root of the depth-first search tree.
Yields chain (list) – A list of edges representing a chain. There is no guarantee on the orientation
of the edges in each chain (for example, if a chain includes the edge joining nodes 1 and 2, the
chain may include either (1, 2) or (2, 1)).
Raises NodeNotFound – If root is not in the graph G.

Notes

The worst-case running time of this implementation is linear in the number of nodes and number of edges1 .
1 Jens M. Schmidt (2013). “A simple test on 2-vertex- and 2-edge-connectivity.” Information Processing Letters, 113, 241–244. Elsevier.

<https://doi.org/10.1016/j.ipl.2013.01.016>

210 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.8 Chordal

Algorithms for chordal graphs.


A graph is chordal if every cycle of length at least 4 has a chord (an edge joining two nodes not adjacent in the cycle).
https://en.wikipedia.org/wiki/Chordal_graph

is_chordal(G) Checks whether G is a chordal graph.


chordal_graph_cliques(G) Returns the set of maximal cliques of a chordal graph.
chordal_graph_treewidth(G) Returns the treewidth of the chordal graph G.
find_induced_nodes(G, s, t[, treewidth_bound]) Returns the set of induced nodes in the path from s to t.

3.8.1 networkx.algorithms.chordal.is_chordal

is_chordal(G)
Checks whether G is a chordal graph.
A graph is chordal if every cycle of length at least 4 has a chord (an edge joining two nodes not adjacent in the
cycle).
Parameters G (graph) – A NetworkX graph.
Returns chordal – True if G is a chordal graph and False otherwise.
Return type bool
Raises NetworkXError – The algorithm does not support DiGraph, MultiGraph and MultiDi-
Graph. If the input graph is an instance of one of these classes, a NetworkXError is raised.

Examples

>>> import networkx as nx


>>> e=[(1,2),(1,3),(2,3),(2,4),(3,4),(3,5),(3,6),(4,5),(4,6),(5,6)]
>>> G=nx.Graph(e)
>>> nx.is_chordal(G)
True

Notes

The routine tries to go through every node following maximum cardinality search. It returns False when it finds
that the separator for any node is not a clique. Based on the algorithms in1 .

References

3.8.2 networkx.algorithms.chordal.chordal_graph_cliques

chordal_graph_cliques(G)
Returns the set of maximal cliques of a chordal graph.
1 R. E. Tarjan and M. Yannakakis, Simple linear-time algorithms to test chordality of graphs, test acyclicity of hypergraphs, and selectively
reduce acyclic hypergraphs, SIAM J. Comput., 13 (1984), pp. 566–579.

3.8. Chordal 211


NetworkX Reference, Release 2.3rc1.dev20190222214247

The algorithm breaks the graph in connected components and performs a maximum cardinality search in each
component to get the cliques.
Parameters G (graph) – A NetworkX graph
Returns cliques
Return type A set containing the maximal cliques in G.
Raises NetworkXError – The algorithm does not support DiGraph, MultiGraph and MultiDi-
Graph. If the input graph is an instance of one of these classes, a NetworkXError is raised.
The algorithm can only be applied to chordal graphs. If the input graph is found to be non-
chordal, a NetworkXError is raised.

Examples

>>> import networkx as nx


>>> e= [(1,2),(1,3),(2,3),(2,4),(3,4),(3,5),(3,6),(4,5),(4,6),(5,6),(7,8)]
>>> G = nx.Graph(e)
>>> G.add_node(9)
>>> setlist = nx.chordal_graph_cliques(G)

3.8.3 networkx.algorithms.chordal.chordal_graph_treewidth

chordal_graph_treewidth(G)
Returns the treewidth of the chordal graph G.
Parameters G (graph) – A NetworkX graph
Returns treewidth – The size of the largest clique in the graph minus one.
Return type int
Raises NetworkXError – The algorithm does not support DiGraph, MultiGraph and MultiDi-
Graph. If the input graph is an instance of one of these classes, a NetworkXError is raised.
The algorithm can only be applied to chordal graphs. If the input graph is found to be non-
chordal, a NetworkXError is raised.

Examples

>>> import networkx as nx


>>> e = [(1,2),(1,3),(2,3),(2,4),(3,4),(3,5),(3,6),(4,5),(4,6),(5,6),(7,8)]
>>> G = nx.Graph(e)
>>> G.add_node(9)
>>> nx.chordal_graph_treewidth(G)
3

References

3.8.4 networkx.algorithms.chordal.find_induced_nodes

find_induced_nodes(G, s, t, treewidth_bound=9223372036854775807)
Returns the set of induced nodes in the path from s to t.
Parameters

212 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• G (graph) – A chordal NetworkX graph


• s (node) – Source node to look for induced nodes
• t (node) – Destination node to look for induced nodes
• treewith_bound (float) – Maximum treewidth acceptable for the graph H. The search for
induced nodes will end as soon as the treewidth_bound is exceeded.
Returns I – The set of induced nodes in the path from s to t in G
Return type Set of nodes
Raises NetworkXError – The algorithm does not support DiGraph, MultiGraph and MultiDi-
Graph. If the input graph is an instance of one of these classes, a NetworkXError is raised.
The algorithm can only be applied to chordal graphs. If the input graph is found to be non-
chordal, a NetworkXError is raised.

Examples

>>> import networkx as nx


>>> G=nx.Graph()
>>> G = nx.generators.classic.path_graph(10)
>>> I = nx.find_induced_nodes(G,1,9,2)
>>> list(I)
[1, 2, 3, 4, 5, 6, 7, 8, 9]

Notes

G must be a chordal graph and (s,t) an edge that is not in G.


If a treewidth_bound is provided, the search for induced nodes will end as soon as the treewidth_bound is
exceeded.
The algorithm is inspired by Algorithm 4 in1 . A formal definition of induced node can also be found on that
reference.

References

3.9 Clique

Functions for finding and manipulating cliques.


Finding the largest clique in a graph is NP-complete problem, so most of these algorithms have an exponential running
time; for more information, see the Wikipedia article on the clique problem1 .

enumerate_all_cliques(G) Returns all cliques in an undirected graph.


find_cliques(G) Returns all maximal cliques in an undirected graph.
make_max_clique_graph(G[, create_using]) Returns the maximal clique graph of the given graph.
make_clique_bipartite(G[, fpos, . . . ]) Returns the bipartite clique graph corresponding to G.
graph_clique_number(G[, cliques]) Returns the clique number of the graph.
Continued on next page
1 Learning Bounded Treewidth Bayesian Networks. Gal Elidan, Stephen Gould; JMLR, 9(Dec):2699–2731, 2008. http://jmlr.csail.mit.edu/

papers/volume9/elidan08a/elidan08a.pdf
1 clique problem:: https://en.wikipedia.org/wiki/Clique_problem

3.9. Clique 213


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 44 – continued from previous page


graph_number_of_cliques(G[, cliques]) Returns the number of maximal cliques in the graph.
node_clique_number(G[, nodes, cliques]) Returns the size of the largest maximal clique contain-
ing each given node.
number_of_cliques(G[, nodes, cliques]) Returns the number of maximal cliques for each node.
cliques_containing_node(G[, nodes, cliques]) Returns a list of cliques containing the given node.

3.9.1 networkx.algorithms.clique.enumerate_all_cliques

enumerate_all_cliques(G)
Returns all cliques in an undirected graph.
This function returns an iterator over cliques, each of which is a list of nodes. The iteration is ordered by
cardinality of the cliques: first all cliques of size one, then all cliques of size two, etc.
Parameters G (NetworkX graph) – An undirected graph.
Returns An iterator over cliques, each of which is a list of nodes in G. The cliques are ordered
according to size.
Return type iterator

Notes

To obtain a list of all cliques, use list(enumerate_all_cliques(G)). However, be aware that in the
worst-case, the length of this list can be exponential in the number of nodes in the graph (for example, when
the graph is the complete graph). This function avoids storing all cliques in memory by only keeping current
candidate node lists in memory during its search.
The implementation is adapted from the algorithm by Zhang, et al. (2005)1 to output all cliques discovered.
This algorithm ignores self-loops and parallel edges, since cliques are not conventionally defined with such
edges.

References

3.9.2 networkx.algorithms.clique.find_cliques

find_cliques(G)
Returns all maximal cliques in an undirected graph.
For each node v, a maximal clique for v is a largest complete subgraph containing v. The largest maximal clique
is sometimes called the maximum clique.
This function returns an iterator over cliques, each of which is a list of nodes. It is an iterative implementation,
so should not suffer from recursion depth issues.
Parameters G (NetworkX graph) – An undirected graph.
Returns An iterator over maximal cliques, each of which is a list of nodes in G. The order of cliques
is arbitrary.
Return type iterator
1 Yun Zhang, Abu-Khzam, F.N., Baldwin, N.E., Chesler, E.J., Langston, M.A., Samatova, N.F., “Genome-Scale Computational Approaches to

Memory-Intensive Applications in Systems Biology”. Supercomputing, 2005. Proceedings of the ACM/IEEE SC 2005 Conference, pp. 12, 12–18
Nov. 2005. <https://doi.org/10.1109/SC.2005.29>.

214 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:

find_cliques_recursive() A recursive version of the same algorithm.

Notes

To obtain a list of all maximal cliques, use list(find_cliques(G)). However, be aware that in the worst-
case, the length of this list can be exponential in the number of nodes in the graph (for example, when the graph
is the complete graph). This function avoids storing all cliques in memory by only keeping current candidate
node lists in memory during its search.
This implementation is based on the algorithm published by Bron and Kerbosch (1973)1 , as adapted by Tomita,
Tanaka and Takahashi (2006)2 and discussed in Cazals and Karande (2008)3 . It essentially unrolls the re-
cursion used in the references to avoid issues of recursion stack depth (for a recursive implementation, see
find_cliques_recursive()).
This algorithm ignores self-loops and parallel edges, since cliques are not conventionally defined with such
edges.

References

3.9.3 networkx.algorithms.clique.make_max_clique_graph

make_max_clique_graph(G, create_using=None)
Returns the maximal clique graph of the given graph.
The nodes of the maximal clique graph of G are the cliques of G and an edge joins two cliques if the cliques are
not disjoint.
Parameters
• G (NetworkX graph)
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
Returns A graph whose nodes are the cliques of G and whose edges join two cliques if they are not
disjoint.
Return type NetworkX graph

Notes

This function behaves like the following code:

import networkx as nx
G = nx.make_clique_bipartite(G)
cliques = [v for v in G.nodes() if G.nodes[v]['bipartite'] == 0]
G = nx.bipartite.project(G, cliques)
G = nx.relabel_nodes(G, {-v: v - 1 for v in G})

1 Bron, C. and Kerbosch, J. “Algorithm 457: finding all cliques of an undirected graph”. Communications of the ACM 16, 9 (Sep. 1973),

575–577. <http://portal.acm.org/citation.cfm?doid=362342.362367>
2 Etsuji Tomita, Akira Tanaka, Haruhisa Takahashi, “The worst-case time complexity for generating all maximal cliques and computational

experiments”, Theoretical Computer Science, Volume 363, Issue 1, Computing and Combinatorics, 10th Annual International Conference on
Computing and Combinatorics (COCOON 2004), 25 October 2006, Pages 28–42 <https://doi.org/10.1016/j.tcs.2006.06.015>
3 F. Cazals, C. Karande, “A note on the problem of reporting maximal cliques”, Theoretical Computer Science, Volume 407, Issues 1–3, 6

November 2008, Pages 564–568, <https://doi.org/10.1016/j.tcs.2008.05.010>

3.9. Clique 215


NetworkX Reference, Release 2.3rc1.dev20190222214247

It should be faster, though, since it skips all the intermediate steps.

3.9.4 networkx.algorithms.clique.make_clique_bipartite

make_clique_bipartite(G, fpos=None, create_using=None, name=None)


Returns the bipartite clique graph corresponding to G.
In the returned bipartite graph, the “bottom” nodes are the nodes of G and the “top” nodes represent the maximal
cliques of G. There is an edge from node v to clique C in the returned graph if and only if v is an element of C.
Parameters
• G (NetworkX graph) – An undirected graph.
• fpos (bool) – If True or not None, the returned graph will have an additional attribute, pos,
a dictionary mapping node to position in the Euclidean plane.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
Returns
A bipartite graph whose “bottom” set is the nodes of the graph G, whose “top” set is the cliques
of G, and whose edges join nodes of G to the cliques that contain them.
The nodes of the graph G have the node attribute ‘bipartite’ set to 1 and the nodes representing
cliques have the node attribute ‘bipartite’ set to 0, as is the convention for bipartite graphs in
NetworkX.
Return type NetworkX graph

3.9.5 networkx.algorithms.clique.graph_clique_number

graph_clique_number(G, cliques=None)
Returns the clique number of the graph.
The clique number of a graph is the size of the largest clique in the graph.
Parameters
• G (NetworkX graph) – An undirected graph.
• cliques (list) – A list of cliques, each of which is itself a list of nodes. If not specified, the
list of all cliques will be computed, as by find_cliques().
Returns The size of the largest clique in G.
Return type int

Notes

You should provide cliques if you have already computed the list of maximal cliques, in order to avoid an
exponential time search for maximal cliques.

216 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.9.6 networkx.algorithms.clique.graph_number_of_cliques

graph_number_of_cliques(G, cliques=None)
Returns the number of maximal cliques in the graph.
Parameters
• G (NetworkX graph) – An undirected graph.
• cliques (list) – A list of cliques, each of which is itself a list of nodes. If not specified, the
list of all cliques will be computed, as by find_cliques().
Returns The number of maximal cliques in G.
Return type int

Notes

You should provide cliques if you have already computed the list of maximal cliques, in order to avoid an
exponential time search for maximal cliques.

3.9.7 networkx.algorithms.clique.node_clique_number

node_clique_number(G, nodes=None, cliques=None)


Returns the size of the largest maximal clique containing each given node.
Returns a single or list depending on input nodes. Optional list of cliques can be input if already computed.

3.9.8 networkx.algorithms.clique.number_of_cliques

number_of_cliques(G, nodes=None, cliques=None)


Returns the number of maximal cliques for each node.
Returns a single or list depending on input nodes. Optional list of cliques can be input if already computed.

3.9.9 networkx.algorithms.clique.cliques_containing_node

cliques_containing_node(G, nodes=None, cliques=None)


Returns a list of cliques containing the given node.
Returns a single list or list of lists depending on input nodes. Optional list of cliques can be input if already
computed.

3.10 Clustering

Algorithms to characterize the number of triangles in a graph.

triangles(G[, nodes]) Compute the number of triangles.


transitivity(G) Compute graph transitivity, the fraction of all possible
triangles present in G.
clustering(G[, nodes, weight]) Compute the clustering coefficient for nodes.
Continued on next page

3.10. Clustering 217


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 45 – continued from previous page


average_clustering(G[, nodes, weight, . . . ]) Compute the average clustering coefficient for the graph
G.
square_clustering(G[, nodes]) Compute the squares clustering coefficient for nodes.
generalized_degree(G[, nodes]) Compute the generalized degree for nodes.

3.10.1 networkx.algorithms.cluster.triangles

triangles(G, nodes=None)
Compute the number of triangles.
Finds the number of triangles that include a node as one vertex.
Parameters
• G (graph) – A networkx graph
• nodes (container of nodes, optional (default= all nodes in G)) – Compute triangles for nodes
in this container.
Returns out – Number of triangles keyed by node label.
Return type dictionary

Examples

>>> G=nx.complete_graph(5)
>>> print(nx.triangles(G,0))
6
>>> print(nx.triangles(G))
{0: 6, 1: 6, 2: 6, 3: 6, 4: 6}
>>> print(list(nx.triangles(G,(0,1)).values()))
[6, 6]

Notes

When computing triangles for the entire graph each triangle is counted three times, once at each node. Self
loops are ignored.

3.10.2 networkx.algorithms.cluster.transitivity

transitivity(G)
Compute graph transitivity, the fraction of all possible triangles present in G.
Possible triangles are identified by the number of “triads” (two edges with a shared vertex).
The transitivity is

#𝑡𝑟𝑖𝑎𝑛𝑔𝑙𝑒𝑠
𝑇 =3 .
#𝑡𝑟𝑖𝑎𝑑𝑠
Parameters G (graph)
Returns out – Transitivity
Return type float

218 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.complete_graph(5)
>>> print(nx.transitivity(G))
1.0

3.10.3 networkx.algorithms.cluster.clustering

clustering(G, nodes=None, weight=None)


Compute the clustering coefficient for nodes.
For unweighted graphs, the clustering of a node 𝑢 is the fraction of possible triangles through that node that
exist,
2𝑇 (𝑢)
𝑐𝑢 = ,
𝑑𝑒𝑔(𝑢)(𝑑𝑒𝑔(𝑢) − 1)

where 𝑇 (𝑢) is the number of triangles through node 𝑢 and 𝑑𝑒𝑔(𝑢) is the degree of 𝑢.
For weighted graphs, there are several ways to define clustering1 . the one used here is defined as the geometric
average of the subgraph edge weights2 ,
1 ∑︁
𝑐𝑢 = (𝑤
ˆ𝑢𝑣 𝑤 ˆ𝑣𝑤 )1/3 .
ˆ𝑢𝑤 𝑤
𝑑𝑒𝑔(𝑢)(𝑑𝑒𝑔(𝑢) − 1)) 𝑣𝑤

The edge weights 𝑤


ˆ𝑢𝑣 are normalized by the maximum weight in the network 𝑤
ˆ𝑢𝑣 = 𝑤𝑢𝑣 / max(𝑤).
The value of 𝑐𝑢 is assigned to 0 if 𝑑𝑒𝑔(𝑢) < 2.
For directed graphs, the clustering is similarly defined as the fraction of all possible directed triangles or geo-
metric average of the subgraph edge weights for unweighted and weighted directed graph respectively3 .
1
𝑐𝑢 = 𝑇 (𝑢),
𝑑𝑒𝑔 𝑡𝑜𝑡 (𝑢)(𝑑𝑒𝑔 𝑡𝑜𝑡 (𝑢) − 1) − 2𝑑𝑒𝑔 ↔ (𝑢)

where 𝑇 (𝑢) is the number of directed triangles through node 𝑢, 𝑑𝑒𝑔 𝑡𝑜𝑡 (𝑢) is the sum of in degree and out degree
of 𝑢 and 𝑑𝑒𝑔 ↔ (𝑢) is the reciprocal degree of 𝑢.
Parameters
• G (graph)
• nodes (container of nodes, optional (default=all nodes in G)) – Compute clustering for
nodes in this container.
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used as a weight. If None, then each edge has weight 1.
Returns out – Clustering coefficient at specified nodes
Return type float, or dictionary
1 Generalizations of the clustering coefficient to weighted complex networks by J. Saramäki, M. Kivelä, J.-P. Onnela, K. Kaski, and J. Kertész,
Physical Review E, 75 027105 (2007). http://jponnela.com/web_documents/a9.pdf
2 Intensity and coherence of motifs in weighted complex networks by J. P. Onnela, J. Saramäki, J. Kertész, and K. Kaski, Physical Review E,

71(6), 065103 (2005).


3 Clustering in complex directed networks by G. Fagiolo, Physical Review E, 76(2), 026107 (2007).

3.10. Clustering 219


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G=nx.complete_graph(5)
>>> print(nx.clustering(G,0))
1.0
>>> print(nx.clustering(G))
{0: 1.0, 1: 1.0, 2: 1.0, 3: 1.0, 4: 1.0}

Notes

Self loops are ignored.

References

3.10.4 networkx.algorithms.cluster.average_clustering

average_clustering(G, nodes=None, weight=None, count_zeros=True)


Compute the average clustering coefficient for the graph G.
The clustering coefficient for the graph is the average,
1 ∑︁
𝐶= 𝑐𝑣 ,
𝑛
𝑣∈𝐺

where 𝑛 is the number of nodes in G.


Parameters
• G (graph)
• nodes (container of nodes, optional (default=all nodes in G)) – Compute average clustering
for nodes in this container.
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used as a weight. If None, then each edge has weight 1.
• count_zeros (bool) – If False include only the nodes with nonzero clustering in the average.
Returns avg – Average clustering
Return type float

Examples

>>> G=nx.complete_graph(5)
>>> print(nx.average_clustering(G))
1.0

Notes

This is a space saving routine; it might be faster to use the clustering function to get a list and then take the
average.
Self loops are ignored.

220 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.10.5 networkx.algorithms.cluster.square_clustering

square_clustering(G, nodes=None)
Compute the squares clustering coefficient for nodes.
For each node return the fraction of possible squares that exist at the node1
∑︀𝑘𝑣 ∑︀𝑘𝑣
𝑤=𝑢+1 𝑞𝑣 (𝑢, 𝑤)
𝐶4 (𝑣) = ∑︀𝑘𝑣 ∑︀𝑢=1
𝑘𝑣
,
𝑢=1 𝑤=𝑢+1 [𝑎𝑣 (𝑢, 𝑤) + 𝑞𝑣 (𝑢, 𝑤)]

where 𝑞𝑣 (𝑢, 𝑤) are the number of common neighbors of 𝑢 and 𝑤 other than 𝑣 (ie squares), and 𝑎𝑣 (𝑢, 𝑤) =
(𝑘𝑢 − (1 + 𝑞𝑣 (𝑢, 𝑤) + 𝜃𝑢𝑣 ))(𝑘𝑤 − (1 + 𝑞𝑣 (𝑢, 𝑤) + 𝜃𝑢𝑤 )), where 𝜃𝑢𝑤 = 1 if 𝑢 and 𝑤 are connected and 0
otherwise.
Parameters
• G (graph)
• nodes (container of nodes, optional (default=all nodes in G)) – Compute clustering for
nodes in this container.
Returns c4 – A dictionary keyed by node with the square clustering coefficient value.
Return type dictionary

Examples

>>> G=nx.complete_graph(5)
>>> print(nx.square_clustering(G,0))
1.0
>>> print(nx.square_clustering(G))
{0: 1.0, 1: 1.0, 2: 1.0, 3: 1.0, 4: 1.0}

Notes

While 𝐶3 (𝑣) (triangle clustering) gives the probability that two neighbors of node v are connected with each
other, 𝐶4 (𝑣) is the probability that two neighbors of node v share a common neighbor different from v. This
algorithm can be applied to both bipartite and unipartite networks.

References

3.10.6 networkx.algorithms.cluster.generalized_degree

generalized_degree(G, nodes=None)
Compute the generalized degree for nodes.
For each node, the generalized degree shows how many edges of given triangle multiplicity the node is connected
to. The triangle multiplicity of an edge is the number of triangles an edge participates in. The generalized degree
(0) (𝑁 −2) (𝑗)
of node 𝑖 can be written as a vector k𝑖 = (𝑘𝑖 , . . . , 𝑘𝑖 ) where 𝑘𝑖 is the number of edges attached to node
𝑖 that participate in 𝑗 triangles.
1 Pedro G. Lind, Marta C. González, and Hans J. Herrmann. 2005 Cycles and clustering in bipartite networks. Physical Review E (72) 056127.

3.10. Clustering 221


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (graph)
• nodes (container of nodes, optional (default=all nodes in G)) – Compute the generalized
degree for nodes in this container.
Returns out – Generalized degree of specified nodes. The Counter is keyed by edge triangle multi-
plicity.
Return type Counter, or dictionary of Counters

Examples

>>> G=nx.complete_graph(5)
>>> print(nx.generalized_degree(G,0))
Counter({3: 4})
>>> print(nx.generalized_degree(G))
{0: Counter({3: 4}), 1: Counter({3: 4}), 2: Counter({3: 4}), 3: Counter({3: 4}),
˓→4: Counter({3: 4})}

To recover the number of triangles attached to a node:

>>> k1 = nx.generalized_degree(G,0)
>>> sum([k*v for k,v in k1.items()])/2 == nx.triangles(G,0)
True

Notes

In a network of N nodes, the highest triangle multiplicty an edge can have is N-2.
The return value does not include a zero entry if no edges of a particular triangle multiplicity are present.
The number of triangles node 𝑖 is attached to can be recovered from the generalized degree k𝑖 =
(0) (𝑁 −2) (1) (2) (𝑁 −2)
(𝑘𝑖 , . . . , 𝑘𝑖 ) by (𝑘𝑖 + 2𝑘𝑖 + . . . + (𝑁 − 2)𝑘𝑖 )/2.

References

3.11 Coloring

greedy_color(G[, strategy, interchange]) Color a graph using various strategies of greedy graph
coloring.

3.11.1 networkx.algorithms.coloring.greedy_color

greedy_color(G, strategy=’largest_first’, interchange=False)


Color a graph using various strategies of greedy graph coloring.
Attempts to color a graph using as few colors as possible, where no neighbours of a node can have same color
as the node itself. The given strategy determines the order in which nodes are colored.
The strategies are described in1 , and smallest-last is based on2 .
1 Adrian Kosowski, and Krzysztof Manuszewski, Classical Coloring of Graphs, Graph Colorings, 2-19, 2004. ISBN 0-8218-3458-4.
2 David W. Matula, and Leland L. Beck, “Smallest-last ordering and clustering and graph coloring algorithms.” J. ACM 30, 3 (July 1983),

222 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (NetworkX graph)
• strategy (string or function(G, colors)) – A function (or a string representing a function) that
provides the coloring strategy, by returning nodes in the ordering they should be colored. G
is the graph, and colors is a dictionary of the currently assigned colors, keyed by nodes.
The function must return an iterable over all the nodes in G.
If the strategy function is an iterator generator (that is, a function with yield statements),
keep in mind that the colors dictionary will be updated after each yield, since this
function chooses colors greedily.
If strategy is a string, it must be one of the following, each of which represents one of
the built-in strategy functions.
– 'largest_first'
– 'random_sequential'
– 'smallest_last'
– 'independent_set'
– 'connected_sequential_bfs'
– 'connected_sequential_dfs'
– 'connected_sequential' (alias for the previous strategy)
– 'strategy_saturation_largest_first'
– 'DSATUR' (alias for the previous strategy)
• interchange (bool) – Will use the color interchange algorithm described by3 if set to True.
Note that strategy_saturation_largest_first and
strategy_independent_set do not work with interchange. Furthermore, if
you use interchange with your own strategy function, you cannot rely on the values in the
colors argument.
Returns
• A dictionary with keys representing nodes and values representing
• corresponding coloring.

Examples

>>> G = nx.cycle_graph(4)
>>> d = nx.coloring.greedy_color(G, strategy='largest_first')
>>> d in [{0: 0, 1: 1, 2: 0, 3: 1}, {0: 1, 1: 0, 2: 1, 3: 0}]
True

Raises NetworkXPointlessConcept – If strategy is


strategy_saturation_largest_first or strategy_independent_set
and interchange is True.
417–427. <https://doi.org/10.1145/2402.322385>
3 Maciej M. Sysło, Marsingh Deo, Janusz S. Kowalik, Discrete Optimization Algorithms with Pascal Programs, 415-424, 1983. ISBN 0-486-

45353-7.

3.11. Coloring 223


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

Some node ordering strategies are provided for use with greedy_color().

strategy_connected_sequential(G, colors[, Returns an iterable over nodes in G in the order given by


. . . ]) a breadth-first or depth-first traversal.
strategy_connected_sequential_dfs(G, Returns an iterable over nodes in G in the order given by
colors) a depth-first traversal.
strategy_connected_sequential_bfs(G, Returns an iterable over nodes in G in the order given by
colors) a breadth-first traversal.
strategy_independent_set(G, colors) Uses a greedy independent set removal strategy to de-
termine the colors.
strategy_largest_first(G, colors) Returns a list of the nodes of G in decreasing order by
degree.
strategy_random_sequential(G, colors[, Returns a random permutation of the nodes of G as a
seed]) list.
strategy_saturation_largest_first(G, Iterates over all the nodes of G in “saturation order”
colors) (also known as “DSATUR”).
strategy_smallest_last(G, colors) Returns a deque of the nodes of G, “smallest” last.

3.11.2 networkx.algorithms.coloring.strategy_connected_sequential

strategy_connected_sequential(G, colors, traversal=’bfs’)


Returns an iterable over nodes in G in the order given by a breadth-first or depth-first traversal.
traversal must be one of the strings 'dfs' or 'bfs', representing depth-first traversal or breadth-first
traversal, respectively.
The generated sequence has the property that for each node except the first, at least one neighbor appeared
earlier in the sequence.
G is a NetworkX graph. colors is ignored.

3.11.3 networkx.algorithms.coloring.strategy_connected_sequential_dfs

strategy_connected_sequential_dfs(G, colors)
Returns an iterable over nodes in G in the order given by a depth-first traversal.
The generated sequence has the property that for each node except the first, at least one neighbor appeared
earlier in the sequence.
G is a NetworkX graph. colors is ignored.

3.11.4 networkx.algorithms.coloring.strategy_connected_sequential_bfs

strategy_connected_sequential_bfs(G, colors)
Returns an iterable over nodes in G in the order given by a breadth-first traversal.
The generated sequence has the property that for each node except the first, at least one neighbor appeared
earlier in the sequence.
G is a NetworkX graph. colors is ignored.

224 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.11.5 networkx.algorithms.coloring.strategy_independent_set

strategy_independent_set(G, colors)
Uses a greedy independent set removal strategy to determine the colors.
This function updates colors in-place and return None, unlike the other strategy functions in this module.
This algorithm repeatedly finds and removes a maximal independent set, assigning each node in the set an
unused color.
G is a NetworkX graph.
This strategy is related to strategy_smallest_last(): in that strategy, an independent set of size one is
chosen at each step instead of a maximal independent set.

3.11.6 networkx.algorithms.coloring.strategy_largest_first

strategy_largest_first(G, colors)
Returns a list of the nodes of G in decreasing order by degree.
G is a NetworkX graph. colors is ignored.

3.11.7 networkx.algorithms.coloring.strategy_random_sequential

strategy_random_sequential(G, colors, seed=None)


Returns a random permutation of the nodes of G as a list.
G is a NetworkX graph. colors is ignored.
seed [integer, random_state, or None (default)] Indicator of random number generation state. See Randomness.

3.11.8 networkx.algorithms.coloring.strategy_saturation_largest_first

strategy_saturation_largest_first(G, colors)
Iterates over all the nodes of G in “saturation order” (also known as “DSATUR”).
G is a NetworkX graph. colors is a dictionary mapping nodes of G to colors, for those nodes that have already
been colored.

3.11.9 networkx.algorithms.coloring.strategy_smallest_last

strategy_smallest_last(G, colors)
Returns a deque of the nodes of G, “smallest” last.
Specifically, the degrees of each node are tracked in a bucket queue. From this, the node of minimum degree is
repeatedly popped from the graph, updating its neighbors’ degrees.
G is a NetworkX graph. colors is ignored.
This implementation of the strategy runs in 𝑂(𝑛 + 𝑚) time (ignoring polylogarithmic factors), where 𝑛 is the
number of nodes and 𝑚 is the number of edges.
This strategy is related to strategy_independent_set(): if we interpret each node removed as an
independent set of size one, then this strategy chooses an independent set of size one instead of a maximal
independent set.

3.11. Coloring 225


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.12 Communicability

Communicability.

communicability(G) Returns communicability between all pairs of nodes in


G.
communicability_exp(G) Returns communicability between all pairs of nodes in
G.

3.12.1 networkx.algorithms.communicability_alg.communicability

communicability(G)
Returns communicability between all pairs of nodes in G.
The communicability between pairs of nodes in G is the sum of closed walks of different lengths starting at node
u and ending at node v.
Parameters G (graph)
Returns comm – Dictionary of dictionaries keyed by nodes with communicability as the value.
Return type dictionary of dictionaries
Raises NetworkXError – If the graph is not undirected and simple.
See also:

communicability_exp() Communicability between all pairs of nodes in G using spectral decomposition.


communicability_betweenness_centrality() Communicability betweeness centrality for each
node in G.

Notes

This algorithm uses a spectral decomposition of the adjacency matrix. Let G=(V,E) be a simple undirected
graph. Using the connection between the powers of the adjacency matrix and the number of walks in the graph,
the communicability between nodes u and v based on the graph spectrum is1
𝑛
∑︁
𝐶(𝑢, 𝑣) = 𝜑𝑗 (𝑢)𝜑𝑗 (𝑣)𝑒𝜆𝑗 ,
𝑗=1

where phi_{j}(u) is the urm{th} element of the jrm{th} orthonormal eigenvector of the adjacency
matrix associated with the eigenvalue lambda_{j}.

References

Examples

>>> G = nx.Graph([(0,1),(1,2),(1,5),(5,4),(2,4),(2,3),(4,3),(3,6)])
>>> c = nx.communicability(G)

1 Ernesto Estrada, Naomichi Hatano, “Communicability in complex networks”, Phys. Rev. E 77, 036111 (2008). https://arxiv.org/abs/0707.0756

226 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.12.2 networkx.algorithms.communicability_alg.communicability_exp

communicability_exp(G)
Returns communicability between all pairs of nodes in G.
Communicability between pair of node (u,v) of node in G is the sum of closed walks of different lengths starting
at node u and ending at node v.
Parameters G (graph)
Returns comm – Dictionary of dictionaries keyed by nodes with communicability as the value.
Return type dictionary of dictionaries
Raises NetworkXError – If the graph is not undirected and simple.
See also:

communicability() Communicability between pairs of nodes in G.


communicability_betweenness_centrality() Communicability betweeness centrality for each
node in G.

Notes

This algorithm uses matrix exponentiation of the adjacency matrix.


Let G=(V,E) be a simple undirected graph. Using the connection between the powers of the adjacency matrix
and the number of walks in the graph, the communicability between nodes u and v is1 ,

𝐶(𝑢, 𝑣) = (𝑒𝐴 )𝑢𝑣 ,

where A is the adjacency matrix of G.

References

Examples

>>> G = nx.Graph([(0,1),(1,2),(1,5),(5,4),(2,4),(2,3),(4,3),(3,6)])
>>> c = nx.communicability_exp(G)

3.13 Communities

Functions for computing and measuring community structure.


The functions in this class are not imported into the top-level networkx namespace. You can access these functions
by importing the networkx.algorithms.community module, then accessing the functions as attributes of
community. For example:

>>> import networkx as nx


>>> from networkx.algorithms import community
>>> G = nx.barbell_graph(5, 1)
>>> communities_generator = community.girvan_newman(G)
(continues on next page)
1 Ernesto Estrada, Naomichi Hatano, “Communicability in complex networks”, Phys. Rev. E 77, 036111 (2008). https://arxiv.org/abs/0707.0756

3.13. Communities 227


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> top_level_communities = next(communities_generator)
>>> next_level_communities = next(communities_generator)
>>> sorted(map(sorted, next_level_communities))
[[0, 1, 2, 3, 4], [5], [6, 7, 8, 9, 10]]

3.13.1 Bipartitions

Functions for computing the Kernighan–Lin bipartition algorithm.

kernighan_lin_bisection(G[, partition, . . . ]) Partition a graph into two blocks using the


Kernighan–Lin algorithm.

networkx.algorithms.community.kernighan_lin.kernighan_lin_bisection

kernighan_lin_bisection(G, partition=None, max_iter=10, weight=’weight’, seed=None)


Partition a graph into two blocks using the Kernighan–Lin algorithm.
This algorithm paritions a network into two sets by iteratively swapping pairs of nodes to reduce the edge cut
between the two sets.
Parameters
• G (graph)
• partition (tuple) – Pair of iterables containing an initial partition. If not specified, a random
balanced partition is used.
• max_iter (int) – Maximum number of times to attempt swaps to find an improvemement
before giving up.
• weight (key) – Edge data key to use as weight. If None, the weights are all set to one.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness. Only used if partition is None
Returns partition – A pair of sets of nodes representing the bipartition.
Return type tuple
Raises NetworkXError – If partition is not a valid partition of the nodes of the graph.

References

3.13.2 Generators

Functions for generating graphs with community structure.

LFR_benchmark_graph(n, tau1, tau2, mu[, . . . ]) Returns the LFR benchmark graph for testing
community-finding algorithms.

228 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.community.community_generators.LFR_benchmark_graph

LFR_benchmark_graph(n, tau1, tau2, mu, average_degree=None, min_degree=None, max_degree=None,


min_community=None, max_community=None, tol=1e-07, max_iters=500,
seed=None)
Returns the LFR benchmark graph for testing community-finding algorithms.
This algorithm proceeds as follows:
1) Find a degree sequence with a power law distribution, and minimum value min_degree, which has
approximate average degree average_degree. This is accomplished by either
a) specifying min_degree and not average_degree,
b) specifying average_degree and not min_degree, in which case a suitable minimum degree
will be found.
max_degree can also be specified, otherwise it will be set to n. Each node u will have mu
mathrm{deg}(u) edges joining it to nodes in communities other than its own and (1 - mu)
mathrm{deg}(u) edges joining it to nodes in its own community.
2) Generate community sizes according to a power law distribution with exponent tau2. If
min_community and max_community are not specified they will be selected to be min_degree
and max_degree, respectively. Community sizes are generated until the sum of their sizes equals n.
3) Each node will be randomly assigned a community with the condition that the community is large enough
for the node’s intra-community degree, (1 - mu) mathrm{deg}(u) as described in step 2. If a
community grows too large, a random node will be selected for reassignment to a new community, until
all nodes have been assigned a community.
4) Each node u then adds (1 - mu) mathrm{deg}(u) intra-community edges and mu
mathrm{deg}(u) inter-community edges.

Parameters
• n (int) – Number of nodes in the created graph.
• tau1 (float) – Power law exponent for the degree distribution of the created graph. This
value must be strictly greater than one.
• tau2 (float) – Power law exponent for the community size distribution in the created graph.
This value must be strictly greater than one.
• mu (float) – Fraction of intra-community edges incident to each node. This value must be
in the interval [0, 1].
• average_degree (float) – Desired average degree of nodes in the created graph. This value
must be in the interval [0, n]. Exactly one of this and min_degree must be specified,
otherwise a NetworkXError is raised.
• min_degree (int) – Minimum degree of nodes in the created graph. This value must be in
the interval [0, n]. Exactly one of this and average_degree must be specified, otherwise
a NetworkXError is raised.
• max_degree (int) – Maximum degree of nodes in the created graph. If not specified, this is
set to n, the total number of nodes in the graph.
• min_community (int) – Minimum size of communities in the graph. If not specified, this is
set to min_degree.
• max_community (int) – Maximum size of communities in the graph. If not specified, this
is set to n, the total number of nodes in the graph.

3.13. Communities 229


NetworkX Reference, Release 2.3rc1.dev20190222214247

• tol (float) – Tolerance when comparing floats, specifically when comparing average degree
values.
• max_iters (int) – Maximum number of iterations to try to create the community sizes, de-
gree distribution, and community affiliations.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns
G – The LFR benchmark graph generated according to the specified parameters.
Each node in the graph has a node attribute 'community' that stores the community (that is,
the set of nodes) that includes it.
Return type NetworkX graph
Raises
• NetworkXError – If any of the parameters do not meet their upper and lower bounds:
– tau1 and tau2 must be strictly greater than 1.
– mu must be in [0, 1].
– max_degree must be in {1, . . . , n}.
– min_community and max_community must be in {0, . . . , n}.
If not exactly one of average_degree and min_degree is specified.
If min_degree is not specified and a suitable min_degree cannot be found.
• ExceededMaxIterations – If a valid degree sequence cannot be created within
max_iters number of iterations.
If a valid set of community sizes cannot be created within max_iters number of iterations.
If a valid community assignment cannot be created within 10 * n * max_iters num-
ber of iterations.

Examples

Basic usage:

>>> from networkx.algorithms.community import LFR_benchmark_graph


>>> n = 250
>>> tau1 = 3
>>> tau2 = 1.5
>>> mu = 0.1
>>> G = LFR_benchmark_graph(n, tau1, tau2, mu, average_degree=5,
... min_community=20, seed=10)

Continuing the example above, you can get the communities from the node attributes of the graph:

>>> communities = {frozenset(G.nodes[v]['community']) for v in G}

Notes

This algorithm differs slightly from the original way it was presented in [1].

230 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

1) Rather than connecting the graph via a configuration model then rewiring to match the intra-community
and inter-community degrees, we do this wiring explicitly at the end, which should be equivalent.
2) The code posted on the author’s website [2] calculates the random power law distributed variables and their
average using continuous approximations, whereas we use the discrete distributions here as both degree
and community size are discrete.
Though the authors describe the algorithm as quite robust, testing during development indicates that a somewhat
narrower parameter set is likely to successfully produce a graph. Some suggestions have been provided in the
event of exceptions.

References

3.13.3 K-Clique

k_clique_communities(G, k[, cliques]) Find k-clique communities in graph using the percola-
tion method.

networkx.algorithms.community.kclique.k_clique_communities

k_clique_communities(G, k, cliques=None)
Find k-clique communities in graph using the percolation method.
A k-clique community is the union of all cliques of size k that can be reached through adjacent (sharing k-1
nodes) k-cliques.
Parameters
• G (NetworkX graph)
• k (int) – Size of smallest clique
• cliques (list or generator) – Precomputed cliques (use networkx.find_cliques(G))
Returns
Return type Yields sets of nodes, one for each k-clique community.

Examples

>>> from networkx.algorithms.community import k_clique_communities


>>> G = nx.complete_graph(5)
>>> K5 = nx.convert_node_labels_to_integers(G,first_label=2)
>>> G.add_edges_from(K5.edges())
>>> c = list(k_clique_communities(G, 4))
>>> list(c[0])
[0, 1, 2, 3, 4, 5, 6]
>>> list(k_clique_communities(G, 6))
[]

References

3.13.4 Modularity-based communities

Functions for detecting communities based on modularity.

3.13. Communities 231


NetworkX Reference, Release 2.3rc1.dev20190222214247

greedy_modularity_communities(G[, Find communities in graph using Clauset-Newman-


weight]) Moore greedy modularity maximization.

networkx.algorithms.community.modularity_max.greedy_modularity_communities

greedy_modularity_communities(G, weight=None)
Find communities in graph using Clauset-Newman-Moore greedy modularity maximization. This method cur-
rently supports the Graph class and does not consider edge weights.
Greedy modularity maximization begins with each node in its own community and joins the pair of communities
that most increases modularity until no such pair exists.
Parameters G (NetworkX graph)
Returns
Return type Yields sets of nodes, one for each community.

Examples

>>> from networkx.algorithms.community import greedy_modularity_communities


>>> G = nx.karate_club_graph()
>>> c = list(greedy_modularity_communities(G))
>>> sorted(c[0])
[8, 14, 15, 18, 20, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33]

References

3.13.5 Label propagation

Label propagation community detection algorithms.

asyn_lpa_communities(G[, weight, seed]) Returns communities in G as detected by asynchronous


label propagation.
label_propagation_communities(G) Generates community sets determined by label propa-
gation

networkx.algorithms.community.label_propagation.asyn_lpa_communities

asyn_lpa_communities(G, weight=None, seed=None)


Returns communities in G as detected by asynchronous label propagation.
The asynchronous label propagation algorithm is described in1 . The algorithm is probabilistic and the found
communities may vary on different executions.
The algorithm proceeds as follows. After initializing each node with a unique label, the algorithm repeatedly sets
the label of a node to be the label that appears most frequently among that nodes neighbors. The algorithm halts
when each node has the label that appears most frequently among its neighbors. The algorithm is asynchronous
because each node is updated without waiting for updates on the remaining nodes.
1Raghavan, Usha Nandini, Réka Albert, and Soundar Kumara. “Near linear time algorithm to detect community structures in large-scale
networks.” Physical Review E 76.3 (2007): 036106.

232 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

This generalized version of the algorithm in1 accepts edge weights.


Parameters
• G (Graph)
• weight (string) – The edge attribute representing the weight of an edge. If None, each
edge is assumed to have weight one. In this algorithm, the weight of an edge is used in
determining the frequency with which a label appears among the neighbors of a node: a
higher weight means the label appears more often.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns communities – Iterable of communities given as sets of nodes.
Return type iterable

Notes

Edge weight attributes must be numerical.

References

networkx.algorithms.community.label_propagation.label_propagation_communities

label_propagation_communities(G)
Generates community sets determined by label propagation
Finds communities in G using a semi-synchronous label propagation method[1]_. This method combines the
advantages of both the synchronous and asynchronous models. Not implemented for directed graphs.
Parameters G (graph) – An undirected NetworkX graph.
Yields communities (generator) – Yields sets of the nodes in each community.
Raises NetworkXNotImplemented – If the graph is directed

References

3.13.6 Fluid Communities

Asynchronous Fluid Communities algorithm for community detection.

asyn_fluidc(G, k[, max_iter, seed]) Returns communities in G as detected by Fluid Commu-


nities algorithm.

networkx.algorithms.community.asyn_fluid.asyn_fluidc

asyn_fluidc(G, k, max_iter=100, seed=None)


Returns communities in G as detected by Fluid Communities algorithm.
The asynchronous fluid communities algorithm is described in1 . The algorithm is based on the simple idea of
1 Parés F., Garcia-Gasulla D. et al. “Fluid Communities: A Competitive and Highly Scalable Community Detection Algorithm”. [https:

//arxiv.org/pdf/1703.09307.pdf].

3.13. Communities 233


NetworkX Reference, Release 2.3rc1.dev20190222214247

fluids interacting in an environment, expanding and pushing each other. It’s initialization is random, so found
communities may vary on different executions.
The algorithm proceeds as follows. First each of the initial k communities is initialized in a random vertex in the
graph. Then the algorithm iterates over all vertices in a random order, updating the community of each vertex
based on its own community and the communities of its neighbours. This process is performed several times
until convergence. At all times, each community has a total density of 1, which is equally distributed among
the vertices it contains. If a vertex changes of community, vertex densities of affected communities are adjusted
immediately. When a complete iteration over all vertices is done, such that no vertex changes the community it
belongs to, the algorithm has converged and returns.
This is the original version of the algorithm described in1 . Unfortunately, it does not support weighted graphs
yet.
Parameters
• G (Graph)
• k (integer) – The number of communities to be found.
• max_iter (integer) – The number of maximum iterations allowed. By default 15.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns communities – Iterable of communities given as sets of nodes.
Return type iterable

Notes

k variable is not an optional argument.

References

3.13.7 Measuring partitions

Functions for measuring the quality of a partition (into communities).

coverage(G, partition) Returns the coverage of a partition.


performance(G, partition) Returns the performance of a partition.

networkx.algorithms.community.quality.coverage

coverage(G, partition)
Returns the coverage of a partition.
The coverage of a partition is the ratio of the number of intra-community edges to the total number of edges in
the graph.
Parameters
• G (NetworkX graph)
• partition (sequence) – Partition of the nodes of G, represented as a sequence of sets of
nodes. Each block of the partition represents a community.
Returns The coverage of the partition, as defined above.

234 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type float


Raises NetworkXError – If partition is not a valid partition of the nodes of G.

Notes

If G is a multigraph, the multiplicity of edges is counted.

References

networkx.algorithms.community.quality.performance

performance(G, partition)
Returns the performance of a partition.
The performance of a partition is the ratio of the number of intra-community edges plus inter-community non-
edges with the total number of potential edges.
Parameters
• G (NetworkX graph) – A simple graph (directed or undirected).
• partition (sequence) – Partition of the nodes of G, represented as a sequence of sets of
nodes. Each block of the partition represents a community.
Returns The performance of the partition, as defined above.
Return type float
Raises NetworkXError – If partition is not a valid partition of the nodes of G.

References

3.13.8 Partitions via centrality measures

Functions for computing communities based on centrality notions.

girvan_newman(G[, most_valuable_edge]) Finds communities in a graph using the Gir-


van–Newman method.

networkx.algorithms.community.centrality.girvan_newman

girvan_newman(G, most_valuable_edge=None)
Finds communities in a graph using the Girvan–Newman method.
Parameters
• G (NetworkX graph)
• most_valuable_edge (function) – Function that takes a graph as input and outputs an edge.
The edge returned by this function will be recomputed and removed at each iteration of the
algorithm.
If not specified, the edge with the highest networkx.
edge_betweenness_centrality() will be used.

3.13. Communities 235


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns Iterator over tuples of sets of nodes in G. Each set of node is a community, each tuple is a
sequence of communities at a particular level of the algorithm.
Return type iterator

Examples

To get the first pair of communities:

>>> G = nx.path_graph(10)
>>> comp = girvan_newman(G)
>>> tuple(sorted(c) for c in next(comp))
([0, 1, 2, 3, 4], [5, 6, 7, 8, 9])

To get only the first k tuples of communities, use itertools.islice():

>>> import itertools


>>> G = nx.path_graph(8)
>>> k = 2
>>> comp = girvan_newman(G)
>>> for communities in itertools.islice(comp, k):
... print(tuple(sorted(c) for c in communities))
...
([0, 1, 2, 3], [4, 5, 6, 7])
([0, 1], [2, 3], [4, 5, 6, 7])

To stop getting tuples of communities once the number of communities is greater than k, use itertools.
takewhile():

>>> import itertools


>>> G = nx.path_graph(8)
>>> k = 4
>>> comp = girvan_newman(G)
>>> limited = itertools.takewhile(lambda c: len(c) <= k, comp)
>>> for communities in limited:
... print(tuple(sorted(c) for c in communities))
...
([0, 1, 2, 3], [4, 5, 6, 7])
([0, 1], [2, 3], [4, 5, 6, 7])
([0, 1], [2, 3], [4, 5], [6, 7])

To just choose an edge to remove based on the weight:

>>> from operator import itemgetter


>>> G = nx.path_graph(10)
>>> edges = G.edges()
>>> nx.set_edge_attributes(G, {(u, v): v for u, v in edges}, 'weight')
>>> def heaviest(G):
... u, v, w = max(G.edges(data='weight'), key=itemgetter(2))
... return (u, v)
...
>>> comp = girvan_newman(G, most_valuable_edge=heaviest)
>>> tuple(sorted(c) for c in next(comp))
([0, 1, 2, 3, 4, 5, 6, 7, 8], [9])

To utilize edge weights when choosing an edge with, for example, the highest betweenness centrality:

236 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> from networkx import edge_betweenness_centrality as betweenness


>>> def most_central_edge(G):
... centrality = betweenness(G, weight='weight')
... return max(centrality, key=centrality.get)
...
>>> G = nx.path_graph(10)
>>> comp = girvan_newman(G, most_valuable_edge=most_central_edge)
>>> tuple(sorted(c) for c in next(comp))
([0, 1, 2, 3, 4], [5, 6, 7, 8, 9])

To specify a different ranking algorithm for edges, use the most_valuable_edge keyword argument:

>>> from networkx import edge_betweenness_centrality


>>> from random import random
>>> def most_central_edge(G):
... centrality = edge_betweenness_centrality(G)
... max_cent = max(centrality.values())
... # Scale the centrality values so they are between 0 and 1,
... # and add some random noise.
... centrality = {e: c / max_cent for e, c in centrality.items()}
... # Add some random noise.
... centrality = {e: c + random() for e, c in centrality.items()}
... return max(centrality, key=centrality.get)
...
>>> G = nx.path_graph(10)
>>> comp = girvan_newman(G, most_valuable_edge=most_central_edge)

Notes

The Girvan–Newman algorithm detects communities by progressively removing edges from the original graph.
The algorithm removes the “most valuable” edge, traditionally the edge with the highest betweenness centrality,
at each step. As the graph breaks down into pieces, the tightly knit community structure is exposed and the
result can be depicted as a dendrogram.

3.13.9 Validating partitions

Helper functions for community-finding algorithms.

is_partition(G, communities) Returns True if and only if communities is a partition


of the nodes of G.

networkx.algorithms.community.community_utils.is_partition

is_partition(G, communities)
Returns True if and only if communities is a partition of the nodes of G.
A partition of a universe set is a family of pairwise disjoint sets whose union is the entire universe set.
G is a NetworkX graph.
communities is an iterable of sets of nodes of G. This iterable will be consumed multiple times during the
execution of this function.

3.13. Communities 237


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.14 Components

3.14.1 Connectivity

is_connected(G) Returns True if the graph is connected, False otherwise.


number_connected_components(G) Returns the number of connected components.
connected_components(G) Generate connected components.
connected_component_subgraphs(G[, copy]) DEPRECATED: Use (G.subgraph(c) for c
in connected_components(G))
node_connected_component(G, n) Returns the set of nodes in the component of graph con-
taining node n.

networkx.algorithms.components.is_connected

is_connected(G)
Returns True if the graph is connected, False otherwise.
Parameters G (NetworkX Graph) – An undirected graph.
Returns connected – True if the graph is connected, false otherwise.
Return type bool
Raises NetworkXNotImplemented: – If G is directed.

Examples

>>> G = nx.path_graph(4)
>>> print(nx.is_connected(G))
True

See also:
is_strongly_connected(), is_weakly_connected(), is_semiconnected(),
is_biconnected(), connected_components()

Notes

For undirected graphs only.

networkx.algorithms.components.number_connected_components

number_connected_components(G)
Returns the number of connected components.
Parameters G (NetworkX graph) – An undirected graph.
Returns n – Number of connected components
Return type integer
See also:
connected_components(), number_weakly_connected_components(),
number_strongly_connected_components()

238 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

For undirected graphs only.

networkx.algorithms.components.connected_components

connected_components(G)
Generate connected components.
Parameters G (NetworkX graph) – An undirected graph
Returns comp – A generator of sets of nodes, one for each component of G.
Return type generator of sets
Raises NetworkXNotImplemented: – If G is directed.

Examples

Generate a sorted list of connected components, largest first.

>>> G = nx.path_graph(4)
>>> nx.add_path(G, [10, 11, 12])
>>> [len(c) for c in sorted(nx.connected_components(G), key=len, reverse=True)]
[4, 3]

If you only want the largest connected component, it’s more efficient to use max instead of sort.

>>> largest_cc = max(nx.connected_components(G), key=len)

See also:
strongly_connected_components(), weakly_connected_components()

Notes

For undirected graphs only.

networkx.algorithms.components.connected_component_subgraphs

connected_component_subgraphs(G, copy=True)
DEPRECATED: Use (G.subgraph(c) for c in connected_components(G))
Or (G.subgraph(c).copy() for c in connected_components(G))

networkx.algorithms.components.node_connected_component

node_connected_component(G, n)
Returns the set of nodes in the component of graph containing node n.
Parameters
• G (NetworkX Graph) – An undirected graph.
• n (node label) – A node in G

3.14. Components 239


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns comp – A set of nodes in the component of G containing node n.


Return type set
Raises NetworkXNotImplemented: – If G is directed.
See also:
connected_components()

Notes

For undirected graphs only.

3.14.2 Strong connectivity

is_strongly_connected(G) Test directed graph for strong connectivity.


number_strongly_connected_components(G) Returns number of strongly connected components in
graph.
strongly_connected_components(G) Generate nodes in strongly connected components of
graph.
DEPRECATED: Use (G.subgraph(c) for c
strongly_connected_component_subgraphs(G[,
copy]) in strongly_connected_components(G))
Generate nodes in strongly connected components of
strongly_connected_components_recursive(G)
graph.
Generate nodes in strongly connected components of
kosaraju_strongly_connected_components(G[,
. . . ]) graph.
condensation(G[, scc]) Returns the condensation of G.

networkx.algorithms.components.is_strongly_connected

is_strongly_connected(G)
Test directed graph for strong connectivity.
A directed graph is strongly connected if and only if every vertex in the graph is reachable from every other
vertex.
Parameters G (NetworkX Graph) – A directed graph.
Returns connected – True if the graph is strongly connected, False otherwise.
Return type bool
Raises NetworkXNotImplemented: – If G is undirected.
See also:
is_weakly_connected(), is_semiconnected(), is_connected(), is_biconnected(),
strongly_connected_components()

Notes

For directed graphs only.

240 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.components.number_strongly_connected_components

number_strongly_connected_components(G)
Returns number of strongly connected components in graph.
Parameters G (NetworkX graph) – A directed graph.
Returns n – Number of strongly connected components
Return type integer
Raises NetworkXNotImplemented: – If G is undirected.
See also:
strongly_connected_components(), number_connected_components(),
number_weakly_connected_components()

Notes

For directed graphs only.

networkx.algorithms.components.strongly_connected_components

strongly_connected_components(G)
Generate nodes in strongly connected components of graph.
Parameters G (NetworkX Graph) – A directed graph.
Returns comp – A generator of sets of nodes, one for each strongly connected component of G.
Return type generator of sets
Raises NetworkXNotImplemented : – If G is undirected.

Examples

Generate a sorted list of strongly connected components, largest first.

>>> G = nx.cycle_graph(4, create_using=nx.DiGraph())


>>> nx.add_cycle(G, [10, 11, 12])
>>> [len(c) for c in sorted(nx.strongly_connected_components(G),
... key=len, reverse=True)]
[4, 3]

If you only want the largest component, it’s more efficient to use max instead of sort.

>>> largest = max(nx.strongly_connected_components(G), key=len)

See also:
connected_components(), weakly_connected_components(),
kosaraju_strongly_connected_components()

Notes

Uses Tarjan’s algorithm[1]_ with Nuutila’s modifications[2]_. Nonrecursive version of algorithm.

3.14. Components 241


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

networkx.algorithms.components.strongly_connected_component_subgraphs

strongly_connected_component_subgraphs(G, copy=True)
DEPRECATED: Use (G.subgraph(c) for c in strongly_connected_components(G))
Or (G.subgraph(c).copy() for c in strongly_connected_components(G))

networkx.algorithms.components.strongly_connected_components_recursive

strongly_connected_components_recursive(G)
Generate nodes in strongly connected components of graph.
Recursive version of algorithm.
Parameters G (NetworkX Graph) – A directed graph.
Returns comp – A generator of sets of nodes, one for each strongly connected component of G.
Return type generator of sets
Raises NetworkXNotImplemented : – If G is undirected.

Examples

Generate a sorted list of strongly connected components, largest first.

>>> G = nx.cycle_graph(4, create_using=nx.DiGraph())


>>> nx.add_cycle(G, [10, 11, 12])
>>> [len(c) for c in sorted(nx.strongly_connected_components_recursive(G),
... key=len, reverse=True)]
[4, 3]

If you only want the largest component, it’s more efficient to use max instead of sort.

>>> largest = max(nx.strongly_connected_components_recursive(G), key=len)

See also:
connected_components()

Notes

Uses Tarjan’s algorithm[1]_ with Nuutila’s modifications[2]_.

References

networkx.algorithms.components.kosaraju_strongly_connected_components

kosaraju_strongly_connected_components(G, source=None)
Generate nodes in strongly connected components of graph.
Parameters G (NetworkX Graph) – A directed graph.
Returns comp – A genrator of sets of nodes, one for each strongly connected component of G.

242 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type generator of sets


Raises NetworkXNotImplemented: – If G is undirected.

Examples

Generate a sorted list of strongly connected components, largest first.

>>> G = nx.cycle_graph(4, create_using=nx.DiGraph())


>>> nx.add_cycle(G, [10, 11, 12])
>>> [len(c) for c in sorted(nx.kosaraju_strongly_connected_components(G),
... key=len, reverse=True)]
[4, 3]

If you only want the largest component, it’s more efficient to use max instead of sort.

>>> largest = max(nx.kosaraju_strongly_connected_components(G), key=len)

See also:
strongly_connected_components()

Notes

Uses Kosaraju’s algorithm.

networkx.algorithms.components.condensation

condensation(G, scc=None)
Returns the condensation of G.
The condensation of G is the graph with each of the strongly connected components contracted into a single
node.
Parameters
• G (NetworkX DiGraph) – A directed graph.
• scc (list or generator (optional, default=None)) – Strongly connected components. If pro-
vided, the elements in scc must partition the nodes in G. If not provided, it will be calculated
as scc=nx.strongly_connected_components(G).
Returns C – The condensation graph C of G. The node labels are integers corresponding to the index
of the component in the list of strongly connected components of G. C has a graph attribute
named ‘mapping’ with a dictionary mapping the original nodes to the nodes in C to which they
belong. Each node in C also has a node attribute ‘members’ with the set of original nodes in G
that form the SCC that the node in C represents.
Return type NetworkX DiGraph
Raises NetworkXNotImplemented: – If G is undirected.

Notes

After contracting all strongly connected components to a single node, the resulting graph is a directed acyclic
graph.

3.14. Components 243


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.14.3 Weak connectivity

is_weakly_connected(G) Test directed graph for weak connectivity.


number_weakly_connected_components(G) Returns the number of weakly connected components in
G.
weakly_connected_components(G) Generate weakly connected components of G.
weakly_connected_component_subgraphs(G[, DEPRECATED: Use (G.subgraph(c) for c
copy]) in weakly_connected_components(G))

networkx.algorithms.components.is_weakly_connected

is_weakly_connected(G)
Test directed graph for weak connectivity.
A directed graph is weakly connected if and only if the graph is connected when the direction of the edge
between nodes is ignored.
Note that if a graph is strongly connected (i.e. the graph is connected even when we account for directionality),
it is by definition weakly connected as well.
Parameters G (NetworkX Graph) – A directed graph.
Returns connected – True if the graph is weakly connected, False otherwise.
Return type bool
Raises NetworkXNotImplemented: – If G is undirected.
See also:
is_strongly_connected(), is_semiconnected(), is_connected(), is_biconnected(),
weakly_connected_components()

Notes

For directed graphs only.

networkx.algorithms.components.number_weakly_connected_components

number_weakly_connected_components(G)
Returns the number of weakly connected components in G.
Parameters G (NetworkX graph) – A directed graph.
Returns n – Number of weakly connected components
Return type integer
Raises NetworkXNotImplemented: – If G is undirected.
See also:
weakly_connected_components(), number_connected_components(),
number_strongly_connected_components()

Notes

For directed graphs only.

244 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.components.weakly_connected_components

weakly_connected_components(G)
Generate weakly connected components of G.
Parameters G (NetworkX graph) – A directed graph
Returns comp – A generator of sets of nodes, one for each weakly connected component of G.
Return type generator of sets
Raises NetworkXNotImplemented: – If G is undirected.

Examples

Generate a sorted list of weakly connected components, largest first.

>>> G = nx.path_graph(4, create_using=nx.DiGraph())


>>> nx.add_path(G, [10, 11, 12])
>>> [len(c) for c in sorted(nx.weakly_connected_components(G),
... key=len, reverse=True)]
[4, 3]

If you only want the largest component, it’s more efficient to use max instead of sort:

>>> largest_cc = max(nx.weakly_connected_components(G), key=len)

See also:
connected_components(), strongly_connected_components()

Notes

For directed graphs only.

networkx.algorithms.components.weakly_connected_component_subgraphs

weakly_connected_component_subgraphs(G, copy=True)
DEPRECATED: Use (G.subgraph(c) for c in weakly_connected_components(G))
Or (G.subgraph(c).copy() for c in weakly_connected_components(G))

3.14.4 Attracting components

is_attracting_component(G) Returns True if G consists of a single attracting compo-


nent.
number_attracting_components(G) Returns the number of attracting components in G.
attracting_components(G) Generates the attracting components in G.
attracting_component_subgraphs(G[, DEPRECATED: Use (G.subgraph(c) for c
copy]) in attracting_components(G))

3.14. Components 245


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.components.is_attracting_component

is_attracting_component(G)
Returns True if G consists of a single attracting component.
Parameters G (DiGraph, MultiDiGraph) – The graph to be analyzed.
Returns attracting – True if G has a single attracting component. Otherwise, False.
Return type bool
Raises NetworkXNotImplemented : – If the input graph is undirected.
See also:
attracting_components(), number_attracting_components()

networkx.algorithms.components.number_attracting_components

number_attracting_components(G)
Returns the number of attracting components in G.
Parameters G (DiGraph, MultiDiGraph) – The graph to be analyzed.
Returns n – The number of attracting components in G.
Return type int
Raises NetworkXNotImplemented : – If the input graph is undirected.
See also:
attracting_components(), is_attracting_component()

networkx.algorithms.components.attracting_components

attracting_components(G)
Generates the attracting components in G.
An attracting component in a directed graph G is a strongly connected component with the property that a
random walker on the graph will never leave the component, once it enters the component.
The nodes in attracting components can also be thought of as recurrent nodes. If a random walker enters the
attractor containing the node, then the node will be visited infinitely often.
Parameters G (DiGraph, MultiDiGraph) – The graph to be analyzed.
Returns attractors – A generator of sets of nodes, one for each attracting component of G.
Return type generator of sets
Raises NetworkXNotImplemented : – If the input graph is undirected.
See also:
number_attracting_components(), is_attracting_component()

246 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.components.attracting_component_subgraphs

attracting_component_subgraphs(G, copy=True)
DEPRECATED: Use (G.subgraph(c) for c in attracting_components(G))
Or (G.subgraph(c).copy() for c in attracting_components(G))

3.14.5 Biconnected components

is_biconnected(G) Returns True if the graph is biconnected, False other-


wise.
biconnected_components(G) Returns a generator of sets of nodes, one set for each
biconnected component of the graph
biconnected_component_edges(G) Returns a generator of lists of edges, one list for each
biconnected component of the input graph.
biconnected_component_subgraphs(G[, DEPRECATED: Use (G.subgraph(c) for c
copy]) in biconnected_components(G))
articulation_points(G) Yield the articulation points, or cut vertices, of a graph.

networkx.algorithms.components.is_biconnected

is_biconnected(G)
Returns True if the graph is biconnected, False otherwise.
A graph is biconnected if, and only if, it cannot be disconnected by removing only one node (and all edges
incident on that node). If removing a node increases the number of disconnected components in the graph, that
node is called an articulation point, or cut vertex. A biconnected graph has no articulation points.
Parameters G (NetworkX Graph) – An undirected graph.
Returns biconnected – True if the graph is biconnected, False otherwise.
Return type bool
Raises NetworkXNotImplemented : – If the input graph is not undirected.

Examples

>>> G = nx.path_graph(4)
>>> print(nx.is_biconnected(G))
False
>>> G.add_edge(0, 3)
>>> print(nx.is_biconnected(G))
True

See also:
biconnected_components(), articulation_points(), biconnected_component_edges(),
is_strongly_connected(), is_weakly_connected(), is_connected(),
is_semiconnected()

3.14. Components 247


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The algorithm to find articulation points and biconnected components is implemented using a non-recursive
depth-first-search (DFS) that keeps track of the highest level that back edges reach in the DFS tree. A node n
is an articulation point if, and only if, there exists a subtree rooted at n such that there is no back edge from
any successor of n that links to a predecessor of n in the DFS tree. By keeping track of all the edges traversed
by the DFS we can obtain the biconnected components because all edges of a bicomponent will be traversed
consecutively between articulation points.

References

networkx.algorithms.components.biconnected_components

biconnected_components(G)
Returns a generator of sets of nodes, one set for each biconnected component of the graph
Biconnected components are maximal subgraphs such that the removal of a node (and all edges incident on that
node) will not disconnect the subgraph. Note that nodes may be part of more than one biconnected component.
Those nodes are articulation points, or cut vertices. The removal of articulation points will increase the number
of connected components of the graph.
Notice that by convention a dyad is considered a biconnected component.
Parameters G (NetworkX Graph) – An undirected graph.
Returns nodes – Generator of sets of nodes, one set for each biconnected component.
Return type generator
Raises NetworkXNotImplemented : – If the input graph is not undirected.
See also:

k_components() this function is a special case where k=2


bridge_components() similar to this function, but is defined using 2-edge-connectivity instead of 2-node-
connectivity.

Examples

>>> G = nx.lollipop_graph(5, 1)
>>> print(nx.is_biconnected(G))
False
>>> bicomponents = list(nx.biconnected_components(G))
>>> len(bicomponents)
2
>>> G.add_edge(0, 5)
>>> print(nx.is_biconnected(G))
True
>>> bicomponents = list(nx.biconnected_components(G))
>>> len(bicomponents)
1

You can generate a sorted list of biconnected components, largest first, using sort.

248 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G.remove_edge(0, 5)
>>> [len(c) for c in sorted(nx.biconnected_components(G), key=len, reverse=True)]
[5, 2]

If you only want the largest connected component, it’s more efficient to use max instead of sort.

>>> Gc = max(nx.biconnected_components(G), key=len)

See also:
is_biconnected(), articulation_points(), biconnected_component_edges()

Notes

The algorithm to find articulation points and biconnected components is implemented using a non-recursive
depth-first-search (DFS) that keeps track of the highest level that back edges reach in the DFS tree. A node n
is an articulation point if, and only if, there exists a subtree rooted at n such that there is no back edge from
any successor of n that links to a predecessor of n in the DFS tree. By keeping track of all the edges traversed
by the DFS we can obtain the biconnected components because all edges of a bicomponent will be traversed
consecutively between articulation points.

References

networkx.algorithms.components.biconnected_component_edges

biconnected_component_edges(G)
Returns a generator of lists of edges, one list for each biconnected component of the input graph.
Biconnected components are maximal subgraphs such that the removal of a node (and all edges incident on that
node) will not disconnect the subgraph. Note that nodes may be part of more than one biconnected compo-
nent. Those nodes are articulation points, or cut vertices. However, each edge belongs to one, and only one,
biconnected component.
Notice that by convention a dyad is considered a biconnected component.
Parameters G (NetworkX Graph) – An undirected graph.
Returns edges – Generator of lists of edges, one list for each bicomponent.
Return type generator of lists
Raises NetworkXNotImplemented : – If the input graph is not undirected.

Examples

>>> G = nx.barbell_graph(4, 2)
>>> print(nx.is_biconnected(G))
False
>>> bicomponents_edges = list(nx.biconnected_component_edges(G))
>>> len(bicomponents_edges)
5
>>> G.add_edge(2, 8)
>>> print(nx.is_biconnected(G))
True
>>> bicomponents_edges = list(nx.biconnected_component_edges(G))
(continues on next page)

3.14. Components 249


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> len(bicomponents_edges)
1

See also:
is_biconnected(), biconnected_components(), articulation_points()

Notes

The algorithm to find articulation points and biconnected components is implemented using a non-recursive
depth-first-search (DFS) that keeps track of the highest level that back edges reach in the DFS tree. A node n
is an articulation point if, and only if, there exists a subtree rooted at n such that there is no back edge from
any successor of n that links to a predecessor of n in the DFS tree. By keeping track of all the edges traversed
by the DFS we can obtain the biconnected components because all edges of a bicomponent will be traversed
consecutively between articulation points.

References

networkx.algorithms.components.biconnected_component_subgraphs

biconnected_component_subgraphs(G, copy=True)
DEPRECATED: Use (G.subgraph(c) for c in biconnected_components(G))
Or (G.subgraph(c).copy() for c in biconnected_components(G))

networkx.algorithms.components.articulation_points

articulation_points(G)
Yield the articulation points, or cut vertices, of a graph.
An articulation point or cut vertex is any node whose removal (along with all its incident edges) increases the
number of connected components of a graph. An undirected connected graph without articulation points is
biconnected. Articulation points belong to more than one biconnected component of a graph.
Notice that by convention a dyad is considered a biconnected component.
Parameters G (NetworkX Graph) – An undirected graph.
Yields node – An articulation point in the graph.
Raises NetworkXNotImplemented : – If the input graph is not undirected.

Examples

>>> G = nx.barbell_graph(4, 2)
>>> print(nx.is_biconnected(G))
False
>>> len(list(nx.articulation_points(G)))
4
>>> G.add_edge(2, 8)
>>> print(nx.is_biconnected(G))
True
>>> len(list(nx.articulation_points(G)))
0

250 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
is_biconnected(), biconnected_components(), biconnected_component_edges()

Notes

The algorithm to find articulation points and biconnected components is implemented using a non-recursive
depth-first-search (DFS) that keeps track of the highest level that back edges reach in the DFS tree. A node n
is an articulation point if, and only if, there exists a subtree rooted at n such that there is no back edge from
any successor of n that links to a predecessor of n in the DFS tree. By keeping track of all the edges traversed
by the DFS we can obtain the biconnected components because all edges of a bicomponent will be traversed
consecutively between articulation points.

References

3.14.6 Semiconnectedness

is_semiconnected(G) Returns True if the graph is semiconnected, False other-


wise.

networkx.algorithms.components.is_semiconnected

is_semiconnected(G)
Returns True if the graph is semiconnected, False otherwise.
A graph is semiconnected if, and only if, for any pair of nodes, either one is reachable from the other, or they
are mutually reachable.
Parameters G (NetworkX graph) – A directed graph.
Returns semiconnected – True if the graph is semiconnected, False otherwise.
Return type bool
Raises
• NetworkXNotImplemented : – If the input graph is undirected.
• NetworkXPointlessConcept : – If the graph is empty.

Examples

>>> G=nx.path_graph(4,create_using=nx.DiGraph())
>>> print(nx.is_semiconnected(G))
True
>>> G=nx.DiGraph([(1, 2), (3, 2)])
>>> print(nx.is_semiconnected(G))
False

See also:
is_strongly_connected(), is_weakly_connected(), is_connected(),
is_biconnected()

3.14. Components 251


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.15 Connectivity

Connectivity and cut algorithms

3.15.1 Edge-augmentation

Algorithms for finding k-edge-augmentations


A k-edge-augmentation is a set of edges, that once added to a graph, ensures that the graph is k-edge-connected; i.e.
the graph cannot be disconnected unless k or more edges are removed. Typically, the goal is to find the augmentation
with minimum weight. In general, it is not guaranteed that a k-edge-augmentation exists.
See also:
edge_kcomponents algorithms for finding k-edge-connected components
connectivity algorithms for determening edge connectivity.

k_edge_augmentation(G, k[, avail, weight, . . . ]) Finds set of edges to k-edge-connect G.


is_k_edge_connected(G, k) Tests to see if a graph is k-edge-connected.
is_locally_k_edge_connected(G, s, t, k) Tests to see if an edge in a graph is locally k-edge-
connected.

networkx.algorithms.connectivity.edge_augmentation.k_edge_augmentation

k_edge_augmentation(G, k, avail=None, weight=None, partial=False)


Finds set of edges to k-edge-connect G.
Adding edges from the augmentation to G make it impossible to disconnect G unless k or more edges are
removed. This function uses the most efficient function available (depending on the value of k and if the problem
is weighted or unweighted) to search for a minimum weight subset of available edges that k-edge-connects G. In
general, finding a k-edge-augmentation is NP-hard, so solutions are not garuenteed to be minimal. Furthermore,
a k-edge-augmentation may not exist.
Parameters
• G (NetworkX graph) – An undirected graph.
• k (integer) – Desired edge connectivity
• avail (dict or a set of 2 or 3 tuples) – The available edges that can be used in the augmenta-
tion.
If unspecified, then all edges in the complement of G are available. Otherwise, each item is
an available edge (with an optional weight).
In the unweighted case, each item is an edge (u, v).
In the weighted case, each item is a 3-tuple (u, v, d) or a dict with items (u, v):
d. The third item, d, can be a dictionary or a real number. If d is a dictionary d[weight]
correspondings to the weight.
• weight (string) – key to use to find weights if avail is a set of 3-tuples where the third
item in each tuple is a dictionary.
• partial (boolean) – If partial is True and no feasible k-edge-augmentation exists, then all a
partial k-edge-augmentation is generated. Adding the edges in a partial augmentation to G,

252 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

minimizes the number of k-edge-connected components and maximizes the edge connectiv-
ity between those components. For details, see partial_k_edge_augmentation().
Yields edge (tuple) – Edges that, once added to G, would cause G to become k-edge-connected.
If partial is False, an error is raised if this is not possible. Otherwise, generated edges form a
partial augmentation, which k-edge-connects any part of G where it is possible, and maximally
connects the remaining parts.
Raises
• NetworkXUnfeasible: – If partial is False and no k-edge-augmentation exists.
• NetworkXNotImplemented: – If the input graph is directed or a multigraph.
• ValueError: – If k is less than 1

Notes

When k=1 this returns an optimal solution.


When k=2 and avail is None, this returns an optimal solution. Otherwise when k=2, this returns a 2-
approximation of the optimal solution.
For k>3, this problem is NP-hard and this uses a randomized algorithm that produces a feasible solution,
but provides no guarantees on the solution weight.

Example

>>> # Unweighted cases


>>> G = nx.path_graph((1, 2, 3, 4))
>>> G.add_node(5)
>>> sorted(nx.k_edge_augmentation(G, k=1))
[(1, 5)]
>>> sorted(nx.k_edge_augmentation(G, k=2))
[(1, 5), (5, 4)]
>>> sorted(nx.k_edge_augmentation(G, k=3))
[(1, 4), (1, 5), (2, 5), (3, 5), (4, 5)]
>>> complement = list(nx.k_edge_augmentation(G, k=5, partial=True))
>>> G.add_edges_from(complement)
>>> nx.edge_connectivity(G)
4

Example

>>> # Weighted cases


>>> G = nx.path_graph((1, 2, 3, 4))
>>> G.add_node(5)
>>> # avail can be a tuple with a dict
>>> avail = [(1, 5, {'weight': 11}), (2, 5, {'weight': 10})]
>>> sorted(nx.k_edge_augmentation(G, k=1, avail=avail, weight='weight'))
[(2, 5)]
>>> # or avail can be a 3-tuple with a real number
>>> avail = [(1, 5, 11), (2, 5, 10), (4, 3, 1), (4, 5, 51)]
>>> sorted(nx.k_edge_augmentation(G, k=2, avail=avail))
[(1, 5), (2, 5), (4, 5)]
(continues on next page)

3.15. Connectivity 253


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> # or avail can be a dict
>>> avail = {(1, 5): 11, (2, 5): 10, (4, 3): 1, (4, 5): 51}
>>> sorted(nx.k_edge_augmentation(G, k=2, avail=avail))
[(1, 5), (2, 5), (4, 5)]
>>> # If augmentation is infeasible, then a partial solution can be found
>>> avail = {(1, 5): 11}
>>> sorted(nx.k_edge_augmentation(G, k=2, avail=avail, partial=True))
[(1, 5)]

networkx.algorithms.connectivity.edge_augmentation.is_k_edge_connected

is_k_edge_connected(G, k)
Tests to see if a graph is k-edge-connected.
Is it impossible to disconnect the graph by removing fewer than k edges? If so, then G is k-edge-connected.
Parameters
• G (NetworkX graph) – An undirected graph.
• k (integer) – edge connectivity to test for
Returns True if G is k-edge-connected.
Return type boolean
See also:
is_locally_k_edge_connected()

Example

>>> G = nx.barbell_graph(10, 0)
>>> nx.is_k_edge_connected(G, k=1)
True
>>> nx.is_k_edge_connected(G, k=2)
False

networkx.algorithms.connectivity.edge_augmentation.is_locally_k_edge_connected

is_locally_k_edge_connected(G, s, t, k)
Tests to see if an edge in a graph is locally k-edge-connected.
Is it impossible to disconnect s and t by removing fewer than k edges? If so, then s and t are locally k-edge-
connected in G.
Parameters
• G (NetworkX graph) – An undirected graph.
• s (node) – Source node
• t (node) – Target node
• k (integer) – local edge connectivity for nodes s and t
Returns True if s and t are locally k-edge-connected in G.

254 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type boolean


See also:
is_k_edge_connected()

Example

>>> from networkx.algorithms.connectivity import is_locally_k_edge_connected


>>> G = nx.barbell_graph(10, 0)
>>> is_locally_k_edge_connected(G, 5, 15, k=1)
True
>>> is_locally_k_edge_connected(G, 5, 15, k=2)
False
>>> is_locally_k_edge_connected(G, 1, 5, k=2)
True

3.15.2 K-edge-components

Algorithms for finding k-edge-connected components and subgraphs.


A k-edge-connected component (k-edge-cc) is a maximal set of nodes in G, such that all pairs of node have an edge-
connectivity of at least k.
A k-edge-connected subgraph (k-edge-subgraph) is a maximal set of nodes in G, such that the subgraph of G defined
by the nodes has an edge-connectivity at least k.

k_edge_components(G, k) Generates nodes in each maximal k-edge-connected


component in G.
k_edge_subgraphs(G, k) Generates nodes in each maximal k-edge-connected
subgraph in G.
bridge_components(G) Finds all bridge-connected components G.
EdgeComponentAuxGraph A simple algorithm to find all k-edge-connected com-
ponents in a graph.

networkx.algorithms.connectivity.edge_kcomponents.k_edge_components

k_edge_components(G, k)
Generates nodes in each maximal k-edge-connected component in G.
Parameters
• G (NetworkX graph)
• k (Integer) – Desired edge connectivity
Returns k_edge_components – will have k-edge-connectivity in the graph G.
Return type a generator of k-edge-ccs. Each set of returned nodes
See also:
local_edge_connectivity()
k_edge_subgraphs() similar to this function, but the subgraph defined by the nodes must also have k-
edge-connectivity.

3.15. Connectivity 255


NetworkX Reference, Release 2.3rc1.dev20190222214247

k_components() similar to this function, but uses node-connectivity instead of edge-connectivity

Raises
• NetworkXNotImplemented: – If the input graph is a multigraph.
• ValueError: – If k is less than 1

Notes

Attempts to use the most efficient implementation available based on k. If k=1, this is simply simply connected
components for directed graphs and connected components for undirected graphs. If k=2 on an efficient bridge
connected component algorithm from _[1] is run based on the chain decomposition. Otherwise, the algorithm
from _[2] is used.

Example

>>> import itertools as it


>>> from networkx.utils import pairwise
>>> paths = [
... (1, 2, 4, 3, 1, 4),
... (5, 6, 7, 8, 5, 7, 8, 6),
... ]
>>> G = nx.Graph()
>>> G.add_nodes_from(it.chain(*paths))
>>> G.add_edges_from(it.chain(*[pairwise(path) for path in paths]))
>>> # note this returns {1, 4} unlike k_edge_subgraphs
>>> sorted(map(sorted, nx.k_edge_components(G, k=3)))
[[1, 4], [2], [3], [5, 6, 7, 8]]

References

networkx.algorithms.connectivity.edge_kcomponents.k_edge_subgraphs

k_edge_subgraphs(G, k)
Generates nodes in each maximal k-edge-connected subgraph in G.
Parameters
• G (NetworkX graph)
• k (Integer) – Desired edge connectivity
Returns k_edge_subgraphs – Each k-edge-subgraph is a maximal set of nodes that defines a sub-
graph of G that is k-edge-connected.
Return type a generator of k-edge-subgraphs
See also:
edge_connectivity()
k_edge_components() similar to this function, but nodes only need to have k-edge-connctivity within the
graph G and the subgraphs might not be k-edge-connected.

Raises
• NetworkXNotImplemented: – If the input graph is a multigraph.

256 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• ValueError: – If k is less than 1

Notes

Attempts to use the most efficient implementation available based on k. If k=1, or k=2 and the graph is undi-
rected, then this simply calls k_edge_components. Otherwise the algorithm from _[1] is used.

Example

>>> import itertools as it


>>> from networkx.utils import pairwise
>>> paths = [
... (1, 2, 4, 3, 1, 4),
... (5, 6, 7, 8, 5, 7, 8, 6),
... ]
>>> G = nx.Graph()
>>> G.add_nodes_from(it.chain(*paths))
>>> G.add_edges_from(it.chain(*[pairwise(path) for path in paths]))
>>> # note this does not return {1, 4} unlike k_edge_components
>>> sorted(map(sorted, nx.k_edge_subgraphs(G, k=3)))
[[1], [2], [3], [4], [5, 6, 7, 8]]

References

networkx.algorithms.connectivity.edge_kcomponents.bridge_components

bridge_components(G)
Finds all bridge-connected components G.
Parameters G (NetworkX undirected graph)
Returns bridge_components
Return type a generator of 2-edge-connected components
See also:

k_edge_subgraphs() this function is a special case for an undirected graph where k=2.
biconnected_components() similar to this function, but is defined using 2-node-connectivity instead of
2-edge-connectivity.

Raises NetworkXNotImplemented: – If the input graph is directed or a multigraph.

Notes

Bridge-connected components are also known as 2-edge-connected components.

Example

3.15. Connectivity 257


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> # The barbell graph with parameter zero has a single bridge
>>> G = nx.barbell_graph(5, 0)
>>> from networkx.algorithms.connectivity.edge_kcomponents import bridge_
˓→components

>>> sorted(map(sorted, bridge_components(G)))


[[0, 1, 2, 3, 4], [5, 6, 7, 8, 9]]

networkx.algorithms.connectivity.edge_kcomponents.EdgeComponentAuxGraph

class EdgeComponentAuxGraph
A simple algorithm to find all k-edge-connected components in a graph.
Constructing the AuxillaryGraph (which may take some time) allows for the k-edge-ccs to be found in linear
time for arbitrary k.

Notes

This implementation is based on1 . The idea is to construct an auxiliary graph from which the k-edge-ccs can be
extracted in linear time. The auxiliary graph is constructed in 𝑂(|𝑉 | · 𝐹 ) operations, where F is the complexity
of max flow. Querying the components takes an additional 𝑂(|𝑉 |) operations. This algorithm can be slow for
large graphs, but it handles an arbitrary k and works for both directed and undirected inputs.
The undirected case for k=1 is exactly connected components. The undirected case for k=2 is exactly bridge
connected components. The directed case for k=1 is exactly strongly connected components.

References

Example

>>> import itertools as it


>>> from networkx.utils import pairwise
>>> from networkx.algorithms.connectivity import EdgeComponentAuxGraph
>>> # Build an interesting graph with multiple levels of k-edge-ccs
>>> paths = [
... (1, 2, 3, 4, 1, 3, 4, 2), # a 3-edge-cc (a 4 clique)
... (5, 6, 7, 5), # a 2-edge-cc (a 3 clique)
... (1, 5), # combine first two ccs into a 1-edge-cc
... (0,), # add an additional disconnected 1-edge-cc
... ]
>>> G = nx.Graph()
>>> G.add_nodes_from(it.chain(*paths))
>>> G.add_edges_from(it.chain(*[pairwise(path) for path in paths]))
>>> # Constructing the AuxGraph takes about O(n ** 4)
>>> aux_graph = EdgeComponentAuxGraph.construct(G)
>>> # Once constructed, querying takes O(n)
>>> sorted(map(sorted, aux_graph.k_edge_components(k=1)))
[[0], [1, 2, 3, 4, 5, 6, 7]]
>>> sorted(map(sorted, aux_graph.k_edge_components(k=2)))
[[0], [1, 2, 3, 4], [5, 6, 7]]
>>> sorted(map(sorted, aux_graph.k_edge_components(k=3)))
[[0], [1, 2, 3, 4], [5], [6], [7]]
(continues on next page)
1Wang, Tianhao, et al. (2015) A simple algorithm for finding all k-edge-connected components. http://journals.plos.org/plosone/article?id=10.
1371/journal.pone.0136264

258 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> sorted(map(sorted, aux_graph.k_edge_components(k=4)))
[[0], [1], [2], [3], [4], [5], [6], [7]]

Example

>>> # The auxiliary graph is primarilly used for k-edge-ccs but it


>>> # can also speed up the queries of k-edge-subgraphs by refining the
>>> # search space.
>>> import itertools as it
>>> from networkx.utils import pairwise
>>> from networkx.algorithms.connectivity import EdgeComponentAuxGraph
>>> paths = [
... (1, 2, 4, 3, 1, 4),
... ]
>>> G = nx.Graph()
>>> G.add_nodes_from(it.chain(*paths))
>>> G.add_edges_from(it.chain(*[pairwise(path) for path in paths]))
>>> aux_graph = EdgeComponentAuxGraph.construct(G)
>>> sorted(map(sorted, aux_graph.k_edge_subgraphs(k=3)))
[[1], [2], [3], [4]]
>>> sorted(map(sorted, aux_graph.k_edge_components(k=3)))
[[1, 4], [2], [3]]

__init__()
Initialize self. See help(type(self)) for accurate signature.

Methods

construct(G) Builds an auxiliary graph encoding edge-


connectivity between nodes.
k_edge_components(k) Queries the auxiliary graph for k-edge-connected
components.
k_edge_subgraphs(k) Queries the auxiliary graph for k-edge-connected
subgraphs.

3.15.3 K-node-components

Moody and White algorithm for k-components

k_components(G[, flow_func]) Returns the k-component structure of a graph G.

networkx.algorithms.connectivity.kcomponents.k_components

k_components(G, flow_func=None)
Returns the k-component structure of a graph G.
A k-component is a maximal subgraph of a graph G that has, at least, node connectivity k: we need to remove at
least k nodes to break it into more components. k-components have an inherent hierarchical structure because
they are nested in terms of connectivity: a connected graph can contain several 2-components, each of which
can contain one or more 3-components, and so forth.

3.15. Connectivity 259


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (NetworkX graph)
• flow_func (function) – Function to perform the underlying flow computations. Default value
edmonds_karp(). This function performs better in sparse graphs with right tailed degree
distributions. shortest_augmenting_path() will perform better in denser graphs.
Returns k_components – Dictionary with all connectivity levels k in the input Graph as keys and a
list of sets of nodes that form a k-component of level k as values.
Return type dict
Raises NetworkXNotImplemented: – If the input graph is directed.

Examples

>>> # Petersen graph has 10 nodes and it is triconnected, thus all


>>> # nodes are in a single component on all three connectivity levels
>>> G = nx.petersen_graph()
>>> k_components = nx.k_components(G)

Notes

Moody and White1 (appendix A) provide an algorithm for identifying k-components in a graph, which is
based on Kanevsky’s algorithm2 for finding all minimum-size node cut-sets of a graph (implemented in
all_node_cuts() function):
1. Compute node connectivity, k, of the input graph G.
2. Identify all k-cutsets at the current level of connectivity using Kanevsky’s algorithm.
3. Generate new graph components based on the removal of these cutsets. Nodes in a cutset belong to both
sides of the induced cut.
4. If the graph is neither complete nor trivial, return to 1; else end.
This implementation also uses some heuristics (see3 for details) to speed up the computation.
See also:
node_connectivity(), all_node_cuts()
biconnected_components() special case of this function when k=2
k_edge_components() similar to this function, but uses edge-connectivity instead of node-connectivity

References

3.15.4 K-node-cutsets

Kanevsky all minimum node k cutsets algorithm.

1 Moody, J. and D. White (2003). Social cohesion and embeddedness: A hierarchical conception of social groups. American Sociological

Review 68(1), 103–28. http://www2.asanet.org/journals/ASRFeb03MoodyWhite.pdf


2 Kanevsky, A. (1993). Finding all minimum-size separating vertex sets in a graph. Networks 23(6), 533–541. http://onlinelibrary.wiley.com/

doi/10.1002/net.3230230604/abstract
3 Torrents, J. and F. Ferraro (2015). Structural Cohesion: Visualization and Heuristics for Fast Computation. https://arxiv.org/pdf/1503.04476v1

260 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

all_node_cuts(G[, k, flow_func]) Returns all minimum k cutsets of an undirected graph


G.

networkx.algorithms.connectivity.kcutsets.all_node_cuts

all_node_cuts(G, k=None, flow_func=None)


Returns all minimum k cutsets of an undirected graph G.
This implementation is based on Kanevsky’s algorithm1 for finding all minimum-size node cut-sets of an undi-
rected graph G; ie the set (or sets) of nodes of cardinality equal to the node connectivity of G. Thus if removed,
would break G into two or more connected components.
Parameters
• G (NetworkX graph) – Undirected graph
• k (Integer) – Node connectivity of the input graph. If k is None, then it is computed. Default
value: None.
• flow_func (function) – Function to perform the underlying flow computations. Default value
edmonds_karp. This function performs better in sparse graphs with right tailed degree dis-
tributions. shortest_augmenting_path will perform better in denser graphs.
Returns cuts – Each node cutset has cardinality equal to the node connectivity of the input graph.
Return type a generator of node cutsets

Examples

>>> # A two-dimensional grid graph has 4 cutsets of cardinality 2


>>> G = nx.grid_2d_graph(5, 5)
>>> cutsets = list(nx.all_node_cuts(G))
>>> len(cutsets)
4
>>> all(2 == len(cutset) for cutset in cutsets)
True
>>> nx.node_connectivity(G)
2

Notes

This implementation is based on the sequential algorithm for finding all minimum-size separating vertex sets in
a graph1 . The main idea is to compute minimum cuts using local maximum flow computations among a set of
nodes of highest degree and all other non-adjacent nodes in the Graph. Once we find a minimum cut, we add an
edge between the high degree node and the target node of the local maximum flow computation to make sure
that we will not find that minimum cut again.
See also:
node_connectivity(), edmonds_karp(), shortest_augmenting_path()
1 Kanevsky, A. (1993). Finding all minimum-size separating vertex sets in a graph. Networks 23(6), 533–541. http://onlinelibrary.wiley.com/
doi/10.1002/net.3230230604/abstract

3.15. Connectivity 261


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.15.5 Flow-based disjoint paths

Flow based node and edge disjoint paths.

edge_disjoint_paths(G, s, t[, flow_func, . . . ]) Returns the edges disjoint paths between source and tar-
get.
node_disjoint_paths(G, s, t[, flow_func, . . . ]) Computes node disjoint paths between source and tar-
get.

networkx.algorithms.connectivity.disjoint_paths.edge_disjoint_paths

edge_disjoint_paths(G, s, t, flow_func=None, cutoff=None, auxiliary=None, residual=None)


Returns the edges disjoint paths between source and target.
Edge disjoint paths are paths that do not share any edge. The number of edge disjoint paths between source and
target is equal to their edge connectivity.
Parameters
• G (NetworkX graph)
• s (node) – Source node for the flow.
• t (node) – Sink node for the flow.
• flow_func (function) – A function for computing the maximum flow among a pair of
nodes. The function has to accept at least three parameters: a Digraph, a source node,
and a target node. And return a residual network that follows NetworkX conventions (see
maximum_flow() for details). If flow_func is None, the default maximum flow function
(edmonds_karp()) is used. The choice of the default function may change from version
to version and should not be relied on. Default value: None.
• cutoff (int) – Maximum number of paths to yield. Some of the maximum flow algorithms,
such as edmonds_karp() (the default) and shortest_augmenting_path() sup-
port the cutoff parameter, and will terminate when the flow value reaches or exceeds the
cutoff. Other algorithms will ignore this parameter. Default value: None.
• auxiliary (NetworkX DiGraph) – Auxiliary digraph to compute flow based edge connectiv-
ity. It has to have a graph attribute called mapping with a dictionary mapping node names
in G and in the auxiliary digraph. If provided it will be reused instead of recreated. Default
value: None.
• residual (NetworkX DiGraph) – Residual network to compute maximum flow. If provided
it will be reused instead of recreated. Default value: None.
Returns paths – A generator of edge independent paths.
Return type generator
Raises
• NetworkXNoPath : exception – If there is no path between source and target.
• NetworkXError : exception – If source or target are not in the graph G.
See also:

262 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

node_disjoint_paths(), edge_connectivity(), maximum_flow(), edmonds_karp(),


preflow_push(), shortest_augmenting_path()

Examples

We use in this example the platonic icosahedral graph, which has node edge connectivity 5, thus there are 5 edge
disjoint paths between any pair of nodes.

>>> G = nx.icosahedral_graph()
>>> len(list(nx.edge_disjoint_paths(G, 0, 6)))
5

If you need to compute edge disjoint paths on several pairs of nodes in the same graph, it is recommended that
you reuse the data structures that NetworkX uses in the computation: the auxiliary digraph for edge connectivity,
and the residual network for the underlying maximum flow computation.
Example of how to compute edge disjoint paths among all pairs of nodes of the platonic icosahedral graph
reusing the data structures.

>>> import itertools


>>> # You also have to explicitly import the function for
>>> # building the auxiliary digraph from the connectivity package
>>> from networkx.algorithms.connectivity import (
... build_auxiliary_edge_connectivity)
>>> H = build_auxiliary_edge_connectivity(G)
>>> # And the function for building the residual network from the
>>> # flow package
>>> from networkx.algorithms.flow import build_residual_network
>>> # Note that the auxiliary digraph has an edge attribute named capacity
>>> R = build_residual_network(H, 'capacity')
>>> result = {n: {} for n in G}
>>> # Reuse the auxiliary digraph and the residual network by passing them
>>> # as arguments
>>> for u, v in itertools.combinations(G, 2):
... k = len(list(nx.edge_disjoint_paths(G, u, v, auxiliary=H, residual=R)))
... result[u][v] = k
>>> all(result[u][v] == 5 for u, v in itertools.combinations(G, 2))
True

You can also use alternative flow algorithms for computing edge disjoint paths. For instance, in dense
networks the algorithm shortest_augmenting_path() will usually perform better than the default
edmonds_karp() which is faster for sparse networks with highly skewed degree distributions. Alternative
flow functions have to be explicitly imported from the flow package.

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> len(list(nx.edge_disjoint_paths(G, 0, 6, flow_func=shortest_augmenting_path)))
5

Notes

This is a flow based implementation of edge disjoint paths. We compute the maximum flow between source and
target on an auxiliary directed network. The saturated edges in the residual network after running the maximum
flow algorithm correspond to edge disjoint paths between source and target in the original network. This function
handles both directed and undirected graphs, and can use all flow algorithms from NetworkX flow package.

3.15. Connectivity 263


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.connectivity.disjoint_paths.node_disjoint_paths

node_disjoint_paths(G, s, t, flow_func=None, cutoff=None, auxiliary=None, residual=None)


Computes node disjoint paths between source and target.
Node dijoint paths are paths that only share their first and last nodes. The number of node independent paths
between two nodes is equal to their local node connectivity.
Parameters
• G (NetworkX graph)
• s (node) – Source node.
• t (node) – Target node.
• flow_func (function) – A function for computing the maximum flow among a pair of
nodes. The function has to accept at least three parameters: a Digraph, a source node,
and a target node. And return a residual network that follows NetworkX conventions (see
maximum_flow() for details). If flow_func is None, the default maximum flow function
(edmonds_karp()) is used. See below for details. The choice of the default function
may change from version to version and should not be relied on. Default value: None.
• cutoff (int) – Maximum number of paths to yield. Some of the maximum flow algorithms,
such as edmonds_karp() (the default) and shortest_augmenting_path() sup-
port the cutoff parameter, and will terminate when the flow value reaches or exceeds the
cutoff. Other algorithms will ignore this parameter. Default value: None.
• auxiliary (NetworkX DiGraph) – Auxiliary digraph to compute flow based node connectiv-
ity. It has to have a graph attribute called mapping with a dictionary mapping node names
in G and in the auxiliary digraph. If provided it will be reused instead of recreated. Default
value: None.
• residual (NetworkX DiGraph) – Residual network to compute maximum flow. If provided
it will be reused instead of recreated. Default value: None.
Returns paths – Generator of node disjoint paths.
Return type generator
Raises
• NetworkXNoPath : exception – If there is no path between source and target.
• NetworkXError : exception – If source or target are not in the graph G.

Examples

We use in this example the platonic icosahedral graph, which has node node connectivity 5, thus there are 5
node disjoint paths between any pair of non neighbor nodes.

>>> G = nx.icosahedral_graph()
>>> len(list(nx.node_disjoint_paths(G, 0, 6)))
5

If you need to compute node disjoint paths between several pairs of nodes in the same graph, it is recommended
that you reuse the data structures that NetworkX uses in the computation: the auxiliary digraph for node con-
nectivity and node cuts, and the residual network for the underlying maximum flow computation.
Example of how to compute node disjoint paths reusing the data structures:

264 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> # You also have to explicitly import the function for


>>> # building the auxiliary digraph from the connectivity package
>>> from networkx.algorithms.connectivity import (
... build_auxiliary_node_connectivity)
>>> H = build_auxiliary_node_connectivity(G)
>>> # And the function for building the residual network from the
>>> # flow package
>>> from networkx.algorithms.flow import build_residual_network
>>> # Note that the auxiliary digraph has an edge attribute named capacity
>>> R = build_residual_network(H, 'capacity')
>>> # Reuse the auxiliary digraph and the residual network by passing them
>>> # as arguments
>>> len(list(nx.node_disjoint_paths(G, 0, 6, auxiliary=H, residual=R)))
5

You can also use alternative flow algorithms for computing node disjoint paths. For instance, in dense
networks the algorithm shortest_augmenting_path() will usually perform better than the default
edmonds_karp() which is faster for sparse networks with highly skewed degree distributions. Alternative
flow functions have to be explicitly imported from the flow package.

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> len(list(nx.node_disjoint_paths(G, 0, 6, flow_func=shortest_augmenting_path)))
5

Notes

This is a flow based implementation of node disjoint paths. We compute the maximum flow between source
and target on an auxiliary directed network. The saturated edges in the residual network after running the
maximum flow algorithm correspond to node disjoint paths between source and target in the original network.
This function handles both directed and undirected graphs, and can use all flow algorithms from NetworkX flow
package.
See also:
edge_disjoint_paths(), node_connectivity(), maximum_flow(), edmonds_karp(),
preflow_push(), shortest_augmenting_path()

3.15.6 Flow-based Connectivity

Flow based connectivity algorithms

average_node_connectivity(G[, flow_func]) Returns the average connectivity of a graph G.


all_pairs_node_connectivity(G[, nbunch, Compute node connectivity between all pairs of nodes
. . . ]) of G.
edge_connectivity(G[, s, t, flow_func, cutoff]) Returns the edge connectivity of the graph or digraph G.
local_edge_connectivity(G, s, t[, . . . ]) Returns local edge connectivity for nodes s and t in G.
local_node_connectivity(G, s, t[, . . . ]) Computes local node connectivity for nodes s and t.
node_connectivity(G[, s, t, flow_func]) Returns node connectivity for a graph or digraph G.

3.15. Connectivity 265


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.connectivity.connectivity.average_node_connectivity

average_node_connectivity(G, flow_func=None)
Returns the average connectivity of a graph G.
The average connectivity bar{kappa} of a graph G is the average of local node connectivity over all pairs of
nodes of G1 .
∑︀
𝑢,𝑣 𝜅𝐺 (𝑢, 𝑣)
𝜅
¯ (𝐺) = (︀𝑛)︀
2

Parameters
• G (NetworkX graph) – Undirected graph
• flow_func (function) – A function for computing the maximum flow among a pair of
nodes. The function has to accept at least three parameters: a Digraph, a source node,
and a target node. And return a residual network that follows NetworkX conventions (see
maximum_flow() for details). If flow_func is None, the default maximum flow function
(edmonds_karp()) is used. See local_node_connectivity() for details. The
choice of the default function may change from version to version and should not be relied
on. Default value: None.
Returns K – Average node connectivity
Return type float
See also:
local_node_connectivity(), node_connectivity(), edge_connectivity(),
maximum_flow(), edmonds_karp(), preflow_push(), shortest_augmenting_path()

References

networkx.algorithms.connectivity.connectivity.all_pairs_node_connectivity

all_pairs_node_connectivity(G, nbunch=None, flow_func=None)


Compute node connectivity between all pairs of nodes of G.
Parameters
• G (NetworkX graph) – Undirected graph
• nbunch (container) – Container of nodes. If provided node connectivity will be computed
only over pairs of nodes in nbunch.
• flow_func (function) – A function for computing the maximum flow among a pair of
nodes. The function has to accept at least three parameters: a Digraph, a source node,
and a target node. And return a residual network that follows NetworkX conventions (see
maximum_flow() for details). If flow_func is None, the default maximum flow function
(edmonds_karp()) is used. See below for details. The choice of the default function
may change from version to version and should not be relied on. Default value: None.
Returns all_pairs – A dictionary with node connectivity between all pairs of nodes in G, or in
nbunch if provided.
Return type dict
1 Beineke, L., O. Oellermann, and R. Pippert (2002). The average connectivity of a graph. Discrete mathematics 252(1-3), 31-45. http:

//www.sciencedirect.com/science/article/pii/S0012365X01001807

266 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
local_node_connectivity(), edge_connectivity(), local_edge_connectivity(),
maximum_flow(), edmonds_karp(), preflow_push(), shortest_augmenting_path()

networkx.algorithms.connectivity.connectivity.edge_connectivity

edge_connectivity(G, s=None, t=None, flow_func=None, cutoff=None)


Returns the edge connectivity of the graph or digraph G.
The edge connectivity is equal to the minimum number of edges that must be removed to disconnect G or render
it trivial. If source and target nodes are provided, this function returns the local edge connectivity: the minimum
number of edges that must be removed to break all paths from source to target in G.
Parameters
• G (NetworkX graph) – Undirected or directed graph
• s (node) – Source node. Optional. Default value: None.
• t (node) – Target node. Optional. Default value: None.
• flow_func (function) – A function for computing the maximum flow among a pair of
nodes. The function has to accept at least three parameters: a Digraph, a source node,
and a target node. And return a residual network that follows NetworkX conventions (see
maximum_flow() for details). If flow_func is None, the default maximum flow function
(edmonds_karp()) is used. See below for details. The choice of the default function
may change from version to version and should not be relied on. Default value: None.
• cutoff (integer, float) – If specified, the maximum flow algorithm will terminate when the
flow value reaches or exceeds the cutoff. This is only for the algorithms that support the
cutoff parameter: edmonds_karp() and shortest_augmenting_path(). Other
algorithms will ignore this parameter. Default value: None.
Returns K – Edge connectivity for G, or local edge connectivity if source and target were provided
Return type integer

Examples

>>> # Platonic icosahedral graph is 5-edge-connected


>>> G = nx.icosahedral_graph()
>>> nx.edge_connectivity(G)
5

You can use alternative flow algorithms for the underlying maximum flow computation. In dense net-
works the algorithm shortest_augmenting_path() will usually perform better than the default
edmonds_karp(), which is faster for sparse networks with highly skewed degree distributions. Alternative
flow functions have to be explicitly imported from the flow package.

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> nx.edge_connectivity(G, flow_func=shortest_augmenting_path)
5

If you specify a pair of nodes (source and target) as parameters, this function returns the value of local edge
connectivity.

3.15. Connectivity 267


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> nx.edge_connectivity(G, 3, 7)
5

If you need to perform several local computations among different pairs of nodes on the same graph,
it is recommended that you reuse the data structures used in the maximum flow computations. See
local_edge_connectivity() for details.

Notes

This is a flow based implementation of global edge connectivity. For undirected graphs the algorithm works by
finding a ‘small’ dominating set of nodes of G (see algorithm 7 in1 ) and computing local maximum flow (see
local_edge_connectivity()) between an arbitrary node in the dominating set and the rest of nodes in
it. This is an implementation of algorithm 6 in1 . For directed graphs, the algorithm does n calls to the maximum
flow function. This is an implementation of algorithm 8 in1 .
See also:
local_edge_connectivity(), local_node_connectivity(), node_connectivity(),
maximum_flow(), edmonds_karp(), preflow_push(), shortest_augmenting_path(),
k_edge_components(), k_edge_subgraphs()

References

networkx.algorithms.connectivity.connectivity.local_edge_connectivity

local_edge_connectivity(G, s, t, flow_func=None, auxiliary=None, residual=None, cutoff=None)


Returns local edge connectivity for nodes s and t in G.
Local edge connectivity for two nodes s and t is the minimum number of edges that must be removed to discon-
nect them.
This is a flow based implementation of edge connectivity. We compute the maximum flow on an auxiliary
digraph build from the original network (see below for details). This is equal to the local edge connectivity
because the value of a maximum s-t-flow is equal to the capacity of a minimum s-t-cut (Ford and Fulkerson
theorem)1 .
Parameters
• G (NetworkX graph) – Undirected or directed graph
• s (node) – Source node
• t (node) – Target node
• flow_func (function) – A function for computing the maximum flow among a pair of
nodes. The function has to accept at least three parameters: a Digraph, a source node,
and a target node. And return a residual network that follows NetworkX conventions (see
maximum_flow() for details). If flow_func is None, the default maximum flow function
(edmonds_karp()) is used. See below for details. The choice of the default function
may change from version to version and should not be relied on. Default value: None.
• auxiliary (NetworkX DiGraph) – Auxiliary digraph for computing flow based edge connec-
tivity. If provided it will be reused instead of recreated. Default value: None.
1 Abdol-Hossein Esfahanian. Connectivity Algorithms. http://www.cse.msu.edu/~cse835/Papers/Graph_connectivity_revised.pdf
1 Abdol-Hossein Esfahanian. Connectivity Algorithms. http://www.cse.msu.edu/~cse835/Papers/Graph_connectivity_revised.pdf

268 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• residual (NetworkX DiGraph) – Residual network to compute maximum flow. If provided


it will be reused instead of recreated. Default value: None.
• cutoff (integer, float) – If specified, the maximum flow algorithm will terminate when the
flow value reaches or exceeds the cutoff. This is only for the algorithms that support the
cutoff parameter: edmonds_karp() and shortest_augmenting_path(). Other
algorithms will ignore this parameter. Default value: None.
Returns K – local edge connectivity for nodes s and t.
Return type integer

Examples

This function is not imported in the base NetworkX namespace, so you have to explicitly import it from the
connectivity package:

>>> from networkx.algorithms.connectivity import local_edge_connectivity

We use in this example the platonic icosahedral graph, which has edge connectivity 5.

>>> G = nx.icosahedral_graph()
>>> local_edge_connectivity(G, 0, 6)
5

If you need to compute local connectivity on several pairs of nodes in the same graph, it is recommended that
you reuse the data structures that NetworkX uses in the computation: the auxiliary digraph for edge connectivity,
and the residual network for the underlying maximum flow computation.
Example of how to compute local edge connectivity among all pairs of nodes of the platonic icosahedral graph
reusing the data structures.

>>> import itertools


>>> # You also have to explicitly import the function for
>>> # building the auxiliary digraph from the connectivity package
>>> from networkx.algorithms.connectivity import (
... build_auxiliary_edge_connectivity)
>>> H = build_auxiliary_edge_connectivity(G)
>>> # And the function for building the residual network from the
>>> # flow package
>>> from networkx.algorithms.flow import build_residual_network
>>> # Note that the auxiliary digraph has an edge attribute named capacity
>>> R = build_residual_network(H, 'capacity')
>>> result = dict.fromkeys(G, dict())
>>> # Reuse the auxiliary digraph and the residual network by passing them
>>> # as parameters
>>> for u, v in itertools.combinations(G, 2):
... k = local_edge_connectivity(G, u, v, auxiliary=H, residual=R)
... result[u][v] = k
>>> all(result[u][v] == 5 for u, v in itertools.combinations(G, 2))
True

You can also use alternative flow algorithms for computing edge connectivity. For instance, in dense
networks the algorithm shortest_augmenting_path() will usually perform better than the default
edmonds_karp() which is faster for sparse networks with highly skewed degree distributions. Alternative
flow functions have to be explicitly imported from the flow package.

3.15. Connectivity 269


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> local_edge_connectivity(G, 0, 6, flow_func=shortest_augmenting_path)
5

Notes

This is a flow based implementation of edge connectivity. We compute the maximum flow using, by default, the
edmonds_karp() algorithm on an auxiliary digraph build from the original input graph:
If the input graph is undirected, we replace each edge (u,‘v‘) with two reciprocal arcs (u, v) and (v, u) and
then we set the attribute ‘capacity’ for each arc to 1. If the input graph is directed we simply add the ‘capacity’
attribute. This is an implementation of algorithm 1 in1 .
The maximum flow in the auxiliary network is equal to the local edge connectivity because the value of a
maximum s-t-flow is equal to the capacity of a minimum s-t-cut (Ford and Fulkerson theorem).
See also:
edge_connectivity(), local_node_connectivity(), node_connectivity(),
maximum_flow(), edmonds_karp(), preflow_push(), shortest_augmenting_path()

References

networkx.algorithms.connectivity.connectivity.local_node_connectivity

local_node_connectivity(G, s, t, flow_func=None, auxiliary=None, residual=None, cutoff=None)


Computes local node connectivity for nodes s and t.
Local node connectivity for two non adjacent nodes s and t is the minimum number of nodes that must be
removed (along with their incident edges) to disconnect them.
This is a flow based implementation of node connectivity. We compute the maximum flow on an auxiliary
digraph build from the original input graph (see below for details).
Parameters
• G (NetworkX graph) – Undirected graph
• s (node) – Source node
• t (node) – Target node
• flow_func (function) – A function for computing the maximum flow among a pair of
nodes. The function has to accept at least three parameters: a Digraph, a source node,
and a target node. And return a residual network that follows NetworkX conventions (see
maximum_flow() for details). If flow_func is None, the default maximum flow function
(edmonds_karp()) is used. See below for details. The choice of the default function
may change from version to version and should not be relied on. Default value: None.
• auxiliary (NetworkX DiGraph) – Auxiliary digraph to compute flow based node connectiv-
ity. It has to have a graph attribute called mapping with a dictionary mapping node names
in G and in the auxiliary digraph. If provided it will be reused instead of recreated. Default
value: None.
• residual (NetworkX DiGraph) – Residual network to compute maximum flow. If provided
it will be reused instead of recreated. Default value: None.

270 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• cutoff (integer, float) – If specified, the maximum flow algorithm will terminate when the
flow value reaches or exceeds the cutoff. This is only for the algorithms that support the
cutoff parameter: edmonds_karp() and shortest_augmenting_path(). Other
algorithms will ignore this parameter. Default value: None.
Returns K – local node connectivity for nodes s and t
Return type integer

Examples

This function is not imported in the base NetworkX namespace, so you have to explicitly import it from the
connectivity package:

>>> from networkx.algorithms.connectivity import local_node_connectivity

We use in this example the platonic icosahedral graph, which has node connectivity 5.

>>> G = nx.icosahedral_graph()
>>> local_node_connectivity(G, 0, 6)
5

If you need to compute local connectivity on several pairs of nodes in the same graph, it is recommended that
you reuse the data structures that NetworkX uses in the computation: the auxiliary digraph for node connectivity,
and the residual network for the underlying maximum flow computation.
Example of how to compute local node connectivity among all pairs of nodes of the platonic icosahedral graph
reusing the data structures.

>>> import itertools


>>> # You also have to explicitly import the function for
>>> # building the auxiliary digraph from the connectivity package
>>> from networkx.algorithms.connectivity import (
... build_auxiliary_node_connectivity)
...
>>> H = build_auxiliary_node_connectivity(G)
>>> # And the function for building the residual network from the
>>> # flow package
>>> from networkx.algorithms.flow import build_residual_network
>>> # Note that the auxiliary digraph has an edge attribute named capacity
>>> R = build_residual_network(H, 'capacity')
>>> result = dict.fromkeys(G, dict())
>>> # Reuse the auxiliary digraph and the residual network by passing them
>>> # as parameters
>>> for u, v in itertools.combinations(G, 2):
... k = local_node_connectivity(G, u, v, auxiliary=H, residual=R)
... result[u][v] = k
...
>>> all(result[u][v] == 5 for u, v in itertools.combinations(G, 2))
True

You can also use alternative flow algorithms for computing node connectivity. For instance, in dense
networks the algorithm shortest_augmenting_path() will usually perform better than the default
edmonds_karp() which is faster for sparse networks with highly skewed degree distributions. Alternative
flow functions have to be explicitly imported from the flow package.

3.15. Connectivity 271


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> local_node_connectivity(G, 0, 6, flow_func=shortest_augmenting_path)
5

Notes

This is a flow based implementation of node connectivity. We compute the maximum flow using, by default,
the edmonds_karp() algorithm (see: maximum_flow()) on an auxiliary digraph build from the original
input graph:
For an undirected graph G having n nodes and m edges we derive a directed graph H with 2n nodes and 2m+n
arcs by replacing each original node v with two nodes v_A, v_B linked by an (internal) arc in H. Then for each
edge (u, v) in G we add two arcs (u_B, v_A) and (v_B, u_A) in H. Finally we set the attribute capacity = 1 for
each arc in H1 .
For a directed graph G having n nodes and m arcs we derive a directed graph H with 2n nodes and m+n arcs by
replacing each original node v with two nodes v_A, v_B linked by an (internal) arc (v_A, v_B) in H. Then for
each arc (u, v) in G we add one arc (u_B, v_A) in H. Finally we set the attribute capacity = 1 for each arc in H.
This is equal to the local node connectivity because the value of a maximum s-t-flow is equal to the capacity of
a minimum s-t-cut.
See also:
local_edge_connectivity(), node_connectivity(), minimum_node_cut(),
maximum_flow(), edmonds_karp(), preflow_push(), shortest_augmenting_path()

References

networkx.algorithms.connectivity.connectivity.node_connectivity

node_connectivity(G, s=None, t=None, flow_func=None)


Returns node connectivity for a graph or digraph G.
Node connectivity is equal to the minimum number of nodes that must be removed to disconnect G or render it
trivial. If source and target nodes are provided, this function returns the local node connectivity: the minimum
number of nodes that must be removed to break all paths from source to target in G.
Parameters
• G (NetworkX graph) – Undirected graph
• s (node) – Source node. Optional. Default value: None.
• t (node) – Target node. Optional. Default value: None.
• flow_func (function) – A function for computing the maximum flow among a pair of
nodes. The function has to accept at least three parameters: a Digraph, a source node,
and a target node. And return a residual network that follows NetworkX conventions (see
maximum_flow() for details). If flow_func is None, the default maximum flow function
(edmonds_karp()) is used. See below for details. The choice of the default function
may change from version to version and should not be relied on. Default value: None.
Returns K – Node connectivity of G, or local node connectivity if source and target are provided.
1 Kammer, Frank and Hanjo Taubig. Graph Connectivity. in Brandes and Erlebach, ‘Network Analysis: Methodological Foundations’,

Lecture Notes in Computer Science, Volume 3418, Springer-Verlag, 2005. http://www.informatik.uni-augsburg.de/thi/personen/kammer/Graph_


Connectivity.pdf

272 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type integer

Examples

>>> # Platonic icosahedral graph is 5-node-connected


>>> G = nx.icosahedral_graph()
>>> nx.node_connectivity(G)
5

You can use alternative flow algorithms for the underlying maximum flow computation. In dense net-
works the algorithm shortest_augmenting_path() will usually perform better than the default
edmonds_karp(), which is faster for sparse networks with highly skewed degree distributions. Alternative
flow functions have to be explicitly imported from the flow package.

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> nx.node_connectivity(G, flow_func=shortest_augmenting_path)
5

If you specify a pair of nodes (source and target) as parameters, this function returns the value of local node
connectivity.

>>> nx.node_connectivity(G, 3, 7)
5

If you need to perform several local computations among different pairs of nodes on the same graph,
it is recommended that you reuse the data structures used in the maximum flow computations. See
local_node_connectivity() for details.

Notes

This is a flow based implementation of node connectivity. The algorithm works by solving 𝑂((𝑛 − 𝛿 − 1 + 𝛿(𝛿 −
1)/2)) maximum flow problems on an auxiliary digraph. Where 𝛿 is the minimum degree of G. For details about
the auxiliary digraph and the computation of local node connectivity see local_node_connectivity().
This implementation is based on algorithm 11 in1 .
See also:
local_node_connectivity(), edge_connectivity(), maximum_flow(),
edmonds_karp(), preflow_push(), shortest_augmenting_path()

References

3.15.7 Flow-based Minimum Cuts

Flow based cut algorithms

minimum_edge_cut(G[, s, t, flow_func]) Returns a set of edges of minimum cardinality that dis-


connects G.
minimum_node_cut(G[, s, t, flow_func]) Returns a set of nodes of minimum cardinality that dis-
connects G.
minimum_st_edge_cut(G, s, t[, flow_func, . . . ]) Returns the edges of the cut-set of a minimum (s, t)-cut.
Continued on next page
1 Abdol-Hossein Esfahanian. Connectivity Algorithms. http://www.cse.msu.edu/~cse835/Papers/Graph_connectivity_revised.pdf

3.15. Connectivity 273


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 71 – continued from previous page


minimum_st_node_cut(G, s, t[, flow_func, . . . ]) Returns a set of nodes of minimum cardinality that dis-
connect source from target in G.

networkx.algorithms.connectivity.cuts.minimum_edge_cut

minimum_edge_cut(G, s=None, t=None, flow_func=None)


Returns a set of edges of minimum cardinality that disconnects G.
If source and target nodes are provided, this function returns the set of edges of minimum cardinality that, if
removed, would break all paths among source and target in G. If not, it returns a set of edges of minimum
cardinality that disconnects G.
Parameters
• G (NetworkX graph)
• s (node) – Source node. Optional. Default value: None.
• t (node) – Target node. Optional. Default value: None.
• flow_func (function) – A function for computing the maximum flow among a pair of
nodes. The function has to accept at least three parameters: a Digraph, a source node,
and a target node. And return a residual network that follows NetworkX conventions (see
maximum_flow() for details). If flow_func is None, the default maximum flow function
(edmonds_karp()) is used. See below for details. The choice of the default function
may change from version to version and should not be relied on. Default value: None.
Returns cutset – Set of edges that, if removed, would disconnect G. If source and target nodes are
provided, the set contains the edges that if removed, would destroy all paths between source and
target.
Return type set

Examples

>>> # Platonic icosahedral graph has edge connectivity 5


>>> G = nx.icosahedral_graph()
>>> len(nx.minimum_edge_cut(G))
5

You can use alternative flow algorithms for the underlying maximum flow computation. In dense net-
works the algorithm shortest_augmenting_path() will usually perform better than the default
edmonds_karp(), which is faster for sparse networks with highly skewed degree distributions. Alternative
flow functions have to be explicitly imported from the flow package.

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> len(nx.minimum_edge_cut(G, flow_func=shortest_augmenting_path))
5

If you specify a pair of nodes (source and target) as parameters, this function returns the value of local edge
connectivity.

>>> nx.edge_connectivity(G, 3, 7)
5

274 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

If you need to perform several local computations among different pairs of nodes on the same graph,
it is recommended that you reuse the data structures used in the maximum flow computations. See
local_edge_connectivity() for details.

Notes

This is a flow based implementation of minimum edge cut. For undirected graphs the algorithm works by
finding a ‘small’ dominating set of nodes of G (see algorithm 7 in1 ) and computing the maximum flow between
an arbitrary node in the dominating set and the rest of nodes in it. This is an implementation of algorithm 6
in1 . For directed graphs, the algorithm does n calls to the max flow function. The function raises an error if the
directed graph is not weakly connected and returns an empty set if it is weakly connected. It is an implementation
of algorithm 8 in1 .
See also:
minimum_st_edge_cut(), minimum_node_cut(), stoer_wagner(),
node_connectivity(), edge_connectivity(), maximum_flow(), edmonds_karp(),
preflow_push(), shortest_augmenting_path()

References

networkx.algorithms.connectivity.cuts.minimum_node_cut

minimum_node_cut(G, s=None, t=None, flow_func=None)


Returns a set of nodes of minimum cardinality that disconnects G.
If source and target nodes are provided, this function returns the set of nodes of minimum cardinality that, if
removed, would destroy all paths among source and target in G. If not, it returns a set of nodes of minimum
cardinality that disconnects G.
Parameters
• G (NetworkX graph)
• s (node) – Source node. Optional. Default value: None.
• t (node) – Target node. Optional. Default value: None.
• flow_func (function) – A function for computing the maximum flow among a pair of
nodes. The function has to accept at least three parameters: a Digraph, a source node,
and a target node. And return a residual network that follows NetworkX conventions (see
maximum_flow() for details). If flow_func is None, the default maximum flow function
(edmonds_karp()) is used. See below for details. The choice of the default function
may change from version to version and should not be relied on. Default value: None.
Returns cutset – Set of nodes that, if removed, would disconnect G. If source and target nodes are
provided, the set contains the nodes that if removed, would destroy all paths between source and
target.
Return type set

Examples

1 Abdol-Hossein Esfahanian. Connectivity Algorithms. http://www.cse.msu.edu/~cse835/Papers/Graph_connectivity_revised.pdf

3.15. Connectivity 275


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> # Platonic icosahedral graph has node connectivity 5


>>> G = nx.icosahedral_graph()
>>> node_cut = nx.minimum_node_cut(G)
>>> len(node_cut)
5

You can use alternative flow algorithms for the underlying maximum flow computation. In dense net-
works the algorithm shortest_augmenting_path() will usually perform better than the default
edmonds_karp(), which is faster for sparse networks with highly skewed degree distributions. Alternative
flow functions have to be explicitly imported from the flow package.

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> node_cut == nx.minimum_node_cut(G, flow_func=shortest_augmenting_path)
True

If you specify a pair of nodes (source and target) as parameters, this function returns a local st node cut.

>>> len(nx.minimum_node_cut(G, 3, 7))


5

If you need to perform several local st cuts among different pairs of nodes on the same graph, it is recommended
that you reuse the data structures used in the maximum flow computations. See minimum_st_node_cut()
for details.

Notes

This is a flow based implementation of minimum node cut. The algorithm is based in solving a number of
maximum flow computations to determine the capacity of the minimum cut on an auxiliary directed network that
corresponds to the minimum node cut of G. It handles both directed and undirected graphs. This implementation
is based on algorithm 11 in1 .
See also:
minimum_st_node_cut(), minimum_cut(), minimum_edge_cut(), stoer_wagner(),
node_connectivity(), edge_connectivity(), maximum_flow(), edmonds_karp(),
preflow_push(), shortest_augmenting_path()

References

networkx.algorithms.connectivity.cuts.minimum_st_edge_cut

minimum_st_edge_cut(G, s, t, flow_func=None, auxiliary=None, residual=None)


Returns the edges of the cut-set of a minimum (s, t)-cut.
This function returns the set of edges of minimum cardinality that, if removed, would destroy all paths among
source and target in G. Edge weights are not considered. See minimum_cut() for computing minimum cuts
considering edge weights.
Parameters
• G (NetworkX graph)
• s (node) – Source node for the flow.
• t (node) – Sink node for the flow.
1 Abdol-Hossein Esfahanian. Connectivity Algorithms. http://www.cse.msu.edu/~cse835/Papers/Graph_connectivity_revised.pdf

276 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• auxiliary (NetworkX DiGraph) – Auxiliary digraph to compute flow based node connectiv-
ity. It has to have a graph attribute called mapping with a dictionary mapping node names
in G and in the auxiliary digraph. If provided it will be reused instead of recreated. Default
value: None.
• flow_func (function) – A function for computing the maximum flow among a pair of
nodes. The function has to accept at least three parameters: a Digraph, a source node,
and a target node. And return a residual network that follows NetworkX conventions (see
maximum_flow() for details). If flow_func is None, the default maximum flow function
(edmonds_karp()) is used. See node_connectivity() for details. The choice of
the default function may change from version to version and should not be relied on. Default
value: None.
• residual (NetworkX DiGraph) – Residual network to compute maximum flow. If provided
it will be reused instead of recreated. Default value: None.
Returns cutset – Set of edges that, if removed from the graph, will disconnect it.
Return type set
See also:
minimum_cut(), minimum_node_cut(), minimum_edge_cut(), stoer_wagner(),
node_connectivity(), edge_connectivity(), maximum_flow(), edmonds_karp(),
preflow_push(), shortest_augmenting_path()

Examples

This function is not imported in the base NetworkX namespace, so you have to explicitly import it from the
connectivity package:
>>> from networkx.algorithms.connectivity import minimum_st_edge_cut

We use in this example the platonic icosahedral graph, which has edge connectivity 5.
>>> G = nx.icosahedral_graph()
>>> len(minimum_st_edge_cut(G, 0, 6))
5

If you need to compute local edge cuts on several pairs of nodes in the same graph, it is recommended that you
reuse the data structures that NetworkX uses in the computation: the auxiliary digraph for edge connectivity,
and the residual network for the underlying maximum flow computation.
Example of how to compute local edge cuts among all pairs of nodes of the platonic icosahedral graph reusing
the data structures.
>>> import itertools
>>> # You also have to explicitly import the function for
>>> # building the auxiliary digraph from the connectivity package
>>> from networkx.algorithms.connectivity import (
... build_auxiliary_edge_connectivity)
>>> H = build_auxiliary_edge_connectivity(G)
>>> # And the function for building the residual network from the
>>> # flow package
>>> from networkx.algorithms.flow import build_residual_network
>>> # Note that the auxiliary digraph has an edge attribute named capacity
>>> R = build_residual_network(H, 'capacity')
>>> result = dict.fromkeys(G, dict())
(continues on next page)

3.15. Connectivity 277


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> # Reuse the auxiliary digraph and the residual network by passing them
>>> # as parameters
>>> for u, v in itertools.combinations(G, 2):
... k = len(minimum_st_edge_cut(G, u, v, auxiliary=H, residual=R))
... result[u][v] = k
>>> all(result[u][v] == 5 for u, v in itertools.combinations(G, 2))
True

You can also use alternative flow algorithms for computing edge cuts. For instance, in dense networks the algo-
rithm shortest_augmenting_path() will usually perform better than the default edmonds_karp()
which is faster for sparse networks with highly skewed degree distributions. Alternative flow functions have to
be explicitly imported from the flow package.

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> len(minimum_st_edge_cut(G, 0, 6, flow_func=shortest_augmenting_path))
5

networkx.algorithms.connectivity.cuts.minimum_st_node_cut

minimum_st_node_cut(G, s, t, flow_func=None, auxiliary=None, residual=None)


Returns a set of nodes of minimum cardinality that disconnect source from target in G.
This function returns the set of nodes of minimum cardinality that, if removed, would destroy all paths among
source and target in G.
Parameters
• G (NetworkX graph)
• s (node) – Source node.
• t (node) – Target node.
• flow_func (function) – A function for computing the maximum flow among a pair of
nodes. The function has to accept at least three parameters: a Digraph, a source node,
and a target node. And return a residual network that follows NetworkX conventions (see
maximum_flow() for details). If flow_func is None, the default maximum flow function
(edmonds_karp()) is used. See below for details. The choice of the default function
may change from version to version and should not be relied on. Default value: None.
• auxiliary (NetworkX DiGraph) – Auxiliary digraph to compute flow based node connectiv-
ity. It has to have a graph attribute called mapping with a dictionary mapping node names
in G and in the auxiliary digraph. If provided it will be reused instead of recreated. Default
value: None.
• residual (NetworkX DiGraph) – Residual network to compute maximum flow. If provided
it will be reused instead of recreated. Default value: None.
Returns cutset – Set of nodes that, if removed, would destroy all paths between source and target
in G.
Return type set

Examples

This function is not imported in the base NetworkX namespace, so you have to explicitly import it from the
connectivity package:

278 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> from networkx.algorithms.connectivity import minimum_st_node_cut

We use in this example the platonic icosahedral graph, which has node connectivity 5.

>>> G = nx.icosahedral_graph()
>>> len(minimum_st_node_cut(G, 0, 6))
5

If you need to compute local st cuts between several pairs of nodes in the same graph, it is recommended that
you reuse the data structures that NetworkX uses in the computation: the auxiliary digraph for node connectivity
and node cuts, and the residual network for the underlying maximum flow computation.
Example of how to compute local st node cuts reusing the data structures:

>>> # You also have to explicitly import the function for


>>> # building the auxiliary digraph from the connectivity package
>>> from networkx.algorithms.connectivity import (
... build_auxiliary_node_connectivity)
>>> H = build_auxiliary_node_connectivity(G)
>>> # And the function for building the residual network from the
>>> # flow package
>>> from networkx.algorithms.flow import build_residual_network
>>> # Note that the auxiliary digraph has an edge attribute named capacity
>>> R = build_residual_network(H, 'capacity')
>>> # Reuse the auxiliary digraph and the residual network by passing them
>>> # as parameters
>>> len(minimum_st_node_cut(G, 0, 6, auxiliary=H, residual=R))
5

You can also use alternative flow algorithms for computing minimum st node cuts. For instance, in dense
networks the algorithm shortest_augmenting_path() will usually perform better than the default
edmonds_karp() which is faster for sparse networks with highly skewed degree distributions. Alternative
flow functions have to be explicitly imported from the flow package.

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> len(minimum_st_node_cut(G, 0, 6, flow_func=shortest_augmenting_path))
5

Notes

This is a flow based implementation of minimum node cut. The algorithm is based in solving a number of
maximum flow computations to determine the capacity of the minimum cut on an auxiliary directed network that
corresponds to the minimum node cut of G. It handles both directed and undirected graphs. This implementation
is based on algorithm 11 in1 .
See also:
minimum_node_cut(), minimum_edge_cut(), stoer_wagner(), node_connectivity(),
edge_connectivity(), maximum_flow(), edmonds_karp(), preflow_push(),
shortest_augmenting_path()
1 Abdol-Hossein Esfahanian. Connectivity Algorithms. http://www.cse.msu.edu/~cse835/Papers/Graph_connectivity_revised.pdf

3.15. Connectivity 279


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.15.8 Stoer-Wagner minimum cut

Stoer-Wagner minimum cut algorithm.

stoer_wagner(G[, weight, heap]) Returns the weighted minimum edge cut using the
Stoer-Wagner algorithm.

networkx.algorithms.connectivity.stoerwagner.stoer_wagner

stoer_wagner(G, weight=’weight’, heap=<class ’networkx.utils.heaps.BinaryHeap’>)


Returns the weighted minimum edge cut using the Stoer-Wagner algorithm.
Determine the minimum edge cut of a connected graph using the Stoer-Wagner algorithm. In weighted cases,
all weights must be nonnegative.
The running time of the algorithm depends on the type of heaps used:

Type of heap Running time


Binary heap 𝑂(𝑛(𝑚 + 𝑛) log 𝑛)
Fibonacci heap 𝑂(𝑛𝑚√
+ 𝑛2 log 𝑛)
Pairing heap 𝑂(22 log log 𝑛 𝑛𝑚 + 𝑛2 log 𝑛)

Parameters
• G (NetworkX graph) – Edges of the graph are expected to have an attribute named by the
weight parameter below. If this attribute is not present, the edge is considered to have unit
weight.
• weight (string) – Name of the weight attribute of the edges. If the attribute is not present,
unit weight is assumed. Default value: ‘weight’.
• heap (class) – Type of heap to be used in the algorithm. It should be a subclass of MinHeap
or implement a compatible interface.
If a stock heap implementation is to be used, BinaryHeap is recommended over
PairingHeap for Python implementations without optimized attribute accesses (e.g.,
CPython) despite a slower asymptotic running time. For Python implementations with opti-
mized attribute accesses (e.g., PyPy), PairingHeap provides better performance. Default
value: BinaryHeap.
Returns
• cut_value (integer or float) – The sum of weights of edges in a minimum cut.
• partition (pair of node lists) – A partitioning of the nodes that defines a minimum cut.
Raises
• NetworkXNotImplemented – If the graph is directed or a multigraph.
• NetworkXError – If the graph has less than two nodes, is not connected or has a negative-
weighted edge.

280 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.Graph()
>>> G.add_edge('x', 'a', weight=3)
>>> G.add_edge('x', 'b', weight=1)
>>> G.add_edge('a', 'c', weight=3)
>>> G.add_edge('b', 'c', weight=5)
>>> G.add_edge('b', 'd', weight=4)
>>> G.add_edge('d', 'e', weight=2)
>>> G.add_edge('c', 'y', weight=2)
>>> G.add_edge('e', 'y', weight=3)
>>> cut_value, partition = nx.stoer_wagner(G)
>>> cut_value
4

3.15.9 Utils for flow-based connectivity

Utilities for connectivity package

build_auxiliary_edge_connectivity(G) Auxiliary digraph for computing flow based edge con-


nectivity
build_auxiliary_node_connectivity(G) Creates a directed graph D from an undirected graph G
to compute flow based node connectivity.

networkx.algorithms.connectivity.utils.build_auxiliary_edge_connectivity

build_auxiliary_edge_connectivity(G)
Auxiliary digraph for computing flow based edge connectivity
If the input graph is undirected, we replace each edge (u,‘v‘) with two reciprocal arcs (u, v) and (v, u) and
then we set the attribute ‘capacity’ for each arc to 1. If the input graph is directed we simply add the ‘capacity’
attribute. Part of algorithm 1 in1 .

References

networkx.algorithms.connectivity.utils.build_auxiliary_node_connectivity

build_auxiliary_node_connectivity(G)
Creates a directed graph D from an undirected graph G to compute flow based node connectivity.
For an undirected graph G having n nodes and m edges we derive a directed graph D with 2n nodes and 2m+n
arcs by replacing each original node v with two nodes vA, vB linked by an (internal) arc in D. Then for each
edge (u, v) in G we add two arcs (uB, vA) and (vB, uA) in D. Finally we set the attribute capacity = 1 for each
arc in D1 .
For a directed graph having n nodes and m arcs we derive a directed graph D with 2n nodes and m+n arcs by
replacing each original node v with two nodes vA, vB linked by an (internal) arc (vA, vB) in D. Then for each
arc (u, v) in G we add one arc (uB, vA) in D. Finally we set the attribute capacity = 1 for each arc in D.
1 Abdol-Hossein Esfahanian. Connectivity Algorithms. (this is a chapter, look for the reference of the book). http://www.cse.msu.edu/~cse835/

Papers/Graph_connectivity_revised.pdf
1 Kammer, Frank and Hanjo Taubig. Graph Connectivity. in Brandes and Erlebach, ‘Network Analysis: Methodological Foundations’,

Lecture Notes in Computer Science, Volume 3418, Springer-Verlag, 2005. http://www.informatik.uni-augsburg.de/thi/personen/kammer/Graph_


Connectivity.pdf

3.15. Connectivity 281


NetworkX Reference, Release 2.3rc1.dev20190222214247

A dictionary with a mapping between nodes in the original graph and the auxiliary digraph is stored as a graph
attribute: H.graph[‘mapping’].

References

3.16 Cores

Find the k-cores of a graph.


The k-core is found by recursively pruning nodes with degrees less than k.
See the following references for details:
An O(m) Algorithm for Cores Decomposition of Networks Vladimir Batagelj and Matjaz Zaversnik, 2003. https:
//arxiv.org/abs/cs.DS/0310049
Generalized Cores Vladimir Batagelj and Matjaz Zaversnik, 2002. https://arxiv.org/pdf/cs/0202039
For directed graphs a more general notion is that of D-cores which looks at (k, l) restrictions on (in, out) degree. The
(k, k) D-core is the k-core.
D-cores: Measuring Collaboration of Directed Graphs Based on Degeneracy Christos Giatsidis, Dimitrios M. Thilikos,
Michalis Vazirgiannis, ICDM 2011. http://www.graphdegeneracy.org/dcores_ICDM_2011.pdf

core_number(G) Returns the core number for each vertex.


k_core(G[, k, core_number]) Returns the k-core of G.
k_shell(G[, k, core_number]) Returns the k-shell of G.
k_crust(G[, k, core_number]) Returns the k-crust of G.
k_corona(G, k[, core_number]) Returns the k-corona of G.

3.16.1 networkx.algorithms.core.core_number

core_number(G)
Returns the core number for each vertex.
A k-core is a maximal subgraph that contains nodes of degree k or more.
The core number of a node is the largest value k of a k-core containing that node.
Parameters G (NetworkX graph) – A graph or directed graph
Returns core_number – A dictionary keyed by node to the core number.
Return type dictionary
Raises NetworkXError – The k-core is not implemented for graphs with self loops or parallel
edges.

Notes

Not implemented for graphs with parallel edges or self loops.


For directed graphs the node degree is defined to be the in-degree + out-degree.

282 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.16.2 networkx.algorithms.core.k_core

k_core(G, k=None, core_number=None)


Returns the k-core of G.
A k-core is a maximal subgraph that contains nodes of degree k or more.
Parameters
• G (NetworkX graph) – A graph or directed graph
• k (int, optional) – The order of the core. If not specified return the main core.
• core_number (dictionary, optional) – Precomputed core numbers for the graph G.
Returns G – The k-core subgraph
Return type NetworkX graph
Raises NetworkXError – The k-core is not defined for graphs with self loops or parallel edges.

Notes

The main core is the core with the largest degree.


Not implemented for graphs with parallel edges or self loops.
For directed graphs the node degree is defined to be the in-degree + out-degree.
Graph, node, and edge attributes are copied to the subgraph.
See also:
core_number()

References

3.16.3 networkx.algorithms.core.k_shell

k_shell(G, k=None, core_number=None)


Returns the k-shell of G.
The k-shell is the subgraph induced by nodes with core number k. That is, nodes in the k-core that are not in the
(k+1)-core.
Parameters
• G (NetworkX graph) – A graph or directed graph.
• k (int, optional) – The order of the shell. If not specified return the outer shell.
• core_number (dictionary, optional) – Precomputed core numbers for the graph G.
Returns G – The k-shell subgraph
Return type NetworkX graph
Raises NetworkXError – The k-shell is not implemented for graphs with self loops or parallel
edges.

3.16. Cores 283


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This is similar to k_corona but in that case only neighbors in the k-core are considered.
Not implemented for graphs with parallel edges or self loops.
For directed graphs the node degree is defined to be the in-degree + out-degree.
Graph, node, and edge attributes are copied to the subgraph.
See also:
core_number(), k_corona()

References

3.16.4 networkx.algorithms.core.k_crust

k_crust(G, k=None, core_number=None)


Returns the k-crust of G.
The k-crust is the graph G with the k-core removed.
Parameters
• G (NetworkX graph) – A graph or directed graph.
• k (int, optional) – The order of the shell. If not specified return the main crust.
• core_number (dictionary, optional) – Precomputed core numbers for the graph G.
Returns G – The k-crust subgraph
Return type NetworkX graph
Raises NetworkXError – The k-crust is not implemented for graphs with self loops or parallel
edges.

Notes

This definition of k-crust is different than the definition in1 . The k-crust in1 is equivalent to the k+1 crust of this
algorithm.
Not implemented for graphs with parallel edges or self loops.
For directed graphs the node degree is defined to be the in-degree + out-degree.
Graph, node, and edge attributes are copied to the subgraph.
See also:
core_number()
1 A model of Internet topology using k-shell decomposition Shai Carmi, Shlomo Havlin, Scott Kirkpatrick, Yuval Shavitt, and Eran Shir, PNAS

July 3, 2007 vol. 104 no. 27 11150-11154 http://www.pnas.org/content/104/27/11150.full

284 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.16.5 networkx.algorithms.core.k_corona

k_corona(G, k, core_number=None)
Returns the k-corona of G.
The k-corona is the subgraph of nodes in the k-core which have exactly k neighbours in the k-core.
Parameters
• G (NetworkX graph) – A graph or directed graph
• k (int) – The order of the corona.
• core_number (dictionary, optional) – Precomputed core numbers for the graph G.
Returns G – The k-corona subgraph
Return type NetworkX graph
Raises NetworkXError – The k-cornoa is not defined for graphs with self loops or parallel edges.

Notes

Not implemented for graphs with parallel edges or self loops.


For directed graphs the node degree is defined to be the in-degree + out-degree.
Graph, node, and edge attributes are copied to the subgraph.
See also:
core_number()

References

3.17 Covering

Functions related to graph covers.

min_edge_cover(G[, matching_algorithm]) Returns a set of edges which constitutes the minimum


edge cover of the graph.
is_edge_cover(G, cover) Decides whether a set of edges is a valid edge cover of
the graph.

3.17.1 networkx.algorithms.covering.min_edge_cover

min_edge_cover(G, matching_algorithm=None)
Returns a set of edges which constitutes the minimum edge cover of the graph.
A smallest edge cover can be found in polynomial time by finding a maximum matching and extending it
greedily so that all nodes are covered.
Parameters
• G (NetworkX graph) – An undirected bipartite graph.

3.17. Covering 285


NetworkX Reference, Release 2.3rc1.dev20190222214247

• matching_algorithm (function) – A function that returns a maximum cardinality matching


in a given bipartite graph. The function must take one input, the graph G, and return a dic-
tionary mapping each node to its mate. If not specified, hopcroft_karp_matching()
will be used. Other possibilities include eppstein_matching(), or matching algo-
rithms in the networkx.algorithms.matching module.
Returns min_cover – It contains all the edges of minimum edge cover in form of tuples. It contains
both the edges (u, v) and (v, u) for given nodes u and v among the edges of minimum
edge cover.
Return type set

Notes

An edge cover of a graph is a set of edges such that every node of the graph is incident to at least one edge of
the set. The minimum edge cover is an edge covering of smallest cardinality.
Due to its implementation, the worst-case running time of this algorithm is bounded by the worst-case running
time of the function matching_algorithm.
Minimum edge cover for bipartite graph can also be found using the function present in networkx.
algorithms.bipartite.covering

3.17.2 networkx.algorithms.covering.is_edge_cover

is_edge_cover(G, cover)
Decides whether a set of edges is a valid edge cover of the graph.
Given a set of edges, whether it is an edge covering can be decided if we just check whether all nodes of the
graph has an edge from the set, incident on it.
Parameters
• G (NetworkX graph) – An undirected bipartite graph.
• cover (set) – Set of edges to be checked.
Returns Whether the set of edges is a valid edge cover of the graph.
Return type bool

Notes

An edge cover of a graph is a set of edges such that every node of the graph is incident to at least one edge of
the set.

3.18 Cycles

3.18.1 Cycle finding algorithms

cycle_basis(G[, root]) Returns a list of cycles which form a basis for cycles of
G.
Continued on next page

286 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 76 – continued from previous page


simple_cycles(G) Find simple cycles (elementary circuits) of a directed
graph.
find_cycle(G[, source, orientation]) Returns a cycle found via depth-first traversal.
minimum_cycle_basis(G[, weight]) Returns a minimum weight cycle basis for G

3.18.2 networkx.algorithms.cycles.cycle_basis

cycle_basis(G, root=None)
Returns a list of cycles which form a basis for cycles of G.
A basis for cycles of a network is a minimal collection of cycles such that any cycle in the network can be written
as a sum of cycles in the basis. Here summation of cycles is defined as “exclusive or” of the edges. Cycle bases
are useful, e.g. when deriving equations for electric circuits using Kirchhoff’s Laws.
Parameters
• G (NetworkX Graph)
• root (node, optional) – Specify starting node for basis.
Returns
• A list of cycle lists. Each cycle list is a list of nodes
• which forms a cycle (loop) in G.

Examples

>>> G = nx.Graph()
>>> nx.add_cycle(G, [0, 1, 2, 3])
>>> nx.add_cycle(G, [0, 3, 4, 5])
>>> print(nx.cycle_basis(G, 0))
[[3, 4, 5, 0], [1, 2, 3, 0]]

Notes

This is adapted from algorithm CACM 4911 .

References

See also:
simple_cycles()

3.18.3 networkx.algorithms.cycles.simple_cycles

simple_cycles(G)
Find simple cycles (elementary circuits) of a directed graph.
A simple cycle, or elementary circuit, is a closed path where no node appears twice. Two elemen-
tary circuits are distinct if they are not cyclic permutations of each other.
1 Paton, K. An algorithm for finding a fundamental set of cycles of a graph. Comm. ACM 12, 9 (Sept 1969), 514-518.

3.18. Cycles 287


NetworkX Reference, Release 2.3rc1.dev20190222214247

This is a nonrecursive, iterator/generator version of Johnson’s algorithm1 . There may be better algorithms for
some cases23 .
Parameters G (NetworkX DiGraph) – A directed graph
Returns cycle_generator – A generator that produces elementary cycles of the graph. Each cycle
is represented by a list of nodes along the cycle.
Return type generator

Examples

>>> edges = [(0, 0), (0, 1), (0, 2), (1, 2), (2, 0), (2, 1), (2, 2)]
>>> G = nx.DiGraph(edges)
>>> len(list(nx.simple_cycles(G)))
5

To filter the cycles so that they don’t include certain nodes or edges, copy your graph and eliminate those nodes
or edges before calling

>>> copyG = G.copy()


>>> copyG.remove_nodes_from([1])
>>> copyG.remove_edges_from([(0, 1)])
>>> len(list(nx.simple_cycles(copyG)))
3

Notes

The implementation follows pp. 79-80 in1 .


The time complexity is 𝑂((𝑛 + 𝑒)(𝑐 + 1)) for 𝑛 nodes, 𝑒 edges and 𝑐 elementary circuits.

References

See also:
cycle_basis()

3.18.4 networkx.algorithms.cycles.find_cycle

find_cycle(G, source=None, orientation=None)


Returns a cycle found via depth-first traversal.
The cycle is a list of edges indicating the cyclic path. Orientation of directed edges is controlled by
orientation.
Parameters
• G (graph) – A directed/undirected graph/multigraph.
1 Finding all the elementary circuits of a directed graph. D. B. Johnson, SIAM Journal on Computing 4, no. 1, 77-84, 1975. https://doi.org/10.

1137/0204007
2 Enumerating the cycles of a digraph: a new preprocessing strategy. G. Loizou and P. Thanish, Information Sciences, v. 27, 163-182, 1982.
3 A search strategy for the elementary cycles of a directed graph. J.L. Szwarcfiter and P.E. Lauer, BIT NUMERICAL MATHEMATICS, v. 16,

no. 2, 192-204, 1976.

288 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• source (node, list of nodes) – The node from which the traversal begins. If None, then a
source is chosen arbitrarily and repeatedly until all edges from each node in the graph are
searched.
• orientation (None | ‘original’ | ‘reverse’ | ‘ignore’ (default: None)) – For directed graphs
and directed multigraphs, edge traversals need not respect the original orientation of the
edges. When set to ‘reverse’ every edge is traversed in the reverse direction. When set to
‘ignore’, every edge is treated as undirected. When set to ‘original’, every edge is treated as
directed. In all three cases, the yielded edge tuples add a last entry to indicate the direction
in which that edge was traversed. If orientation is None, the yielded edge has no direction
indicated. The direction is respected, but not reported.
Returns edges – A list of directed edges indicating the path taken for the loop. If no cycle is found,
then an exception is raised. For graphs, an edge is of the form (u, v) where u and v are the
tail and head of the edge as determined by the traversal. For multigraphs, an edge is of the form
(u, v, key), where key is the key of the edge. When the graph is directed, then u and v
are always in the order of the actual directed edge. If orientation is not None then the edge tuple
is extended to include the direction of traversal (‘forward’ or ‘reverse’) on that edge.
Return type directed edges
Raises NetworkXNoCycle – If no cycle was found.

Examples

In this example, we construct a DAG and find, in the first call, that there are no directed cycles, and so an
exception is raised. In the second call, we ignore edge orientations and find that there is an undirected cycle.
Note that the second call finds a directed cycle while effectively traversing an undirected graph, and so, we
found an “undirected cycle”. This means that this DAG structure does not form a directed tree (which is also
known as a polytree).

>>> import networkx as nx


>>> G = nx.DiGraph([(0, 1), (0, 2), (1, 2)])
>>> try:
... nx.find_cycle(G, orientation='original')
... except:
... pass
...
>>> list(nx.find_cycle(G, orientation='ignore'))
[(0, 1, 'forward'), (1, 2, 'forward'), (0, 2, 'reverse')]

3.18.5 networkx.algorithms.cycles.minimum_cycle_basis

minimum_cycle_basis(G, weight=None)
Returns a minimum weight cycle basis for G
Minimum weight means a cycle basis for which the total weight (length for unweighted graphs) of all the cycles
is minimum.
Parameters
• G (NetworkX Graph)
• weight (string) – name of the edge attribute to use for edge weights
Returns
• A list of cycle lists. Each cycle list is a list of nodes

3.18. Cycles 289


NetworkX Reference, Release 2.3rc1.dev20190222214247

• which forms a cycle (loop) in G. Note that the nodes are not
• necessarily returned in a order by which they appear in the cycle

Examples

>>> G=nx.Graph()
>>> G.add_cycle([0,1,2,3])
>>> G.add_cycle([0,3,4,5])
>>> print(nx.minimum_cycle_basis(G))
[[0, 1, 2, 3], [0, 3, 4, 5]]

References

[1] Kavitha, Telikepalli, et al. “An O(m^2n) Algorithm for Minimum Cycle Basis of Graphs.” http://link.
springer.com/article/10.1007/s00453-007-9064-z [2] de Pina, J. 1995. Applications of shortest path methods.
Ph.D. thesis, University of Amsterdam, Netherlands
See also:
simple_cycles(), cycle_basis()

3.19 Cuts

Functions for finding and evaluating cuts in a graph.

boundary_expansion(G, S) Returns the boundary expansion of the set S.


conductance(G, S[, T, weight]) Returns the conductance of two sets of nodes.
cut_size(G, S[, T, weight]) Returns the size of the cut between two sets of nodes.
edge_expansion(G, S[, T, weight]) Returns the edge expansion between two node sets.
mixing_expansion(G, S[, T, weight]) Returns the mixing expansion between two node sets.
node_expansion(G, S) Returns the node expansion of the set S.
normalized_cut_size(G, S[, T, weight]) Returns the normalized size of the cut between two sets
of nodes.
volume(G, S[, weight]) Returns the volume of a set of nodes.

3.19.1 networkx.algorithms.cuts.boundary_expansion

boundary_expansion(G, S)
Returns the boundary expansion of the set S.
The boundary expansion is the quotient of the size of the edge boundary and the cardinality of S. [1]
Parameters
• G (NetworkX graph)
• S (sequence) – A sequence of nodes in G.
Returns The boundary expansion of the set S.
Return type number
See also:

290 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

edge_expansion(), mixing_expansion(), node_expansion()

References

3.19.2 networkx.algorithms.cuts.conductance

conductance(G, S, T=None, weight=None)


Returns the conductance of two sets of nodes.
The conductance is the quotient of the cut size and the smaller of the volumes of the two sets. [1]
Parameters
• G (NetworkX graph)
• S (sequence) – A sequence of nodes in G.
• T (sequence) – A sequence of nodes in G.
• weight (object) – Edge attribute key to use as weight. If not specified, edges have weight
one.
Returns The conductance between the two sets S and T.
Return type number
See also:
cut_size(), edge_expansion(), normalized_cut_size(), volume()

References

3.19.3 networkx.algorithms.cuts.cut_size

cut_size(G, S, T=None, weight=None)


Returns the size of the cut between two sets of nodes.
A cut is a partition of the nodes of a graph into two sets. The cut size is the sum of the weights of the edges
“between” the two sets of nodes.
Parameters
• G (NetworkX graph)
• S (sequence) – A sequence of nodes in G.
• T (sequence) – A sequence of nodes in G. If not specified, this is taken to be the set comple-
ment of S.
• weight (object) – Edge attribute key to use as weight. If not specified, edges have weight
one.
Returns Total weight of all edges from nodes in set S to nodes in set T (and, in the case of directed
graphs, all edges from nodes in T to nodes in S).
Return type number

3.19. Cuts 291


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

In the graph with two cliques joined by a single edges, the natural bipartition of the graph into two blocks, one
for each clique, yields a cut of weight one:

>>> G = nx.barbell_graph(3, 0)
>>> S = {0, 1, 2}
>>> T = {3, 4, 5}
>>> nx.cut_size(G, S, T)
1

Each parallel edge in a multigraph is counted when determining the cut size:

>>> G = nx.MultiGraph(['ab', 'ab'])


>>> S = {'a'}
>>> T = {'b'}
>>> nx.cut_size(G, S, T)
2

Notes

In a multigraph, the cut size is the total weight of edges including multiplicity.

3.19.4 networkx.algorithms.cuts.edge_expansion

edge_expansion(G, S, T=None, weight=None)


Returns the edge expansion between two node sets.
The edge expansion is the quotient of the cut size and the smaller of the cardinalities of the two sets. [1]
Parameters
• G (NetworkX graph)
• S (sequence) – A sequence of nodes in G.
• T (sequence) – A sequence of nodes in G.
• weight (object) – Edge attribute key to use as weight. If not specified, edges have weight
one.
Returns The edge expansion between the two sets S and T.
Return type number
See also:
boundary_expansion(), mixing_expansion(), node_expansion()

References

3.19.5 networkx.algorithms.cuts.mixing_expansion

mixing_expansion(G, S, T=None, weight=None)


Returns the mixing expansion between two node sets.
The mixing expansion is the quotient of the cut size and twice the number of edges in the graph. [1]

292 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (NetworkX graph)
• S (sequence) – A sequence of nodes in G.
• T (sequence) – A sequence of nodes in G.
• weight (object) – Edge attribute key to use as weight. If not specified, edges have weight
one.
Returns The mixing expansion between the two sets S and T.
Return type number
See also:
boundary_expansion(), edge_expansion(), node_expansion()

References

3.19.6 networkx.algorithms.cuts.node_expansion

node_expansion(G, S)
Returns the node expansion of the set S.
The node expansion is the quotient of the size of the node boundary of S and the cardinality of S. [1]
Parameters
• G (NetworkX graph)
• S (sequence) – A sequence of nodes in G.
Returns The node expansion of the set S.
Return type number
See also:
boundary_expansion(), edge_expansion(), mixing_expansion()

References

3.19.7 networkx.algorithms.cuts.normalized_cut_size

normalized_cut_size(G, S, T=None, weight=None)


Returns the normalized size of the cut between two sets of nodes.
The normalized cut size is the cut size times the sum of the reciprocal sizes of the volumes of the two sets. [1]
Parameters
• G (NetworkX graph)
• S (sequence) – A sequence of nodes in G.
• T (sequence) – A sequence of nodes in G.
• weight (object) – Edge attribute key to use as weight. If not specified, edges have weight
one.
Returns The normalized cut size between the two sets S and T.

3.19. Cuts 293


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type number

Notes

In a multigraph, the cut size is the total weight of edges including multiplicity.
See also:
conductance(), cut_size(), edge_expansion(), volume()

References

3.19.8 networkx.algorithms.cuts.volume

volume(G, S, weight=None)
Returns the volume of a set of nodes.
The volume of a set S is the sum of the (out-)degrees of nodes in S (taking into account parallel edges in
multigraphs). [1]
Parameters
• G (NetworkX graph)
• S (sequence) – A sequence of nodes in G.
• weight (object) – Edge attribute key to use as weight. If not specified, edges have weight
one.
Returns The volume of the set of nodes represented by S in the graph G.
Return type number
See also:
conductance(), cut_size(), edge_expansion(), edge_boundary(),
normalized_cut_size()

References

3.20 Directed Acyclic Graphs

Algorithms for directed acyclic graphs (DAGs).


Note that most of these functions are only guaranteed to work for DAGs. In general, these functions do not check for
acyclic-ness, so it is up to the user to check for that.

ancestors(G, source) Returns all nodes having a path to source in G.


descendants(G, source) Returns all nodes reachable from source in G.
topological_sort(G) Returns a generator of nodes in topologically sorted or-
der.
all_topological_sorts(G) Returns a generator of _all_ topological sorts of the di-
rected graph G.
lexicographical_topological_sort(G[, Returns a generator of nodes in lexicographically topo-
key]) logically sorted order.
Continued on next page

294 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 78 – continued from previous page


is_directed_acyclic_graph(G) Returns True if the graph G is a directed acyclic graph
(DAG) or False if not.
is_aperiodic(G) Returns True if G is aperiodic.
transitive_closure(G) Returns transitive closure of a directed graph
transitive_reduction(G) Returns transitive reduction of a directed graph
antichains(G) Generates antichains from a directed acyclic graph
(DAG).
dag_longest_path(G[, weight, default_weight]) Returns the longest path in a directed acyclic graph
(DAG).
dag_longest_path_length(G[, weight, . . . ]) Returns the longest path length in a DAG
dag_to_branching(G) Returns a branching representing all (overlapping) paths
from root nodes to leaf nodes in the given directed
acyclic graph.

3.20.1 networkx.algorithms.dag.ancestors

ancestors(G, source)
Returns all nodes having a path to source in G.
Parameters
• G (NetworkX DiGraph) – A directed acyclic graph (DAG)
• source (node in G)
Returns The ancestors of source in G
Return type set()

3.20.2 networkx.algorithms.dag.descendants

descendants(G, source)
Returns all nodes reachable from source in G.
Parameters
• G (NetworkX DiGraph) – A directed acyclic graph (DAG)
• source (node in G)
Returns The descendants of source in G
Return type set()

3.20.3 networkx.algorithms.dag.topological_sort

topological_sort(G)
Returns a generator of nodes in topologically sorted order.
A topological sort is a nonunique permutation of the nodes such that an edge from u to v implies that u appears
before v in the topological sort order.
Parameters G (NetworkX digraph) – A directed acyclic graph (DAG)
Returns An iterable of node names in topological sorted order.
Return type iterable

3.20. Directed Acyclic Graphs 295


NetworkX Reference, Release 2.3rc1.dev20190222214247

Raises
• NetworkXError – Topological sort is defined for directed graphs only. If the graph G is
undirected, a NetworkXError is raised.
• NetworkXUnfeasible – If G is not a directed acyclic graph (DAG) no topological sort
exists and a NetworkXUnfeasible exception is raised. This can also be raised if G is
changed while the returned iterator is being processed
• RuntimeError – If G is changed while the returned iterator is being processed.

Examples

To get the reverse order of the topological sort:

>>> DG = nx.DiGraph([(1, 2), (2, 3)])


>>> list(reversed(list(nx.topological_sort(DG))))
[3, 2, 1]

If your DiGraph naturally has the edges representing tasks/inputs and nodes representing people/processes that
initiate tasks, then topological_sort is not quite what you need. You will have to change the tasks to nodes with
dependence reflected by edges. The result is a kind of topological sort of the edges. This can be done with
networkx.line_graph() as follows:

>>> list(nx.topological_sort(nx.line_graph(DG)))
[(1, 2), (2, 3)]

Notes

This algorithm is based on a description and proof in “Introduction to Algorithms: A Creative Approach”1 .
See also:
is_directed_acyclic_graph(), lexicographical_topological_sort()

References

3.20.4 networkx.algorithms.dag.all_topological_sorts

all_topological_sorts(G)
Returns a generator of _all_ topological sorts of the directed graph G.
A topological sort is a nonunique permutation of the nodes such that an edge from u to v implies that u appears
before v in the topological sort order.
Parameters G (NetworkX DiGraph) – A directed graph
Returns All topological sorts of the digraph G
Return type generator
Raises
• NetworkXNotImplemented – If G is not directed
• NetworkXUnfeasible – If G is not acyclic
1 Manber, U. (1989). Introduction to Algorithms - A Creative Approach. Addison-Wesley.

296 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

To enumerate all topological sorts of directed graph:

>>> DG = nx.DiGraph([(1, 2), (2, 3), (2, 4)])


>>> list(nx.all_topological_sorts(DG))
[[1, 2, 4, 3], [1, 2, 3, 4]]

Notes

Implements an iterative version of the algorithm given in [1].

References

3.20.5 networkx.algorithms.dag.lexicographical_topological_sort

lexicographical_topological_sort(G, key=None)
Returns a generator of nodes in lexicographically topologically sorted order.
A topological sort is a nonunique permutation of the nodes such that an edge from u to v implies that u appears
before v in the topological sort order.
Parameters
• G (NetworkX digraph) – A directed acyclic graph (DAG)
• key (function, optional) – This function maps nodes to keys with which to resolve ambigui-
ties in the sort order. Defaults to the identity function.
Returns An iterable of node names in lexicographical topological sort order.
Return type iterable
Raises
• NetworkXError – Topological sort is defined for directed graphs only. If the graph G is
undirected, a NetworkXError is raised.
• NetworkXUnfeasible – If G is not a directed acyclic graph (DAG) no topological sort
exists and a NetworkXUnfeasible exception is raised. This can also be raised if G is
changed while the returned iterator is being processed
• RuntimeError – If G is changed while the returned iterator is being processed.

Notes

This algorithm is based on a description and proof in “Introduction to Algorithms: A Creative Approach”1 .
See also:
topological_sort()
1 Manber, U. (1989). Introduction to Algorithms - A Creative Approach. Addison-Wesley.

3.20. Directed Acyclic Graphs 297


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.20.6 networkx.algorithms.dag.is_directed_acyclic_graph

is_directed_acyclic_graph(G)
Returns True if the graph G is a directed acyclic graph (DAG) or False if not.
Parameters G (NetworkX graph)
Returns True if G is a DAG, False otherwise
Return type bool

3.20.7 networkx.algorithms.dag.is_aperiodic

is_aperiodic(G)
Returns True if G is aperiodic.
A directed graph is aperiodic if there is no integer k > 1 that divides the length of every cycle in the graph.
Parameters G (NetworkX DiGraph) – A directed graph
Returns True if the graph is aperiodic False otherwise
Return type bool
Raises NetworkXError – If G is not directed

Notes

This uses the method outlined in1 , which runs in 𝑂(𝑚) time given 𝑚 edges in G. Note that a graph is not
aperiodic if it is acyclic as every integer trivial divides length 0 cycles.

References

3.20.8 networkx.algorithms.dag.transitive_closure

transitive_closure(G)
Returns transitive closure of a directed graph
The transitive closure of G = (V,E) is a graph G+ = (V,E+) such that for all v,w in V there is an edge (v,w) in E+
if and only if there is a non-null path from v to w in G.
Parameters G (NetworkX DiGraph) – A directed graph
Returns The transitive closure of G
Return type NetworkX DiGraph
Raises NetworkXNotImplemented – If G is not directed
1 Jarvis, J. P.; Shier, D. R. (1996), “Graph-theoretic analysis of finite Markov chains,” in Shier, D. R.; Wallenius, K. T., Applied Mathematical

Modeling: A Multidisciplinary Approach, CRC Press.

298 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.20.9 networkx.algorithms.dag.transitive_reduction

transitive_reduction(G)
Returns transitive reduction of a directed graph
The transitive reduction of G = (V,E) is a graph G- = (V,E-) such that for all v,w in V there is an edge (v,w) in
E- if and only if (v,w) is in E and there is no path from v to w in G with length greater than 1.
Parameters G (NetworkX DiGraph) – A directed acyclic graph (DAG)
Returns The transitive reduction of G
Return type NetworkX DiGraph
Raises NetworkXError – If G is not a directed acyclic graph (DAG) transitive reduction is not
uniquely defined and a NetworkXError exception is raised.

References

https://en.wikipedia.org/wiki/Transitive_reduction

3.20.10 networkx.algorithms.dag.antichains

antichains(G)
Generates antichains from a directed acyclic graph (DAG).
An antichain is a subset of a partially ordered set such that any two elements in the subset are incomparable.
Parameters G (NetworkX DiGraph) – A directed acyclic graph (DAG)
Returns
Return type generator object
Raises
• NetworkXNotImplemented – If G is not directed
• NetworkXUnfeasible – If G contains a cycle

Notes

This function was originally developed by Peter Jipsen and Franco Saliola for the SAGE project. It’s included
in NetworkX with permission from the authors. Original SAGE code at:
https://github.com/sagemath/sage/blob/master/src/sage/combinat/posets/hasse_diagram.py

References

3.20.11 networkx.algorithms.dag.dag_longest_path

dag_longest_path(G, weight=’weight’, default_weight=1)


Returns the longest path in a directed acyclic graph (DAG).
If G has edges with weight attribute the edge data are used as weight values.

3.20. Directed Acyclic Graphs 299


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (NetworkX DiGraph) – A directed acyclic graph (DAG)
• weight (str, optional) – Edge data key to use for weight
• default_weight (int, optional) – The weight of edges that do not have a weight attribute
Returns Longest path
Return type list
Raises NetworkXNotImplemented – If G is not directed
See also:
dag_longest_path_length()

3.20.12 networkx.algorithms.dag.dag_longest_path_length

dag_longest_path_length(G, weight=’weight’, default_weight=1)


Returns the longest path length in a DAG
Parameters
• G (NetworkX DiGraph) – A directed acyclic graph (DAG)
• weight (string, optional) – Edge data key to use for weight
• default_weight (int, optional) – The weight of edges that do not have a weight attribute
Returns Longest path length
Return type int
Raises NetworkXNotImplemented – If G is not directed
See also:
dag_longest_path()

3.20.13 networkx.algorithms.dag.dag_to_branching

dag_to_branching(G)
Returns a branching representing all (overlapping) paths from root nodes to leaf nodes in the given directed
acyclic graph.
As described in networkx.algorithms.tree.recognition, a branching is a directed forest in which
each node has at most one parent. In other words, a branching is a disjoint union of arborescences. For this
function, each node of in-degree zero in G becomes a root of one of the arborescences, and there will be one leaf
node for each distinct path from that root to a leaf node in G.
Each node v in G with k parents becomes k distinct nodes in the returned branching, one for each parent, and
the sub-DAG rooted at v is duplicated for each copy. The algorithm then recurses on the children of each copy
of v.
Parameters G (NetworkX graph) – A directed acyclic graph.
Returns
The branching in which there is a bijection between root-to-leaf paths in G (in which multiple
paths may share the same leaf) and root-to-leaf paths in the branching (in which there is a unique
path from a root to a leaf).

300 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Each node has an attribute ‘source’ whose value is the original node to which this node corre-
sponds. No other graph, node, or edge attributes are copied into this new graph.
Return type DiGraph
Raises
• NetworkXNotImplemented – If G is not directed, or if G is a multigraph.
• HasACycle – If G is not acyclic.

Examples

To examine which nodes in the returned branching were produced by which original node in the directed acyclic
graph, we can collect the mapping from source node to new nodes into a dictionary. For example, consider the
directed diamond graph:

>>> from collections import defaultdict


>>> from operator import itemgetter
>>>
>>> G = nx.DiGraph(nx.utils.pairwise('abd'))
>>> G.add_edges_from(nx.utils.pairwise('acd'))
>>> B = nx.dag_to_branching(G)
>>>
>>> sources = defaultdict(set)
>>> for v, source in B.nodes(data='source'):
... sources[source].add(v)
>>> len(sources['a'])
1
>>> len(sources['d'])
2

To copy node attributes from the original graph to the new graph, you can use a dictionary like the one con-
structed in the above example:

>>> for source, nodes in sources.items():


... for v in nodes:
... B.node[v].update(G.node[source])

Notes

This function is not idempotent in the sense that the node labels in the returned branching may be uniquely
generated each time the function is invoked. In fact, the node labels may not be integers; in order to relabel
the nodes to be more readable, you can use the networkx.convert_node_labels_to_integers()
function.
The current implementation of this function uses networkx.prefix_tree(), so it is subject to the limita-
tions of that function.

3.21 Dispersion

3.21.1 Dispersion

3.21. Dispersion 301


NetworkX Reference, Release 2.3rc1.dev20190222214247

dispersion(G[, u, v, normalized, alpha, b, c]) Calculate dispersion between u and v in G.

networkx.algorithms.centrality.dispersion

dispersion(G, u=None, v=None, normalized=True, alpha=1.0, b=0.0, c=0.0)


Calculate dispersion between u and v in G.
A link between two actors (u and v) has a high dispersion when their mutual ties (s and t) are not well connected
with each other.
Parameters
• G (graph) – A NetworkX graph.
• u (node, optional) – The source for the dispersion score (e.g. ego node of the network).
• v (node, optional) – The target of the dispersion score if specified.
• normalized (bool) – If True (default) normalize by the embededness of the nodes (u and v).
Returns nodes – If u (v) is specified, returns a dictionary of nodes with dispersion score for all
“target” (“source”) nodes. If neither u nor v is specified, returns a dictionary of dictionaries for
all nodes ‘u’ in the graph with a dispersion score for each node ‘v’.
Return type dictionary

Notes

This implementation follows Lars Backstrom and Jon Kleinberg1 . Typical usage would be to run dispersion on
the ego network 𝐺𝑢 if 𝑢 were specified. Running dispersion() with neither 𝑢 nor 𝑣 specified can take some
time to complete.

References

3.22 Distance Measures

Graph diameter, radius, eccentricity and other properties.

center(G[, e, usebounds]) Returns the center of the graph G.


diameter(G[, e, usebounds]) Returns the diameter of the graph G.
eccentricity(G[, v, sp]) Returns the eccentricity of nodes in G.
extrema_bounding(G[, compute]) Compute requested extreme distance metric of undi-
rected graph G
periphery(G[, e, usebounds]) Returns the periphery of the graph G.
radius(G[, e, usebounds]) Returns the radius of the graph G.

3.22.1 networkx.algorithms.distance_measures.center

center(G, e=None, usebounds=False)


Returns the center of the graph G.
1 Romantic Partnerships and the Dispersion of Social Ties: A Network Analysis of Relationship Status on Facebook. Lars Backstrom, Jon

Kleinberg. https://arxiv.org/pdf/1310.6753v1.pdf

302 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

The center is the set of nodes with eccentricity equal to radius.


Parameters
• G (NetworkX graph) – A graph
• e (eccentricity dictionary, optional) – A precomputed dictionary of eccentricities.
Returns c – List of nodes in center
Return type list

3.22.2 networkx.algorithms.distance_measures.diameter

diameter(G, e=None, usebounds=False)


Returns the diameter of the graph G.
The diameter is the maximum eccentricity.
Parameters
• G (NetworkX graph) – A graph
• e (eccentricity dictionary, optional) – A precomputed dictionary of eccentricities.
Returns d – Diameter of graph
Return type integer
See also:
eccentricity()

3.22.3 networkx.algorithms.distance_measures.eccentricity

eccentricity(G, v=None, sp=None)


Returns the eccentricity of nodes in G.
The eccentricity of a node v is the maximum distance from v to all other nodes in G.
Parameters
• G (NetworkX graph) – A graph
• v (node, optional) – Return value of specified node
• sp (dict of dicts, optional) – All pairs shortest path lengths as a dictionary of dictionaries
Returns ecc – A dictionary of eccentricity values keyed by node.
Return type dictionary

3.22.4 networkx.algorithms.distance_measures.extrema_bounding

extrema_bounding(G, compute=’diameter’)
Compute requested extreme distance metric of undirected graph G
Computation is based on smart lower and upper bounds, and in practice linear in the number of nodes, rather
than quadratic (except for some border cases such as complete graphs or circle shaped graphs).
Parameters
• G (NetworkX graph) – An undirected graph

3.22. Distance Measures 303


NetworkX Reference, Release 2.3rc1.dev20190222214247

• compute (string denoting the requesting metric) – “diameter” for the maximal eccentricity
value, “radius” for the minimal eccentricity value, “periphery” for the set of nodes with
eccentricity equal to the diameter “center” for the set of nodes with eccentricity equal to the
radius
Returns value – int for “diameter” and “radius” or list of nodes for “center” and “periphery”
Return type value of the requested metric
Raises NetworkXError – If the graph consists of multiple components

Notes

This algorithm was proposed in the following papers:


F.W. Takes and W.A. Kosters, Determining the Diameter of Small World Networks, in Proceedings of the 20th
ACM International Conference on Information and Knowledge Management (CIKM 2011), pp. 1191-1196,
2011. doi: https://doi.org/10.1145/2063576.2063748
F.W. Takes and W.A. Kosters, Computing the Eccentricity Distribution of Large Graphs, Algorithms 6(1): 100-
118, 2013. doi: https://doi.org/10.3390/a6010100
M. Borassi, P. Crescenzi, M. Habib, W.A. Kosters, A. Marino and F.W. Takes, Fast Graph Diameter and Radius
BFS-Based Computation in (Weakly Connected) Real-World Graphs, Theoretical Computer Science 586: 59-
80, 2015. doi: https://doi.org/10.1016/j.tcs.2015.02.033

3.22.5 networkx.algorithms.distance_measures.periphery

periphery(G, e=None, usebounds=False)


Returns the periphery of the graph G.
The periphery is the set of nodes with eccentricity equal to the diameter.
Parameters
• G (NetworkX graph) – A graph
• e (eccentricity dictionary, optional) – A precomputed dictionary of eccentricities.
Returns p – List of nodes in periphery
Return type list

3.22.6 networkx.algorithms.distance_measures.radius

radius(G, e=None, usebounds=False)


Returns the radius of the graph G.
The radius is the minimum eccentricity.
Parameters
• G (NetworkX graph) – A graph
• e (eccentricity dictionary, optional) – A precomputed dictionary of eccentricities.
Returns r – Radius of graph
Return type integer

304 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.23 Distance-Regular Graphs

3.23.1 Distance-regular graphs

is_distance_regular(G) Returns True if the graph is distance regular, False oth-


erwise.
is_strongly_regular(G) Returns True if and only if the given graph is strongly
regular.
intersection_array(G) Returns the intersection array of a distance-regular
graph.
global_parameters(b, c) Returns global parameters for a given intersection array.

3.23.2 networkx.algorithms.distance_regular.is_distance_regular

is_distance_regular(G)
Returns True if the graph is distance regular, False otherwise.
A connected graph G is distance-regular if for any nodes x,y and any integers i,j=0,1,. . . ,d (where d is the graph
diameter), the number of vertices at distance i from x and distance j from y depends only on i,j and the graph
distance between x and y, independently of the choice of x and y.
Parameters G (Networkx graph (undirected))
Returns True if the graph is Distance Regular, False otherwise
Return type bool

Examples

>>> G=nx.hypercube_graph(6)
>>> nx.is_distance_regular(G)
True

See also:
intersection_array(), global_parameters()

Notes

For undirected and simple graphs only

References

3.23.3 networkx.algorithms.distance_regular.is_strongly_regular

is_strongly_regular(G)
Returns True if and only if the given graph is strongly regular.
An undirected graph is strongly regular if
• it is regular,
• each pair of adjacent vertices has the same number of neighbors in common,

3.23. Distance-Regular Graphs 305


NetworkX Reference, Release 2.3rc1.dev20190222214247

• each pair of nonadjacent vertices has the same number of neighbors in common.
Each strongly regular graph is a distance-regular graph. Conversely, if a distance-regular graph has di-
ameter two, then it is a strongly regular graph. For more information on distance-regular graphs, see
is_distance_regular().
Parameters G (NetworkX graph) – An undirected graph.
Returns Whether G is strongly regular.
Return type bool

Examples

The cycle graph on five vertices is strongly regular. It is two-regular, each pair of adjacent vertices has no shared
neighbors, and each pair of nonadjacent vertices has one shared neighbor:

>>> import networkx as nx


>>> G = nx.cycle_graph(5)
>>> nx.is_strongly_regular(G)
True

3.23.4 networkx.algorithms.distance_regular.intersection_array

intersection_array(G)
Returns the intersection array of a distance-regular graph.
Given a distance-regular graph G with integers b_i, c_i,i = 0,. . . .,d such that for any 2 vertices x,y in G at a
distance i=d(x,y), there are exactly c_i neighbors of y at a distance of i-1 from x and b_i neighbors of y at a
distance of i+1 from x.
A distance regular graph’s intersection array is given by, [b_0,b_1,. . . ..b_{d-1};c_1,c_2,. . . ..c_d]
Parameters G (Networkx graph (undirected))
Returns b,c
Return type tuple of lists

Examples

>>> G=nx.icosahedral_graph()
>>> nx.intersection_array(G)
([5, 2, 1], [1, 2, 5])

References

See also:
global_parameters()

306 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.23.5 networkx.algorithms.distance_regular.global_parameters

global_parameters(b, c)
Returns global parameters for a given intersection array.
Given a distance-regular graph G with integers b_i, c_i,i = 0,. . . .,d such that for any 2 vertices x,y in G at a
distance i=d(x,y), there are exactly c_i neighbors of y at a distance of i-1 from x and b_i neighbors of y at a
distance of i+1 from x.
Thus, a distance regular graph has the global parameters, [[c_0,a_0,b_0],[c_1,a_1,b_1],. . . . . . ,[c_d,a_d,b_d]] for
the intersection array [b_0,b_1,. . . ..b_{d-1};c_1,c_2,. . . ..c_d] where a_i+b_i+c_i=k , k= degree of every vertex.
Parameters
• b (list)
• c (list)
Returns An iterable over three tuples.
Return type iterable

Examples

>>> G = nx.dodecahedral_graph()
>>> b, c = nx.intersection_array(G)
>>> list(nx.global_parameters(b, c))
[(0, 0, 3), (1, 0, 2), (1, 1, 1), (1, 1, 1), (2, 0, 1), (3, 0, 0)]

References

See also:
intersection_array()

3.24 Dominance

Dominance algorithms.

immediate_dominators(G, start) Returns the immediate dominators of all nodes of a di-


rected graph.
dominance_frontiers(G, start) Returns the dominance frontiers of all nodes of a di-
rected graph.

3.24.1 networkx.algorithms.dominance.immediate_dominators

immediate_dominators(G, start)
Returns the immediate dominators of all nodes of a directed graph.
Parameters
• G (a DiGraph or MultiDiGraph) – The graph where dominance is to be computed.
• start (node) – The start node of dominance computation.

3.24. Dominance 307


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns idom – A dict containing the immediate dominators of each node reachable from start.
Return type dict keyed by nodes
Raises
• NetworkXNotImplemented – If G is undirected.
• NetworkXError – If start is not in G.

Notes

Except for start, the immediate dominators are the parents of their corresponding nodes in the dominator tree.

Examples

>>> G = nx.DiGraph([(1, 2), (1, 3), (2, 5), (3, 4), (4, 5)])
>>> sorted(nx.immediate_dominators(G, 1).items())
[(1, 1), (2, 1), (3, 1), (4, 3), (5, 1)]

References

3.24.2 networkx.algorithms.dominance.dominance_frontiers

dominance_frontiers(G, start)
Returns the dominance frontiers of all nodes of a directed graph.
Parameters
• G (a DiGraph or MultiDiGraph) – The graph where dominance is to be computed.
• start (node) – The start node of dominance computation.
Returns df – A dict containing the dominance frontiers of each node reachable from start as lists.
Return type dict keyed by nodes
Raises
• NetworkXNotImplemented – If G is undirected.
• NetworkXError – If start is not in G.

Examples

>>> G = nx.DiGraph([(1, 2), (1, 3), (2, 5), (3, 4), (4, 5)])
>>> sorted((u, sorted(df)) for u, df in nx.dominance_frontiers(G, 1).items())
[(1, []), (2, [5]), (3, [5]), (4, [5]), (5, [])]

References

3.25 Dominating Sets

Functions for computing dominating sets in a graph.

308 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

dominating_set(G[, start_with]) Finds a dominating set for the graph G.


is_dominating_set(G, nbunch) Checks if nbunch is a dominating set for G.

3.25.1 networkx.algorithms.dominating.dominating_set

dominating_set(G, start_with=None)
Finds a dominating set for the graph G.
A dominating set for a graph with node set V is a subset D of V such that every node not in D is adjacent to at
least one member of D1 .
Parameters
• G (NetworkX graph)
• start_with (node (default=None)) – Node to use as a starting point for the algorithm.
Returns D – A dominating set for G.
Return type set

Notes

This function is an implementation of algorithm 7 in2 which finds some dominating set, not necessarily the
smallest one.
See also:
is_dominating_set()

References

3.25.2 networkx.algorithms.dominating.is_dominating_set

is_dominating_set(G, nbunch)
Checks if nbunch is a dominating set for G.
A dominating set for a graph with node set V is a subset D of V such that every node not in D is adjacent to at
least one member of D1 .
Parameters
• G (NetworkX graph)
• nbunch (iterable) – An iterable of nodes in the graph G.
See also:
dominating_set()
1 https://en.wikipedia.org/wiki/Dominating_set
2 Abdol-Hossein Esfahanian. Connectivity Algorithms. http://www.cse.msu.edu/~cse835/Papers/Graph_connectivity_revised.pdf
1 https://en.wikipedia.org/wiki/Dominating_set

3.25. Dominating Sets 309


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.26 Efficiency

Provides functions for computing the efficiency of nodes and graphs.

efficiency(G, u, v) Returns the efficiency of a pair of nodes in a graph.


local_efficiency(G) Returns the average local efficiency of the graph.
global_efficiency(G) Returns the average global efficiency of the graph.

3.26.1 networkx.algorithms.efficiency.efficiency

efficiency(G, u, v)
Returns the efficiency of a pair of nodes in a graph.
The efficiency of a pair of nodes is the multiplicative inverse of the shortest path distance between the nodes1 .
Returns 0 if no path between nodes.
Parameters
• G (networkx.Graph) – An undirected graph for which to compute the average local
efficiency.
• u, v (node) – Nodes in the graph G.
Returns Multiplicative inverse of the shortest path distance between the nodes.
Return type float

Notes

Edge weights are ignored when computing the shortest path distances.
See also:
local_efficiency(), global_efficiency()

References

3.26.2 networkx.algorithms.efficiency.local_efficiency

local_efficiency(G)
Returns the average local efficiency of the graph.
The efficiency of a pair of nodes in a graph is the multiplicative inverse of the shortest path distance between the
nodes. The local efficiency of a node in the graph is the average global efficiency of the subgraph induced by
the neighbors of the node. The average local efficiency is the average of the local efficiencies of each node1 .
Parameters G (networkx.Graph) – An undirected graph for which to compute the average local
efficiency.
1 Latora, Vito, and Massimo Marchiori. “Efficient behavior of small-world networks.” Physical Review Letters 87.19 (2001): 198701. <https:
//doi.org/10.1103/PhysRevLett.87.198701>
1 Latora, Vito, and Massimo Marchiori. “Efficient behavior of small-world networks.” Physical Review Letters 87.19 (2001): 198701. <https:

//doi.org/10.1103/PhysRevLett.87.198701>

310 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns The average local efficiency of the graph.


Return type float

Notes

Edge weights are ignored when computing the shortest path distances.
See also:
global_efficiency()

References

3.26.3 networkx.algorithms.efficiency.global_efficiency

global_efficiency(G)
Returns the average global efficiency of the graph.
The efficiency of a pair of nodes in a graph is the multiplicative inverse of the shortest path distance between the
nodes. The average global efficiency of a graph is the average efficiency of all pairs of nodes1 .
Parameters G (networkx.Graph) – An undirected graph for which to compute the average
global efficiency.
Returns The average global efficiency of the graph.
Return type float

Notes

Edge weights are ignored when computing the shortest path distances.
See also:
local_efficiency()

References

3.27 Eulerian

Eulerian circuits and graphs.

is_eulerian(G) Returns True if and only if G is Eulerian.


eulerian_circuit(G[, source, keys]) Returns an iterator over the edges of an Eulerian circuit
in G.
eulerize(G) Transforms a graph into an Eulerian graph
1 Latora, Vito, and Massimo Marchiori. “Efficient behavior of small-world networks.” Physical Review Letters 87.19 (2001): 198701. <https:

//doi.org/10.1103/PhysRevLett.87.198701>

3.27. Eulerian 311


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.27.1 networkx.algorithms.euler.is_eulerian

is_eulerian(G)
Returns True if and only if G is Eulerian.
A graph is Eulerian if it has an Eulerian circuit. An Eulerian circuit is a closed walk that includes each edge of
a graph exactly once.
Parameters G (NetworkX graph) – A graph, either directed or undirected.

Examples

>>> nx.is_eulerian(nx.DiGraph({0: [3], 1: [2], 2: [3], 3: [0, 1]}))


True
>>> nx.is_eulerian(nx.complete_graph(5))
True
>>> nx.is_eulerian(nx.petersen_graph())
False

Notes

If the graph is not connected (or not strongly connected, for directed graphs), this function returns False.

3.27.2 networkx.algorithms.euler.eulerian_circuit

eulerian_circuit(G, source=None, keys=False)


Returns an iterator over the edges of an Eulerian circuit in G.
An Eulerian circuit is a closed walk that includes each edge of a graph exactly once.
Parameters
• G (NetworkX graph) – A graph, either directed or undirected.
• source (node, optional) – Starting node for circuit.
• keys (bool) – If False, edges generated by this function will be of the form (u, v). Oth-
erwise, edges will be of the form (u, v, k). This option is ignored unless G is a multi-
graph.
Returns edges – An iterator over edges in the Eulerian circuit.
Return type iterator
Raises NetworkXError – If the graph is not Eulerian.
See also:
is_eulerian()

Notes

This is a linear time implementation of an algorithm adapted from1 .


For general information about Euler tours, see2 .
1 J. Edmonds, E. L. Johnson. Matching, Euler tours and the Chinese postman. Mathematical programming, Volume 5, Issue 1 (1973), 111-114.
2 https://en.wikipedia.org/wiki/Eulerian_path

312 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

Examples

To get an Eulerian circuit in an undirected graph:

>>> G = nx.complete_graph(3)
>>> list(nx.eulerian_circuit(G))
[(0, 2), (2, 1), (1, 0)]
>>> list(nx.eulerian_circuit(G, source=1))
[(1, 2), (2, 0), (0, 1)]

To get the sequence of vertices in an Eulerian circuit:

>>> [u for u, v in nx.eulerian_circuit(G)]


[0, 2, 1]

3.27.3 networkx.algorithms.euler.eulerize

eulerize(G)
Transforms a graph into an Eulerian graph
Parameters G (NetworkX graph) – An undirected graph
Returns G
Return type NetworkX multigraph
Raises NetworkXError – If the graph is not connected.
See also:
is_eulerian(), eulerian_circuit()

References

Examples

>>> G = nx.complete_graph(10)
>>> H = nx.eulerize(G)
>>> nx.is_eulerian(H)
True

3.28 Flows

3.28.1 Maximum Flow

maximum_flow(flowG, _s, _t[, capacity, . . . ]) Find a maximum single-commodity flow.


maximum_flow_value(flowG, _s, _t[, . . . ]) Find the value of maximum single-commodity flow.
minimum_cut(flowG, _s, _t[, capacity, flow_func]) Compute the value and the node partition of a minimum
(s, t)-cut.
minimum_cut_value(flowG, _s, _t[, capacity, . . . ]) Compute the value of a minimum (s, t)-cut.

3.28. Flows 313


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.flow.maximum_flow

maximum_flow(flowG, _s, _t, capacity=’capacity’, flow_func=None, **kwargs)


Find a maximum single-commodity flow.
Parameters
• flowG (NetworkX graph) – Edges of the graph are expected to have an attribute called ‘ca-
pacity’. If this attribute is not present, the edge is considered to have infinite capacity.
• _s (node) – Source node for the flow.
• _t (node) – Sink node for the flow.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• flow_func (function) – A function for computing the maximum flow among a pair of nodes
in a capacitated graph. The function has to accept at least three parameters: a Graph or
Digraph, a source node, and a target node. And return a residual network that follows Net-
workX conventions (see Notes). If flow_func is None, the default maximum flow function
(preflow_push()) is used. See below for alternative algorithms. The choice of the
default function may change from version to version and should not be relied on. Default
value: None.
• kwargs (Any other keyword parameter is passed to the function that) – computes the maxi-
mum flow.
Returns
• flow_value (integer, float) – Value of the maximum flow, i.e., net outflow from the source.
• flow_dict (dict) – A dictionary containing the value of the flow that went through each edge.
Raises
• NetworkXError – The algorithm does not support MultiGraph and MultiDiGraph. If the
input graph is an instance of one of these two classes, a NetworkXError is raised.
• NetworkXUnbounded – If the graph has a path of infinite capacity, the value of a feasible
flow on the graph is unbounded above and the function raises a NetworkXUnbounded.
See also:
maximum_flow_value(), minimum_cut(), minimum_cut_value(), edmonds_karp(),
preflow_push(), shortest_augmenting_path()

Notes

The function used in the flow_func parameter has to return a residual network that follows NetworkX conven-
tions:
The residual network R from an input graph G has the same nodes as G. R is a DiGraph that contains a pair of
edges (u, v) and (v, u) iff (u, v) is not a self-loop, and at least one of (u, v) and (v, u) exists in
G.
For each edge (u, v) in R, R[u][v]['capacity'] is equal to the capacity of (u, v) in G if it exists in
G or zero otherwise. If the capacity is infinite, R[u][v]['capacity'] will have a high arbitrary finite value
that does not affect the solution of the problem. This value is stored in R.graph['inf']. For each edge (u,
v) in R, R[u][v]['flow'] represents the flow function of (u, v) and satisfies R[u][v]['flow']
== -R[v][u]['flow'].

314 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

The flow value, defined as the total flow into t, the sink, is stored in R.graph['flow_value']. Reacha-
bility to t using only edges (u, v) such that R[u][v]['flow'] < R[u][v]['capacity'] induces
a minimum s-t cut.
Specific algorithms may store extra data in R.
The function should supports an optional boolean parameter value_only. When True, it can optionally terminate
the algorithm as soon as the maximum flow value and the minimum cut can be determined.

Examples

>>> import networkx as nx


>>> G = nx.DiGraph()
>>> G.add_edge('x','a', capacity=3.0)
>>> G.add_edge('x','b', capacity=1.0)
>>> G.add_edge('a','c', capacity=3.0)
>>> G.add_edge('b','c', capacity=5.0)
>>> G.add_edge('b','d', capacity=4.0)
>>> G.add_edge('d','e', capacity=2.0)
>>> G.add_edge('c','y', capacity=2.0)
>>> G.add_edge('e','y', capacity=3.0)

maximum_flow returns both the value of the maximum flow and a dictionary with all flows.

>>> flow_value, flow_dict = nx.maximum_flow(G, 'x', 'y')


>>> flow_value
3.0
>>> print(flow_dict['x']['b'])
1.0

You can also use alternative algorithms for computing the maximum flow by using the flow_func parameter.

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> flow_value == nx.maximum_flow(G, 'x', 'y',
... flow_func=shortest_augmenting_path)[0]
True

networkx.algorithms.flow.maximum_flow_value

maximum_flow_value(flowG, _s, _t, capacity=’capacity’, flow_func=None, **kwargs)


Find the value of maximum single-commodity flow.
Parameters
• flowG (NetworkX graph) – Edges of the graph are expected to have an attribute called ‘ca-
pacity’. If this attribute is not present, the edge is considered to have infinite capacity.
• _s (node) – Source node for the flow.
• _t (node) – Sink node for the flow.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• flow_func (function) – A function for computing the maximum flow among a pair of nodes
in a capacitated graph. The function has to accept at least three parameters: a Graph or

3.28. Flows 315


NetworkX Reference, Release 2.3rc1.dev20190222214247

Digraph, a source node, and a target node. And return a residual network that follows Net-
workX conventions (see Notes). If flow_func is None, the default maximum flow function
(preflow_push()) is used. See below for alternative algorithms. The choice of the
default function may change from version to version and should not be relied on. Default
value: None.
• kwargs (Any other keyword parameter is passed to the function that) – computes the maxi-
mum flow.
Returns flow_value – Value of the maximum flow, i.e., net outflow from the source.
Return type integer, float
Raises
• NetworkXError – The algorithm does not support MultiGraph and MultiDiGraph. If the
input graph is an instance of one of these two classes, a NetworkXError is raised.
• NetworkXUnbounded – If the graph has a path of infinite capacity, the value of a feasible
flow on the graph is unbounded above and the function raises a NetworkXUnbounded.
See also:
maximum_flow(), minimum_cut(), minimum_cut_value(), edmonds_karp(),
preflow_push(), shortest_augmenting_path()

Notes

The function used in the flow_func parameter has to return a residual network that follows NetworkX conven-
tions:
The residual network R from an input graph G has the same nodes as G. R is a DiGraph that contains a pair of
edges (u, v) and (v, u) iff (u, v) is not a self-loop, and at least one of (u, v) and (v, u) exists in
G.
For each edge (u, v) in R, R[u][v]['capacity'] is equal to the capacity of (u, v) in G if it exists in
G or zero otherwise. If the capacity is infinite, R[u][v]['capacity'] will have a high arbitrary finite value
that does not affect the solution of the problem. This value is stored in R.graph['inf']. For each edge (u,
v) in R, R[u][v]['flow'] represents the flow function of (u, v) and satisfies R[u][v]['flow']
== -R[v][u]['flow'].
The flow value, defined as the total flow into t, the sink, is stored in R.graph['flow_value']. Reacha-
bility to t using only edges (u, v) such that R[u][v]['flow'] < R[u][v]['capacity'] induces
a minimum s-t cut.
Specific algorithms may store extra data in R.
The function should supports an optional boolean parameter value_only. When True, it can optionally terminate
the algorithm as soon as the maximum flow value and the minimum cut can be determined.

Examples

>>> import networkx as nx


>>> G = nx.DiGraph()
>>> G.add_edge('x','a', capacity=3.0)
>>> G.add_edge('x','b', capacity=1.0)
>>> G.add_edge('a','c', capacity=3.0)
>>> G.add_edge('b','c', capacity=5.0)
(continues on next page)

316 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> G.add_edge('b','d', capacity=4.0)
>>> G.add_edge('d','e', capacity=2.0)
>>> G.add_edge('c','y', capacity=2.0)
>>> G.add_edge('e','y', capacity=3.0)

maximum_flow_value computes only the value of the maximum flow:

>>> flow_value = nx.maximum_flow_value(G, 'x', 'y')


>>> flow_value
3.0

You can also use alternative algorithms for computing the maximum flow by using the flow_func parameter.

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> flow_value == nx.maximum_flow_value(G, 'x', 'y',
... flow_func=shortest_augmenting_path)
True

networkx.algorithms.flow.minimum_cut

minimum_cut(flowG, _s, _t, capacity=’capacity’, flow_func=None, **kwargs)


Compute the value and the node partition of a minimum (s, t)-cut.
Use the max-flow min-cut theorem, i.e., the capacity of a minimum capacity cut is equal to the flow value of a
maximum flow.
Parameters
• flowG (NetworkX graph) – Edges of the graph are expected to have an attribute called ‘ca-
pacity’. If this attribute is not present, the edge is considered to have infinite capacity.
• _s (node) – Source node for the flow.
• _t (node) – Sink node for the flow.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• flow_func (function) – A function for computing the maximum flow among a pair of nodes
in a capacitated graph. The function has to accept at least three parameters: a Graph or
Digraph, a source node, and a target node. And return a residual network that follows Net-
workX conventions (see Notes). If flow_func is None, the default maximum flow function
(preflow_push()) is used. See below for alternative algorithms. The choice of the
default function may change from version to version and should not be relied on. Default
value: None.
• kwargs (Any other keyword parameter is passed to the function that) – computes the maxi-
mum flow.
Returns
• cut_value (integer, float) – Value of the minimum cut.
• partition (pair of node sets) – A partitioning of the nodes that defines a minimum cut.
Raises NetworkXUnbounded – If the graph has a path of infinite capacity, all cuts have infinite
capacity and the function raises a NetworkXError.

3.28. Flows 317


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
maximum_flow(), maximum_flow_value(), minimum_cut_value(), edmonds_karp(),
preflow_push(), shortest_augmenting_path()

Notes

The function used in the flow_func parameter has to return a residual network that follows NetworkX conven-
tions:
The residual network R from an input graph G has the same nodes as G. R is a DiGraph that contains a pair of
edges (u, v) and (v, u) iff (u, v) is not a self-loop, and at least one of (u, v) and (v, u) exists in
G.
For each edge (u, v) in R, R[u][v]['capacity'] is equal to the capacity of (u, v) in G if it exists in
G or zero otherwise. If the capacity is infinite, R[u][v]['capacity'] will have a high arbitrary finite value
that does not affect the solution of the problem. This value is stored in R.graph['inf']. For each edge (u,
v) in R, R[u][v]['flow'] represents the flow function of (u, v) and satisfies R[u][v]['flow']
== -R[v][u]['flow'].
The flow value, defined as the total flow into t, the sink, is stored in R.graph['flow_value']. Reacha-
bility to t using only edges (u, v) such that R[u][v]['flow'] < R[u][v]['capacity'] induces
a minimum s-t cut.
Specific algorithms may store extra data in R.
The function should supports an optional boolean parameter value_only. When True, it can optionally terminate
the algorithm as soon as the maximum flow value and the minimum cut can be determined.

Examples

>>> import networkx as nx


>>> G = nx.DiGraph()
>>> G.add_edge('x','a', capacity = 3.0)
>>> G.add_edge('x','b', capacity = 1.0)
>>> G.add_edge('a','c', capacity = 3.0)
>>> G.add_edge('b','c', capacity = 5.0)
>>> G.add_edge('b','d', capacity = 4.0)
>>> G.add_edge('d','e', capacity = 2.0)
>>> G.add_edge('c','y', capacity = 2.0)
>>> G.add_edge('e','y', capacity = 3.0)

minimum_cut computes both the value of the minimum cut and the node partition:

>>> cut_value, partition = nx.minimum_cut(G, 'x', 'y')


>>> reachable, non_reachable = partition

‘partition’ here is a tuple with the two sets of nodes that define the minimum cut. You can compute the cut set
of edges that induce the minimum cut as follows:

>>> cutset = set()


>>> for u, nbrs in ((n, G[n]) for n in reachable):
... cutset.update((u, v) for v in nbrs if v in non_reachable)
>>> print(sorted(cutset))
[('c', 'y'), ('x', 'b')]
>>> cut_value == sum(G.edges[u, v]['capacity'] for (u, v) in cutset)
True

318 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

You can also use alternative algorithms for computing the minimum cut by using the flow_func parameter.

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> cut_value == nx.minimum_cut(G, 'x', 'y',
... flow_func=shortest_augmenting_path)[0]
True

networkx.algorithms.flow.minimum_cut_value

minimum_cut_value(flowG, _s, _t, capacity=’capacity’, flow_func=None, **kwargs)


Compute the value of a minimum (s, t)-cut.
Use the max-flow min-cut theorem, i.e., the capacity of a minimum capacity cut is equal to the flow value of a
maximum flow.
Parameters
• flowG (NetworkX graph) – Edges of the graph are expected to have an attribute called ‘ca-
pacity’. If this attribute is not present, the edge is considered to have infinite capacity.
• _s (node) – Source node for the flow.
• _t (node) – Sink node for the flow.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• flow_func (function) – A function for computing the maximum flow among a pair of nodes
in a capacitated graph. The function has to accept at least three parameters: a Graph or
Digraph, a source node, and a target node. And return a residual network that follows Net-
workX conventions (see Notes). If flow_func is None, the default maximum flow function
(preflow_push()) is used. See below for alternative algorithms. The choice of the
default function may change from version to version and should not be relied on. Default
value: None.
• kwargs (Any other keyword parameter is passed to the function that) – computes the maxi-
mum flow.
Returns cut_value – Value of the minimum cut.
Return type integer, float
Raises NetworkXUnbounded – If the graph has a path of infinite capacity, all cuts have infinite
capacity and the function raises a NetworkXError.
See also:
maximum_flow(), maximum_flow_value(), minimum_cut(), edmonds_karp(),
preflow_push(), shortest_augmenting_path()

Notes

The function used in the flow_func parameter has to return a residual network that follows NetworkX conven-
tions:
The residual network R from an input graph G has the same nodes as G. R is a DiGraph that contains a pair of
edges (u, v) and (v, u) iff (u, v) is not a self-loop, and at least one of (u, v) and (v, u) exists in
G.

3.28. Flows 319


NetworkX Reference, Release 2.3rc1.dev20190222214247

For each edge (u, v) in R, R[u][v]['capacity'] is equal to the capacity of (u, v) in G if it exists in
G or zero otherwise. If the capacity is infinite, R[u][v]['capacity'] will have a high arbitrary finite value
that does not affect the solution of the problem. This value is stored in R.graph['inf']. For each edge (u,
v) in R, R[u][v]['flow'] represents the flow function of (u, v) and satisfies R[u][v]['flow']
== -R[v][u]['flow'].
The flow value, defined as the total flow into t, the sink, is stored in R.graph['flow_value']. Reacha-
bility to t using only edges (u, v) such that R[u][v]['flow'] < R[u][v]['capacity'] induces
a minimum s-t cut.
Specific algorithms may store extra data in R.
The function should supports an optional boolean parameter value_only. When True, it can optionally terminate
the algorithm as soon as the maximum flow value and the minimum cut can be determined.

Examples

>>> import networkx as nx


>>> G = nx.DiGraph()
>>> G.add_edge('x','a', capacity = 3.0)
>>> G.add_edge('x','b', capacity = 1.0)
>>> G.add_edge('a','c', capacity = 3.0)
>>> G.add_edge('b','c', capacity = 5.0)
>>> G.add_edge('b','d', capacity = 4.0)
>>> G.add_edge('d','e', capacity = 2.0)
>>> G.add_edge('c','y', capacity = 2.0)
>>> G.add_edge('e','y', capacity = 3.0)

minimum_cut_value computes only the value of the minimum cut:

>>> cut_value = nx.minimum_cut_value(G, 'x', 'y')


>>> cut_value
3.0

You can also use alternative algorithms for computing the minimum cut by using the flow_func parameter.

>>> from networkx.algorithms.flow import shortest_augmenting_path


>>> cut_value == nx.minimum_cut_value(G, 'x', 'y',
... flow_func=shortest_augmenting_path)
True

3.28.2 Edmonds-Karp

edmonds_karp(G, s, t[, capacity, residual, . . . ]) Find a maximum single-commodity flow using the
Edmonds-Karp algorithm.

networkx.algorithms.flow.edmonds_karp

edmonds_karp(G, s, t, capacity=’capacity’, residual=None, value_only=False, cutoff=None)


Find a maximum single-commodity flow using the Edmonds-Karp algorithm.
This function returns the residual network resulting after computing the maximum flow. See below for details
about the conventions NetworkX uses for defining residual networks.
This algorithm has a running time of 𝑂(𝑛𝑚2 ) for 𝑛 nodes and 𝑚 edges.

320 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (NetworkX graph) – Edges of the graph are expected to have an attribute called ‘capacity’.
If this attribute is not present, the edge is considered to have infinite capacity.
• s (node) – Source node for the flow.
• t (node) – Sink node for the flow.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• residual (NetworkX graph) – Residual network on which the algorithm is to be executed. If
None, a new residual network is created. Default value: None.
• value_only (bool) – If True compute only the value of the maximum flow. This parameter
will be ignored by this algorithm because it is not applicable.
• cutoff (integer, float) – If specified, the algorithm will terminate when the flow value reaches
or exceeds the cutoff. In this case, it may be unable to immediately determine a minimum
cut. Default value: None.
Returns R – Residual network after computing the maximum flow.
Return type NetworkX DiGraph
Raises
• NetworkXError – The algorithm does not support MultiGraph and MultiDiGraph. If the
input graph is an instance of one of these two classes, a NetworkXError is raised.
• NetworkXUnbounded – If the graph has a path of infinite capacity, the value of a feasible
flow on the graph is unbounded above and the function raises a NetworkXUnbounded.
See also:
maximum_flow(), minimum_cut(), preflow_push(), shortest_augmenting_path()

Notes

The residual network R from an input graph G has the same nodes as G. R is a DiGraph that contains a pair of
edges (u, v) and (v, u) iff (u, v) is not a self-loop, and at least one of (u, v) and (v, u) exists in
G.
For each edge (u, v) in R, R[u][v]['capacity'] is equal to the capacity of (u, v) in G if it exists in
G or zero otherwise. If the capacity is infinite, R[u][v]['capacity'] will have a high arbitrary finite value
that does not affect the solution of the problem. This value is stored in R.graph['inf']. For each edge (u,
v) in R, R[u][v]['flow'] represents the flow function of (u, v) and satisfies R[u][v]['flow']
== -R[v][u]['flow'].
The flow value, defined as the total flow into t, the sink, is stored in R.graph['flow_value']. If
cutoff is not specified, reachability to t using only edges (u, v) such that R[u][v]['flow'] <
R[u][v]['capacity'] induces a minimum s-t cut.

Examples

>>> import networkx as nx


>>> from networkx.algorithms.flow import edmonds_karp

3.28. Flows 321


NetworkX Reference, Release 2.3rc1.dev20190222214247

The functions that implement flow algorithms and output a residual network, such as this one, are not imported
to the base NetworkX namespace, so you have to explicitly import them from the flow package.

>>> G = nx.DiGraph()
>>> G.add_edge('x','a', capacity=3.0)
>>> G.add_edge('x','b', capacity=1.0)
>>> G.add_edge('a','c', capacity=3.0)
>>> G.add_edge('b','c', capacity=5.0)
>>> G.add_edge('b','d', capacity=4.0)
>>> G.add_edge('d','e', capacity=2.0)
>>> G.add_edge('c','y', capacity=2.0)
>>> G.add_edge('e','y', capacity=3.0)
>>> R = edmonds_karp(G, 'x', 'y')
>>> flow_value = nx.maximum_flow_value(G, 'x', 'y')
>>> flow_value
3.0
>>> flow_value == R.graph['flow_value']
True

3.28.3 Shortest Augmenting Path

shortest_augmenting_path(G, s, t[, . . . ]) Find a maximum single-commodity flow using the


shortest augmenting path algorithm.

networkx.algorithms.flow.shortest_augmenting_path

shortest_augmenting_path(G, s, t, capacity=’capacity’, residual=None, value_only=False,


two_phase=False, cutoff=None)
Find a maximum single-commodity flow using the shortest augmenting path algorithm.
This function returns the residual network resulting after computing the maximum flow. See below for details
about the conventions NetworkX uses for defining residual networks.
This algorithm has a running time of 𝑂(𝑛2 𝑚) for 𝑛 nodes and 𝑚 edges.
Parameters
• G (NetworkX graph) – Edges of the graph are expected to have an attribute called ‘capacity’.
If this attribute is not present, the edge is considered to have infinite capacity.
• s (node) – Source node for the flow.
• t (node) – Sink node for the flow.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• residual (NetworkX graph) – Residual network on which the algorithm is to be executed. If
None, a new residual network is created. Default value: None.
• value_only (bool) – If True compute only the value of the maximum flow. This parameter
will be ignored by this algorithm because it is not applicable.
• two_phase (bool) – If True, a two-phase variant is used. The two-phase variant improves
the running time on unit-capacity networks from 𝑂(𝑛𝑚) to 𝑂(min(𝑛2/3 , 𝑚1/2 )𝑚). Default
value: False.

322 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• cutoff (integer, float) – If specified, the algorithm will terminate when the flow value reaches
or exceeds the cutoff. In this case, it may be unable to immediately determine a minimum
cut. Default value: None.
Returns R – Residual network after computing the maximum flow.
Return type NetworkX DiGraph
Raises
• NetworkXError – The algorithm does not support MultiGraph and MultiDiGraph. If the
input graph is an instance of one of these two classes, a NetworkXError is raised.
• NetworkXUnbounded – If the graph has a path of infinite capacity, the value of a feasible
flow on the graph is unbounded above and the function raises a NetworkXUnbounded.
See also:
maximum_flow(), minimum_cut(), edmonds_karp(), preflow_push()

Notes

The residual network R from an input graph G has the same nodes as G. R is a DiGraph that contains a pair of
edges (u, v) and (v, u) iff (u, v) is not a self-loop, and at least one of (u, v) and (v, u) exists in
G.
For each edge (u, v) in R, R[u][v]['capacity'] is equal to the capacity of (u, v) in G if it exists in
G or zero otherwise. If the capacity is infinite, R[u][v]['capacity'] will have a high arbitrary finite value
that does not affect the solution of the problem. This value is stored in R.graph['inf']. For each edge (u,
v) in R, R[u][v]['flow'] represents the flow function of (u, v) and satisfies R[u][v]['flow']
== -R[v][u]['flow'].
The flow value, defined as the total flow into t, the sink, is stored in R.graph['flow_value']. If
cutoff is not specified, reachability to t using only edges (u, v) such that R[u][v]['flow'] <
R[u][v]['capacity'] induces a minimum s-t cut.

Examples

>>> import networkx as nx


>>> from networkx.algorithms.flow import shortest_augmenting_path

The functions that implement flow algorithms and output a residual network, such as this one, are not imported
to the base NetworkX namespace, so you have to explicitly import them from the flow package.

>>> G = nx.DiGraph()
>>> G.add_edge('x','a', capacity=3.0)
>>> G.add_edge('x','b', capacity=1.0)
>>> G.add_edge('a','c', capacity=3.0)
>>> G.add_edge('b','c', capacity=5.0)
>>> G.add_edge('b','d', capacity=4.0)
>>> G.add_edge('d','e', capacity=2.0)
>>> G.add_edge('c','y', capacity=2.0)
>>> G.add_edge('e','y', capacity=3.0)
>>> R = shortest_augmenting_path(G, 'x', 'y')
>>> flow_value = nx.maximum_flow_value(G, 'x', 'y')
>>> flow_value
3.0
(continues on next page)

3.28. Flows 323


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> flow_value == R.graph['flow_value']
True

3.28.4 Preflow-Push

preflow_push(G, s, t[, capacity, residual, . . . ]) Find a maximum single-commodity flow using the
highest-label preflow-push algorithm.

networkx.algorithms.flow.preflow_push

preflow_push(G, s, t, capacity=’capacity’, residual=None, global_relabel_freq=1, value_only=False)


Find a maximum single-commodity flow using the highest-label preflow-push algorithm.
This function returns the residual network resulting after computing the maximum flow. See below for details
about the conventions NetworkX uses for defining residual networks.

This algorithm has a running time of 𝑂(𝑛2 𝑚) for 𝑛 nodes and 𝑚 edges.
Parameters
• G (NetworkX graph) – Edges of the graph are expected to have an attribute called ‘capacity’.
If this attribute is not present, the edge is considered to have infinite capacity.
• s (node) – Source node for the flow.
• t (node) – Sink node for the flow.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• residual (NetworkX graph) – Residual network on which the algorithm is to be executed. If
None, a new residual network is created. Default value: None.
• global_relabel_freq (integer, float) – Relative frequency of applying the global relabeling
heuristic to speed up the algorithm. If it is None, the heuristic is disabled. Default value: 1.
• value_only (bool) – If False, compute a maximum flow; otherwise, compute a maximum
preflow which is enough for computing the maximum flow value. Default value: False.
Returns R – Residual network after computing the maximum flow.
Return type NetworkX DiGraph
Raises
• NetworkXError – The algorithm does not support MultiGraph and MultiDiGraph. If the
input graph is an instance of one of these two classes, a NetworkXError is raised.
• NetworkXUnbounded – If the graph has a path of infinite capacity, the value of a feasible
flow on the graph is unbounded above and the function raises a NetworkXUnbounded.
See also:
maximum_flow(), minimum_cut(), edmonds_karp(), shortest_augmenting_path()

324 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The residual network R from an input graph G has the same nodes as G. R is a DiGraph that contains a pair of
edges (u, v) and (v, u) iff (u, v) is not a self-loop, and at least one of (u, v) and (v, u) exists in
G. For each node u in R, R.nodes[u]['excess'] represents the difference between flow into u and flow
out of u.
For each edge (u, v) in R, R[u][v]['capacity'] is equal to the capacity of (u, v) in G if it exists in
G or zero otherwise. If the capacity is infinite, R[u][v]['capacity'] will have a high arbitrary finite value
that does not affect the solution of the problem. This value is stored in R.graph['inf']. For each edge (u,
v) in R, R[u][v]['flow'] represents the flow function of (u, v) and satisfies R[u][v]['flow']
== -R[v][u]['flow'].
The flow value, defined as the total flow into t, the sink, is stored in R.graph['flow_value']. Reacha-
bility to t using only edges (u, v) such that R[u][v]['flow'] < R[u][v]['capacity'] induces
a minimum s-t cut.

Examples

>>> import networkx as nx


>>> from networkx.algorithms.flow import preflow_push

The functions that implement flow algorithms and output a residual network, such as this one, are not imported
to the base NetworkX namespace, so you have to explicitly import them from the flow package.

>>> G = nx.DiGraph()
>>> G.add_edge('x','a', capacity=3.0)
>>> G.add_edge('x','b', capacity=1.0)
>>> G.add_edge('a','c', capacity=3.0)
>>> G.add_edge('b','c', capacity=5.0)
>>> G.add_edge('b','d', capacity=4.0)
>>> G.add_edge('d','e', capacity=2.0)
>>> G.add_edge('c','y', capacity=2.0)
>>> G.add_edge('e','y', capacity=3.0)
>>> R = preflow_push(G, 'x', 'y')
>>> flow_value = nx.maximum_flow_value(G, 'x', 'y')
>>> flow_value == R.graph['flow_value']
True
>>> # preflow_push also stores the maximum flow value
>>> # in the excess attribute of the sink node t
>>> flow_value == R.nodes['y']['excess']
True
>>> # For some problems, you might only want to compute a
>>> # maximum preflow.
>>> R = preflow_push(G, 'x', 'y', value_only=True)
>>> flow_value == R.graph['flow_value']
True
>>> flow_value == R.nodes['y']['excess']
True

3.28.5 Dinitz

3.28. Flows 325


NetworkX Reference, Release 2.3rc1.dev20190222214247

dinitz(G, s, t[, capacity, residual, . . . ]) Find a maximum single-commodity flow using Dinitz’
algorithm.

networkx.algorithms.flow.dinitz

dinitz(G, s, t, capacity=’capacity’, residual=None, value_only=False, cutoff=None)


Find a maximum single-commodity flow using Dinitz’ algorithm.
This function returns the residual network resulting after computing the maximum flow. See below for details
about the conventions NetworkX uses for defining residual networks.
This algorithm has a running time of 𝑂(𝑛2 𝑚) for 𝑛 nodes and 𝑚 edges1 .
Parameters
• G (NetworkX graph) – Edges of the graph are expected to have an attribute called ‘capacity’.
If this attribute is not present, the edge is considered to have infinite capacity.
• s (node) – Source node for the flow.
• t (node) – Sink node for the flow.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• residual (NetworkX graph) – Residual network on which the algorithm is to be executed. If
None, a new residual network is created. Default value: None.
• value_only (bool) – If True compute only the value of the maximum flow. This parameter
will be ignored by this algorithm because it is not applicable.
• cutoff (integer, float) – If specified, the algorithm will terminate when the flow value reaches
or exceeds the cutoff. In this case, it may be unable to immediately determine a minimum
cut. Default value: None.
Returns R – Residual network after computing the maximum flow.
Return type NetworkX DiGraph
Raises
• NetworkXError – The algorithm does not support MultiGraph and MultiDiGraph. If the
input graph is an instance of one of these two classes, a NetworkXError is raised.
• NetworkXUnbounded – If the graph has a path of infinite capacity, the value of a feasible
flow on the graph is unbounded above and the function raises a NetworkXUnbounded.
See also:
maximum_flow(), minimum_cut(), preflow_push(), shortest_augmenting_path()

Notes

The residual network R from an input graph G has the same nodes as G. R is a DiGraph that contains a pair of
edges (u, v) and (v, u) iff (u, v) is not a self-loop, and at least one of (u, v) and (v, u) exists in
G.
1 Dinitz’ Algorithm: The Original Version and Even’s Version. 2006. Yefim Dinitz. In Theoretical Computer Science. Lecture Notes in

Computer Science. Volume 3895. pp 218-240. http://www.cs.bgu.ac.il/~dinitz/Papers/Dinitz_alg.pdf

326 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

For each edge (u, v) in R, R[u][v]['capacity'] is equal to the capacity of (u, v) in G if it exists in
G or zero otherwise. If the capacity is infinite, R[u][v]['capacity'] will have a high arbitrary finite value
that does not affect the solution of the problem. This value is stored in R.graph['inf']. For each edge (u,
v) in R, R[u][v]['flow'] represents the flow function of (u, v) and satisfies R[u][v]['flow']
== -R[v][u]['flow'].
The flow value, defined as the total flow into t, the sink, is stored in R.graph['flow_value']. If
cutoff is not specified, reachability to t using only edges (u, v) such that R[u][v]['flow'] <
R[u][v]['capacity'] induces a minimum s-t cut.

Examples

>>> import networkx as nx


>>> from networkx.algorithms.flow import dinitz

The functions that implement flow algorithms and output a residual network, such as this one, are not imported
to the base NetworkX namespace, so you have to explicitly import them from the flow package.

>>> G = nx.DiGraph()
>>> G.add_edge('x','a', capacity=3.0)
>>> G.add_edge('x','b', capacity=1.0)
>>> G.add_edge('a','c', capacity=3.0)
>>> G.add_edge('b','c', capacity=5.0)
>>> G.add_edge('b','d', capacity=4.0)
>>> G.add_edge('d','e', capacity=2.0)
>>> G.add_edge('c','y', capacity=2.0)
>>> G.add_edge('e','y', capacity=3.0)
>>> R = dinitz(G, 'x', 'y')
>>> flow_value = nx.maximum_flow_value(G, 'x', 'y')
>>> flow_value
3.0
>>> flow_value == R.graph['flow_value']
True

References

3.28.6 Boykov-Kolmogorov

boykov_kolmogorov(G, s, t[, capacity, . . . ]) Find a maximum single-commodity flow using Boykov-


Kolmogorov algorithm.

networkx.algorithms.flow.boykov_kolmogorov

boykov_kolmogorov(G, s, t, capacity=’capacity’, residual=None, value_only=False, cutoff=None)


Find a maximum single-commodity flow using Boykov-Kolmogorov algorithm.
This function returns the residual network resulting after computing the maximum flow. See below for details
about the conventions NetworkX uses for defining residual networks.
This algorithm has worse case complexity 𝑂(𝑛2 𝑚|𝐶|) for 𝑛 nodes, 𝑚 edges, and |𝐶| the cost of the minimum
cut1 . This implementation uses the marking heuristic defined in2 which improves its running time in many
1 Boykov, Y., & Kolmogorov, V. (2004). An experimental comparison of min-cut/max-flow algorithms for energy minimization in vision.

Pattern Analysis and Machine Intelligence, IEEE Transactions on, 26(9), 1124-1137. http://www.csd.uwo.ca/~yuri/Papers/pami04.pdf
2 Vladimir Kolmogorov. Graph-based Algorithms for Multi-camera Reconstruction Problem. PhD thesis, Cornell University, CS Department,

3.28. Flows 327


NetworkX Reference, Release 2.3rc1.dev20190222214247

practical problems.
Parameters
• G (NetworkX graph) – Edges of the graph are expected to have an attribute called ‘capacity’.
If this attribute is not present, the edge is considered to have infinite capacity.
• s (node) – Source node for the flow.
• t (node) – Sink node for the flow.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• residual (NetworkX graph) – Residual network on which the algorithm is to be executed. If
None, a new residual network is created. Default value: None.
• value_only (bool) – If True compute only the value of the maximum flow. This parameter
will be ignored by this algorithm because it is not applicable.
• cutoff (integer, float) – If specified, the algorithm will terminate when the flow value reaches
or exceeds the cutoff. In this case, it may be unable to immediately determine a minimum
cut. Default value: None.
Returns R – Residual network after computing the maximum flow.
Return type NetworkX DiGraph
Raises
• NetworkXError – The algorithm does not support MultiGraph and MultiDiGraph. If the
input graph is an instance of one of these two classes, a NetworkXError is raised.
• NetworkXUnbounded – If the graph has a path of infinite capacity, the value of a feasible
flow on the graph is unbounded above and the function raises a NetworkXUnbounded.
See also:
maximum_flow(), minimum_cut(), preflow_push(), shortest_augmenting_path()

Notes

The residual network R from an input graph G has the same nodes as G. R is a DiGraph that contains a pair of
edges (u, v) and (v, u) iff (u, v) is not a self-loop, and at least one of (u, v) and (v, u) exists in
G.
For each edge (u, v) in R, R[u][v]['capacity'] is equal to the capacity of (u, v) in G if it exists in
G or zero otherwise. If the capacity is infinite, R[u][v]['capacity'] will have a high arbitrary finite value
that does not affect the solution of the problem. This value is stored in R.graph['inf']. For each edge (u,
v) in R, R[u][v]['flow'] represents the flow function of (u, v) and satisfies R[u][v]['flow']
== -R[v][u]['flow'].
The flow value, defined as the total flow into t, the sink, is stored in R.graph['flow_value']. If
cutoff is not specified, reachability to t using only edges (u, v) such that R[u][v]['flow'] <
R[u][v]['capacity'] induces a minimum s-t cut.
2003. pp. 109-114. https://pub.ist.ac.at/~vnk/papers/thesis.pdf

328 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> import networkx as nx


>>> from networkx.algorithms.flow import boykov_kolmogorov

The functions that implement flow algorithms and output a residual network, such as this one, are not imported
to the base NetworkX namespace, so you have to explicitly import them from the flow package.

>>> G = nx.DiGraph()
>>> G.add_edge('x','a', capacity=3.0)
>>> G.add_edge('x','b', capacity=1.0)
>>> G.add_edge('a','c', capacity=3.0)
>>> G.add_edge('b','c', capacity=5.0)
>>> G.add_edge('b','d', capacity=4.0)
>>> G.add_edge('d','e', capacity=2.0)
>>> G.add_edge('c','y', capacity=2.0)
>>> G.add_edge('e','y', capacity=3.0)
>>> R = boykov_kolmogorov(G, 'x', 'y')
>>> flow_value = nx.maximum_flow_value(G, 'x', 'y')
>>> flow_value
3.0
>>> flow_value == R.graph['flow_value']
True

A nice feature of the Boykov-Kolmogorov algorithm is that a partition of the nodes that defines a minimum cut
can be easily computed based on the search trees used during the algorithm. These trees are stored in the graph
attribute trees of the residual network.

>>> source_tree, target_tree = R.graph['trees']


>>> partition = (set(source_tree), set(G) - set(source_tree))

Or equivalently:

>>> partition = (set(G) - set(target_tree), set(target_tree))

References

3.28.7 Gomory-Hu Tree

gomory_hu_tree(G[, capacity, flow_func]) Returns the Gomory-Hu tree of an undirected graph G.

networkx.algorithms.flow.gomory_hu_tree

gomory_hu_tree(G, capacity=’capacity’, flow_func=None)


Returns the Gomory-Hu tree of an undirected graph G.
A Gomory-Hu tree of an undirected graph with capacities is a weighted tree that represents the minimum s-t
cuts for all s-t pairs in the graph.
It only requires n-1 minimum cut computations instead of the obvious n(n-1)/2. The tree represents all
s-t cuts as the minimum cut value among any pair of nodes is the minimum edge weight in the shortest path
between the two nodes in the Gomory-Hu tree.
The Gomory-Hu tree also has the property that removing the edge with the minimum weight in the shortest path
between any two nodes leaves two connected components that form a partition of the nodes in G that defines the

3.28. Flows 329


NetworkX Reference, Release 2.3rc1.dev20190222214247

minimum s-t cut.


See Examples section below for details.
Parameters
• G (NetworkX graph) – Undirected graph
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• flow_func (function) – Function to perform the underlying flow computations. Default value
edmonds_karp(). This function performs better in sparse graphs with right tailed degree
distributions. shortest_augmenting_path() will perform better in denser graphs.
Returns Tree – A NetworkX graph representing the Gomory-Hu tree of the input graph.
Return type NetworkX graph
Raises
• NetworkXNotImplemented : Exception – Raised if the input graph is directed.
• NetworkXError: Exception – Raised if the input graph is an empty Graph.

Examples

>>> G = nx.karate_club_graph()
>>> nx.set_edge_attributes(G, 1, 'capacity')
>>> T = nx.gomory_hu_tree(G)
>>> # The value of the minimum cut between any pair
... # of nodes in G is the minimum edge weight in the
... # shortest path between the two nodes in the
... # Gomory-Hu tree.
... def minimum_edge_weight_in_shortest_path(T, u, v):
... path = nx.shortest_path(T, u, v, weight='weight')
... return min((T[u][v]['weight'], (u,v)) for (u, v) in zip(path, path[1:]))
>>> u, v = 0, 33
>>> cut_value, edge = minimum_edge_weight_in_shortest_path(T, u, v)
>>> cut_value
10
>>> nx.minimum_cut_value(G, u, v)
10
>>> # The Comory-Hu tree also has the property that removing the
... # edge with the minimum weight in the shortest path between
... # any two nodes leaves two connected components that form
... # a partition of the nodes in G that defines the minimum s-t
... # cut.
... cut_value, edge = minimum_edge_weight_in_shortest_path(T, u, v)
>>> T.remove_edge(*edge)
>>> U, V = list(nx.connected_components(T))
>>> # Thus U and V form a partition that defines a minimum cut
... # between u and v in G. You can compute the edge cut set,
... # that is, the set of edges that if removed from G will
... # disconnect u from v in G, with this information:
... cutset = set()
>>> for x, nbrs in ((n, G[n]) for n in U):
... cutset.update((x, y) for y in nbrs if y in V)
>>> # Because we have set the capacities of all edges to 1
(continues on next page)

330 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


... # the cutset contains ten edges
... len(cutset)
10
>>> # You can use any maximum flow algorithm for the underlying
... # flow computations using the argument flow_func
... from networkx.algorithms import flow
>>> T = nx.gomory_hu_tree(G, flow_func=flow.boykov_kolmogorov)
>>> cut_value, edge = minimum_edge_weight_in_shortest_path(T, u, v)
>>> cut_value
10
>>> nx.minimum_cut_value(G, u, v, flow_func=flow.boykov_kolmogorov)
10

Notes

This implementation is based on Gusfield approach1 to compute Comory-Hu trees, which does not require node
contractions and has the same computational complexity than the original method.
See also:
minimum_cut(), maximum_flow()

References

3.28.8 Utils

build_residual_network(G, capacity) Build a residual network and initialize a zero flow.

networkx.algorithms.flow.build_residual_network

build_residual_network(G, capacity)
Build a residual network and initialize a zero flow.
The residual network R from an input graph G has the same nodes as G. R is a DiGraph that contains a pair of
edges (u, v) and (v, u) iff (u, v) is not a self-loop, and at least one of (u, v) and (v, u) exists in
G.
For each edge (u, v) in R, R[u][v]['capacity'] is equal to the capacity of (u, v) in G if it exists in
G or zero otherwise. If the capacity is infinite, R[u][v]['capacity'] will have a high arbitrary finite value
that does not affect the solution of the problem. This value is stored in R.graph['inf']. For each edge (u,
v) in R, R[u][v]['flow'] represents the flow function of (u, v) and satisfies R[u][v]['flow']
== -R[v][u]['flow'].
The flow value, defined as the total flow into t, the sink, is stored in R.graph['flow_value']. If
cutoff is not specified, reachability to t using only edges (u, v) such that R[u][v]['flow'] <
R[u][v]['capacity'] induces a minimum s-t cut.

3.28.9 Network Simplex

1 Gusfield D: Very simple methods for all pairs network flow analysis. SIAM J Comput 19(1):143-155, 1990.

3.28. Flows 331


NetworkX Reference, Release 2.3rc1.dev20190222214247

network_simplex(G[, demand, capacity, weight]) Find a minimum cost flow satisfying all demands in di-
graph G.
min_cost_flow_cost(G[, demand, capacity, Find the cost of a minimum cost flow satisfying all de-
weight]) mands in digraph G.
min_cost_flow(G[, demand, capacity, weight]) Returns a minimum cost flow satisfying all demands in
digraph G.
cost_of_flow(G, flowDict[, weight]) Compute the cost of the flow given by flowDict on graph
G.
max_flow_min_cost(G, s, t[, capacity, weight]) Returns a maximum (s, t)-flow of minimum cost.

networkx.algorithms.flow.network_simplex

network_simplex(G, demand=’demand’, capacity=’capacity’, weight=’weight’)


Find a minimum cost flow satisfying all demands in digraph G.
This is a primal network simplex algorithm that uses the leaving arc rule to prevent cycling.
G is a digraph with edge costs and capacities and in which nodes have demand, i.e., they want to send or receive
some amount of flow. A negative demand means that the node wants to send flow, a positive demand means
that the node want to receive flow. A flow on the digraph G satisfies all demand if the net flow into each node is
equal to the demand of that node.
Parameters
• G (NetworkX graph) – DiGraph on which a minimum cost flow satisfying all demands is to
be found.
• demand (string) – Nodes of the graph G are expected to have an attribute demand that indi-
cates how much flow a node wants to send (negative demand) or receive (positive demand).
Note that the sum of the demands should be 0 otherwise the problem in not feasible. If this
attribute is not present, a node is considered to have 0 demand. Default value: ‘demand’.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• weight (string) – Edges of the graph G are expected to have an attribute weight that indicates
the cost incurred by sending one unit of flow on that edge. If not present, the weight is
considered to be 0. Default value: ‘weight’.
Returns
• flowCost (integer, float) – Cost of a minimum cost flow satisfying all demands.
• flowDict (dictionary) – Dictionary of dictionaries keyed by nodes such that flowDict[u][v]
is the flow edge (u, v).
Raises
• NetworkXError – This exception is raised if the input graph is not directed, not con-
nected or is a multigraph.
• NetworkXUnfeasible – This exception is raised in the following situations:
– The sum of the demands is not zero. Then, there is no flow satisfying all demands.
– There is no flow satisfying all demand.
• NetworkXUnbounded – This exception is raised if the digraph G has a cycle of negative
cost and infinite capacity. Then, the cost of a flow satisfying all demands is unbounded
below.

332 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This algorithm is not guaranteed to work if edge weights or demands are floating point numbers (overflows and
roundoff errors can cause problems). As a workaround you can use integer numbers by multiplying the relevant
edge attributes by a convenient constant factor (eg 100).
See also:
cost_of_flow(), max_flow_min_cost(), min_cost_flow(), min_cost_flow_cost()

Examples

A simple example of a min cost flow problem.


>>> import networkx as nx
>>> G = nx.DiGraph()
>>> G.add_node('a', demand=-5)
>>> G.add_node('d', demand=5)
>>> G.add_edge('a', 'b', weight=3, capacity=4)
>>> G.add_edge('a', 'c', weight=6, capacity=10)
>>> G.add_edge('b', 'd', weight=1, capacity=9)
>>> G.add_edge('c', 'd', weight=2, capacity=5)
>>> flowCost, flowDict = nx.network_simplex(G)
>>> flowCost
24
>>> flowDict # doctest: +SKIP
{'a': {'c': 1, 'b': 4}, 'c': {'d': 1}, 'b': {'d': 4}, 'd': {}}

The mincost flow algorithm can also be used to solve shortest path problems. To find the shortest path between
two nodes u and v, give all edges an infinite capacity, give node u a demand of -1 and node v a demand a 1. Then
run the network simplex. The value of a min cost flow will be the distance between u and v and edges carrying
positive flow will indicate the path.
>>> G=nx.DiGraph()
>>> G.add_weighted_edges_from([('s', 'u' ,10), ('s' ,'x' ,5),
... ('u', 'v' ,1), ('u' ,'x' ,2),
... ('v', 'y' ,1), ('x' ,'u' ,3),
... ('x', 'v' ,5), ('x' ,'y' ,2),
... ('y', 's' ,7), ('y' ,'v' ,6)])
>>> G.add_node('s', demand = -1)
>>> G.add_node('v', demand = 1)
>>> flowCost, flowDict = nx.network_simplex(G)
>>> flowCost == nx.shortest_path_length(G, 's', 'v', weight='weight')
True
>>> sorted([(u, v) for u in flowDict for v in flowDict[u] if flowDict[u][v] > 0])
[('s', 'x'), ('u', 'v'), ('x', 'u')]
>>> nx.shortest_path(G, 's', 'v', weight = 'weight')
['s', 'x', 'u', 'v']

It is possible to change the name of the attributes used for the algorithm.
>>> G = nx.DiGraph()
>>> G.add_node('p', spam=-4)
>>> G.add_node('q', spam=2)
>>> G.add_node('a', spam=-2)
>>> G.add_node('d', spam=-1)
>>> G.add_node('t', spam=2)
(continues on next page)

3.28. Flows 333


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> G.add_node('w', spam=3)
>>> G.add_edge('p', 'q', cost=7, vacancies=5)
>>> G.add_edge('p', 'a', cost=1, vacancies=4)
>>> G.add_edge('q', 'd', cost=2, vacancies=3)
>>> G.add_edge('t', 'q', cost=1, vacancies=2)
>>> G.add_edge('a', 't', cost=2, vacancies=4)
>>> G.add_edge('d', 'w', cost=3, vacancies=4)
>>> G.add_edge('t', 'w', cost=4, vacancies=1)
>>> flowCost, flowDict = nx.network_simplex(G, demand='spam',
... capacity='vacancies',
... weight='cost')
>>> flowCost
37
>>> flowDict # doctest: +SKIP
{'a': {'t': 4}, 'd': {'w': 2}, 'q': {'d': 1}, 'p': {'q': 2, 'a': 2}, 't': {'q': 1,
˓→ 'w': 1}, 'w': {}}

References

networkx.algorithms.flow.min_cost_flow_cost

min_cost_flow_cost(G, demand=’demand’, capacity=’capacity’, weight=’weight’)


Find the cost of a minimum cost flow satisfying all demands in digraph G.
G is a digraph with edge costs and capacities and in which nodes have demand, i.e., they want to send or receive
some amount of flow. A negative demand means that the node wants to send flow, a positive demand means
that the node want to receive flow. A flow on the digraph G satisfies all demand if the net flow into each node is
equal to the demand of that node.
Parameters
• G (NetworkX graph) – DiGraph on which a minimum cost flow satisfying all demands is to
be found.
• demand (string) – Nodes of the graph G are expected to have an attribute demand that indi-
cates how much flow a node wants to send (negative demand) or receive (positive demand).
Note that the sum of the demands should be 0 otherwise the problem in not feasible. If this
attribute is not present, a node is considered to have 0 demand. Default value: ‘demand’.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• weight (string) – Edges of the graph G are expected to have an attribute weight that indicates
the cost incurred by sending one unit of flow on that edge. If not present, the weight is
considered to be 0. Default value: ‘weight’.
Returns flowCost – Cost of a minimum cost flow satisfying all demands.
Return type integer, float
Raises
• NetworkXError – This exception is raised if the input graph is not directed or not con-
nected.
• NetworkXUnfeasible – This exception is raised in the following situations:
– The sum of the demands is not zero. Then, there is no flow satisfying all demands.

334 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

– There is no flow satisfying all demand.


• NetworkXUnbounded – This exception is raised if the digraph G has a cycle of negative
cost and infinite capacity. Then, the cost of a flow satisfying all demands is unbounded
below.
See also:
cost_of_flow(), max_flow_min_cost(), min_cost_flow(), network_simplex()

Notes

This algorithm is not guaranteed to work if edge weights or demands are floating point numbers (overflows and
roundoff errors can cause problems). As a workaround you can use integer numbers by multiplying the relevant
edge attributes by a convenient constant factor (eg 100).

Examples

A simple example of a min cost flow problem.

>>> import networkx as nx


>>> G = nx.DiGraph()
>>> G.add_node('a', demand = -5)
>>> G.add_node('d', demand = 5)
>>> G.add_edge('a', 'b', weight = 3, capacity = 4)
>>> G.add_edge('a', 'c', weight = 6, capacity = 10)
>>> G.add_edge('b', 'd', weight = 1, capacity = 9)
>>> G.add_edge('c', 'd', weight = 2, capacity = 5)
>>> flowCost = nx.min_cost_flow_cost(G)
>>> flowCost
24

networkx.algorithms.flow.min_cost_flow

min_cost_flow(G, demand=’demand’, capacity=’capacity’, weight=’weight’)


Returns a minimum cost flow satisfying all demands in digraph G.
G is a digraph with edge costs and capacities and in which nodes have demand, i.e., they want to send or receive
some amount of flow. A negative demand means that the node wants to send flow, a positive demand means
that the node want to receive flow. A flow on the digraph G satisfies all demand if the net flow into each node is
equal to the demand of that node.
Parameters
• G (NetworkX graph) – DiGraph on which a minimum cost flow satisfying all demands is to
be found.
• demand (string) – Nodes of the graph G are expected to have an attribute demand that indi-
cates how much flow a node wants to send (negative demand) or receive (positive demand).
Note that the sum of the demands should be 0 otherwise the problem in not feasible. If this
attribute is not present, a node is considered to have 0 demand. Default value: ‘demand’.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.

3.28. Flows 335


NetworkX Reference, Release 2.3rc1.dev20190222214247

• weight (string) – Edges of the graph G are expected to have an attribute weight that indicates
the cost incurred by sending one unit of flow on that edge. If not present, the weight is
considered to be 0. Default value: ‘weight’.
Returns flowDict – Dictionary of dictionaries keyed by nodes such that flowDict[u][v] is the flow
edge (u, v).
Return type dictionary
Raises
• NetworkXError – This exception is raised if the input graph is not directed or not con-
nected.
• NetworkXUnfeasible – This exception is raised in the following situations:
– The sum of the demands is not zero. Then, there is no flow satisfying all demands.
– There is no flow satisfying all demand.
• NetworkXUnbounded – This exception is raised if the digraph G has a cycle of negative
cost and infinite capacity. Then, the cost of a flow satisfying all demands is unbounded
below.
See also:
cost_of_flow(), max_flow_min_cost(), min_cost_flow_cost(), network_simplex()

Notes

This algorithm is not guaranteed to work if edge weights or demands are floating point numbers (overflows and
roundoff errors can cause problems). As a workaround you can use integer numbers by multiplying the relevant
edge attributes by a convenient constant factor (eg 100).

Examples

A simple example of a min cost flow problem.

>>> import networkx as nx


>>> G = nx.DiGraph()
>>> G.add_node('a', demand = -5)
>>> G.add_node('d', demand = 5)
>>> G.add_edge('a', 'b', weight = 3, capacity = 4)
>>> G.add_edge('a', 'c', weight = 6, capacity = 10)
>>> G.add_edge('b', 'd', weight = 1, capacity = 9)
>>> G.add_edge('c', 'd', weight = 2, capacity = 5)
>>> flowDict = nx.min_cost_flow(G)

networkx.algorithms.flow.cost_of_flow

cost_of_flow(G, flowDict, weight=’weight’)


Compute the cost of the flow given by flowDict on graph G.
Note that this function does not check for the validity of the flow flowDict. This function will fail if the graph G
and the flow don’t have the same edge set.
Parameters

336 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• G (NetworkX graph) – DiGraph on which a minimum cost flow satisfying all demands is to
be found.
• weight (string) – Edges of the graph G are expected to have an attribute weight that indicates
the cost incurred by sending one unit of flow on that edge. If not present, the weight is
considered to be 0. Default value: ‘weight’.
• flowDict (dictionary) – Dictionary of dictionaries keyed by nodes such that flowDict[u][v]
is the flow edge (u, v).
Returns cost – The total cost of the flow. This is given by the sum over all edges of the product of
the edge’s flow and the edge’s weight.
Return type Integer, float
See also:
max_flow_min_cost(), min_cost_flow(), min_cost_flow_cost(), network_simplex()

Notes

This algorithm is not guaranteed to work if edge weights or demands are floating point numbers (overflows and
roundoff errors can cause problems). As a workaround you can use integer numbers by multiplying the relevant
edge attributes by a convenient constant factor (eg 100).

networkx.algorithms.flow.max_flow_min_cost

max_flow_min_cost(G, s, t, capacity=’capacity’, weight=’weight’)


Returns a maximum (s, t)-flow of minimum cost.
G is a digraph with edge costs and capacities. There is a source node s and a sink node t. This function finds a
maximum flow from s to t whose total cost is minimized.
Parameters
• G (NetworkX graph) – DiGraph on which a minimum cost flow satisfying all demands is to
be found.
• s (node label) – Source of the flow.
• t (node label) – Destination of the flow.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• weight (string) – Edges of the graph G are expected to have an attribute weight that indicates
the cost incurred by sending one unit of flow on that edge. If not present, the weight is
considered to be 0. Default value: ‘weight’.
Returns flowDict – Dictionary of dictionaries keyed by nodes such that flowDict[u][v] is the flow
edge (u, v).
Return type dictionary
Raises
• NetworkXError – This exception is raised if the input graph is not directed or not con-
nected.

3.28. Flows 337


NetworkX Reference, Release 2.3rc1.dev20190222214247

• NetworkXUnbounded – This exception is raised if there is an infinite capacity path from


s to t in G. In this case there is no maximum flow. This exception is also raised if the digraph
G has a cycle of negative cost and infinite capacity. Then, the cost of a flow is unbounded
below.
See also:
cost_of_flow(), min_cost_flow(), min_cost_flow_cost(), network_simplex()

Notes

This algorithm is not guaranteed to work if edge weights or demands are floating point numbers (overflows and
roundoff errors can cause problems). As a workaround you can use integer numbers by multiplying the relevant
edge attributes by a convenient constant factor (eg 100).

Examples

>>> G = nx.DiGraph()
>>> G.add_edges_from([(1, 2, {'capacity': 12, 'weight': 4}),
... (1, 3, {'capacity': 20, 'weight': 6}),
... (2, 3, {'capacity': 6, 'weight': -3}),
... (2, 6, {'capacity': 14, 'weight': 1}),
... (3, 4, {'weight': 9}),
... (3, 5, {'capacity': 10, 'weight': 5}),
... (4, 2, {'capacity': 19, 'weight': 13}),
... (4, 5, {'capacity': 4, 'weight': 0}),
... (5, 7, {'capacity': 28, 'weight': 2}),
... (6, 5, {'capacity': 11, 'weight': 1}),
... (6, 7, {'weight': 8}),
... (7, 4, {'capacity': 6, 'weight': 6})])
>>> mincostFlow = nx.max_flow_min_cost(G, 1, 7)
>>> mincost = nx.cost_of_flow(G, mincostFlow)
>>> mincost
373
>>> from networkx.algorithms.flow import maximum_flow
>>> maxFlow = maximum_flow(G, 1, 7)[1]
>>> nx.cost_of_flow(G, maxFlow) >= mincost
True
>>> mincostFlowValue = (sum((mincostFlow[u][7] for u in G.predecessors(7)))
... - sum((mincostFlow[7][v] for v in G.successors(7))))
>>> mincostFlowValue == nx.maximum_flow_value(G, 1, 7)
True

3.28.10 Capacity Scaling Minimum Cost Flow

capacity_scaling(G[, demand, capacity, . . . ]) Find a minimum cost flow satisfying all demands in di-
graph G.

networkx.algorithms.flow.capacity_scaling

capacity_scaling(G, demand=’demand’, capacity=’capacity’, weight=’weight’, heap=<class ’net-


workx.utils.heaps.BinaryHeap’>)
Find a minimum cost flow satisfying all demands in digraph G.

338 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

This is a capacity scaling successive shortest augmenting path algorithm.


G is a digraph with edge costs and capacities and in which nodes have demand, i.e., they want to send or receive
some amount of flow. A negative demand means that the node wants to send flow, a positive demand means
that the node want to receive flow. A flow on the digraph G satisfies all demand if the net flow into each node is
equal to the demand of that node.
Parameters
• G (NetworkX graph) – DiGraph or MultiDiGraph on which a minimum cost flow satisfying
all demands is to be found.
• demand (string) – Nodes of the graph G are expected to have an attribute demand that indi-
cates how much flow a node wants to send (negative demand) or receive (positive demand).
Note that the sum of the demands should be 0 otherwise the problem in not feasible. If this
attribute is not present, a node is considered to have 0 demand. Default value: ‘demand’.
• capacity (string) – Edges of the graph G are expected to have an attribute capacity that
indicates how much flow the edge can support. If this attribute is not present, the edge is
considered to have infinite capacity. Default value: ‘capacity’.
• weight (string) – Edges of the graph G are expected to have an attribute weight that indicates
the cost incurred by sending one unit of flow on that edge. If not present, the weight is
considered to be 0. Default value: ‘weight’.
• heap (class) – Type of heap to be used in the algorithm. It should be a subclass of MinHeap
or implement a compatible interface.
If a stock heap implementation is to be used, BinaryHeap is recommended over
PairingHeap for Python implementations without optimized attribute accesses (e.g.,
CPython) despite a slower asymptotic running time. For Python implementations with opti-
mized attribute accesses (e.g., PyPy), PairingHeap provides better performance. Default
value: BinaryHeap.
Returns
• flowCost (integer) – Cost of a minimum cost flow satisfying all demands.
• flowDict (dictionary) – If G is a digraph, a dict-of-dicts keyed by nodes such that flow-
Dict[u][v] is the flow on edge (u, v). If G is a MultiDiGraph, a dict-of-dicts-of-dicts keyed
by nodes so that flowDict[u][v][key] is the flow on edge (u, v, key).
Raises
• NetworkXError – This exception is raised if the input graph is not directed, not con-
nected.
• NetworkXUnfeasible – This exception is raised in the following situations:
– The sum of the demands is not zero. Then, there is no flow satisfying all demands.
– There is no flow satisfying all demand.
• NetworkXUnbounded – This exception is raised if the digraph G has a cycle of negative
cost and infinite capacity. Then, the cost of a flow satisfying all demands is unbounded
below.

Notes

This algorithm does not work if edge weights are floating-point numbers.
See also:

3.28. Flows 339


NetworkX Reference, Release 2.3rc1.dev20190222214247

network_simplex()

Examples

A simple example of a min cost flow problem.

>>> import networkx as nx


>>> G = nx.DiGraph()
>>> G.add_node('a', demand = -5)
>>> G.add_node('d', demand = 5)
>>> G.add_edge('a', 'b', weight = 3, capacity = 4)
>>> G.add_edge('a', 'c', weight = 6, capacity = 10)
>>> G.add_edge('b', 'd', weight = 1, capacity = 9)
>>> G.add_edge('c', 'd', weight = 2, capacity = 5)
>>> flowCost, flowDict = nx.capacity_scaling(G)
>>> flowCost
24
>>> flowDict # doctest: +SKIP
{'a': {'c': 1, 'b': 4}, 'c': {'d': 1}, 'b': {'d': 4}, 'd': {}}

It is possible to change the name of the attributes used for the algorithm.

>>> G = nx.DiGraph()
>>> G.add_node('p', spam = -4)
>>> G.add_node('q', spam = 2)
>>> G.add_node('a', spam = -2)
>>> G.add_node('d', spam = -1)
>>> G.add_node('t', spam = 2)
>>> G.add_node('w', spam = 3)
>>> G.add_edge('p', 'q', cost = 7, vacancies = 5)
>>> G.add_edge('p', 'a', cost = 1, vacancies = 4)
>>> G.add_edge('q', 'd', cost = 2, vacancies = 3)
>>> G.add_edge('t', 'q', cost = 1, vacancies = 2)
>>> G.add_edge('a', 't', cost = 2, vacancies = 4)
>>> G.add_edge('d', 'w', cost = 3, vacancies = 4)
>>> G.add_edge('t', 'w', cost = 4, vacancies = 1)
>>> flowCost, flowDict = nx.capacity_scaling(G, demand = 'spam',
... capacity = 'vacancies',
... weight = 'cost')
>>> flowCost
37
>>> flowDict # doctest: +SKIP
{'a': {'t': 4}, 'd': {'w': 2}, 'q': {'d': 1}, 'p': {'q': 2, 'a': 2}, 't': {'q': 1,
˓→ 'w': 1}, 'w': {}}

3.29 Graphical degree sequence

Test sequences for graphiness.

is_graphical(sequence[, method]) Returns True if sequence is a valid degree sequence.


is_digraphical(in_sequence, out_sequence) Returns True if some directed graph can realize the in-
and out-degree sequences.
Continued on next page

340 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 96 – continued from previous page


is_multigraphical(sequence) Returns True if some multigraph can realize the se-
quence.
is_pseudographical(sequence) Returns True if some pseudograph can realize the se-
quence.
is_valid_degree_sequence_havel_hakimi(. . .Returns
) True if deg_sequence can be realized by a sim-
ple graph.
is_valid_degree_sequence_erdos_gallai(. . .Returns
) True if deg_sequence can be realized by a sim-
ple graph.

3.29.1 networkx.algorithms.graphical.is_graphical

is_graphical(sequence, method=’eg’)
Returns True if sequence is a valid degree sequence.
A degree sequence is valid if some graph can realize it.
Parameters
• sequence (list or iterable container) – A sequence of integer node degrees
• method (“eg” | “hh”) – The method used to validate the degree sequence. “eg” corresponds
to the Erdős-Gallai algorithm, and “hh” to the Havel-Hakimi algorithm.
Returns valid – True if the sequence is a valid degree sequence and False if not.
Return type bool

Examples

>>> G = nx.path_graph(4)
>>> sequence = (d for n, d in G.degree())
>>> nx.is_graphical(sequence)
True

References

Erdős-Gallai [?], [?]


Havel-Hakimi [?], [?], [?]

3.29.2 networkx.algorithms.graphical.is_digraphical

is_digraphical(in_sequence, out_sequence)
Returns True if some directed graph can realize the in- and out-degree sequences.
Parameters
• in_sequence (list or iterable container) – A sequence of integer node in-degrees
• out_sequence (list or iterable container) – A sequence of integer node out-degrees
Returns valid – True if in and out-sequences are digraphic False if not.
Return type bool

3.29. Graphical degree sequence 341


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This algorithm is from Kleitman and Wang1 . The worst case runtime is 𝑂(𝑠 × log 𝑛) where 𝑠 and 𝑛 are the sum
and length of the sequences respectively.

References

3.29.3 networkx.algorithms.graphical.is_multigraphical

is_multigraphical(sequence)
Returns True if some multigraph can realize the sequence.
Parameters sequence (list) – A list of integers
Returns valid – True if deg_sequence is a multigraphic degree sequence and False if not.
Return type bool

Notes

The worst-case run time is 𝑂(𝑛) where 𝑛 is the length of the sequence.

References

3.29.4 networkx.algorithms.graphical.is_pseudographical

is_pseudographical(sequence)
Returns True if some pseudograph can realize the sequence.
Every nonnegative integer sequence with an even sum is pseudographical (see1 ).
Parameters sequence (list or iterable container) – A sequence of integer node degrees
Returns valid – True if the sequence is a pseudographic degree sequence and False if not.
Return type bool

Notes

The worst-case run time is 𝑂(𝑛) where n is the length of the sequence.

References

3.29.5 networkx.algorithms.graphical.is_valid_degree_sequence_havel_hakimi

is_valid_degree_sequence_havel_hakimi(deg_sequence)
Returns True if deg_sequence can be realized by a simple graph.
The validation proceeds using the Havel-Hakimi theorem. Worst-case run time is 𝑂(𝑠) where 𝑠 is the sum of
the sequence.
1 D.J. Kleitman and D.L. Wang Algorithms for Constructing Graphs and Digraphs with Given Valences and Factors, Discrete Mathematics,

6(1), pp. 79-88 (1973)


1 F. Boesch and F. Harary. “Line removal algorithms for graphs and their degree lists”, IEEE Trans. Circuits and Systems, CAS-23(12), pp.

778-782 (1976).

342 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters deg_sequence (list) – A list of integers where each element specifies the degree of a
node in a graph.
Returns valid – True if deg_sequence is graphical and False if not.
Return type bool

Notes

The ZZ condition says that for the sequence d if


(max(𝑑) + min(𝑑) + 1)2
|𝑑| >=
4 * min(𝑑)
then d is graphical. This was shown in Theorem 6 in1 .

References

[?], [?], [?]

3.29.6 networkx.algorithms.graphical.is_valid_degree_sequence_erdos_gallai

is_valid_degree_sequence_erdos_gallai(deg_sequence)
Returns True if deg_sequence can be realized by a simple graph.
The validation is done using the Erdős-Gallai theorem [?].
Parameters deg_sequence (list) – A list of integers
Returns valid – True if deg_sequence is graphical and False if not.
Return type bool

Notes

This implementation uses an equivalent form of the Erdős-Gallai criterion. Worst-case run time is 𝑂(𝑛) where
𝑛 is the length of the sequence.
Specifically, a sequence d is graphical if and only if the sum of the sequence is even and for all strong indices k
in the sequence,

𝑘
∑︁ 𝑛
∑︁ 𝑘−1
∑︁ 𝑘−1
∑︁
𝑑𝑖 ≤ 𝑘(𝑘 − 1) + min(𝑑𝑖 , 𝑘) = 𝑘(𝑛 − 1) − (𝑘 𝑛𝑗 − 𝑗𝑛𝑗 )
𝑖=1 𝑗=𝑘+1 𝑗=0 𝑗=0

A strong index k is any index where d_k >= k and the value n_j is the number of occurrences of j in d. The
maximal strong index is called the Durfee index.
This particular rearrangement comes from the proof of Theorem 3 in2 .
The ZZ condition says that for the sequence d if
(max(𝑑) + min(𝑑) + 1)2
|𝑑| >=
4 * min(𝑑)
1 I.E. Zverovich and V.E. Zverovich. “Contributions to the theory of graphic sequences”, Discrete Mathematics, 105, pp. 292-303 (1992).
2 I.E. Zverovich and V.E. Zverovich. “Contributions to the theory of graphic sequences”, Discrete Mathematics, 105, pp. 292-303 (1992).

3.29. Graphical degree sequence 343


NetworkX Reference, Release 2.3rc1.dev20190222214247

then d is graphical. This was shown in Theorem 6 in2 .

References

[?], [?]

3.30 Hierarchy

Flow Hierarchy.

flow_hierarchy(G[, weight]) Returns the flow hierarchy of a directed network.

3.30.1 networkx.algorithms.hierarchy.flow_hierarchy

flow_hierarchy(G, weight=None)
Returns the flow hierarchy of a directed network.
Flow hierarchy is defined as the fraction of edges not participating in cycles in a directed graph1 .
Parameters
• G (DiGraph or MultiDiGraph) – A directed graph
• weight (key,optional (default=None)) – Attribute to use for node weights. If None the
weight defaults to 1.
Returns h – Flow hierarchy value
Return type float

Notes

The algorithm described in1 computes the flow hierarchy through exponentiation of the adjacency matrix. This
function implements an alternative approach that finds strongly connected components. An edge is in a cycle if
and only if it is in a strongly connected component, which can be found in 𝑂(𝑚) time using Tarjan’s algorithm.

References

3.31 Hybrid

Provides functions for finding and testing for locally (k, l)-connected graphs.

kl_connected_subgraph(G, k, l[, low_memory, Returns the maximum locally (k, l)-connected sub-
. . . ]) graph of G.
is_kl_connected(G, k, l[, low_memory]) Returns True if and only if G is locally (k, l)-
connected.
1 Luo, J.; Magee, C.L. (2011), Detecting evolving patterns of self-organizing networks by flow hierarchy measurement, Complexity, Volume 16

Issue 6 53-61. DOI: 10.1002/cplx.20368 http://web.mit.edu/~cmagee/www/documents/28-DetectingEvolvingPatterns_FlowHierarchy.pdf

344 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.31.1 networkx.algorithms.hybrid.kl_connected_subgraph

kl_connected_subgraph(G, k, l, low_memory=False, same_as_graph=False)


Returns the maximum locally (k, l)-connected subgraph of G.
A graph is locally (k, l)-connected if for each edge (u, v) in the graph there are at least l edge-disjoint
paths of length at most k joining u to v.
Parameters
• G (NetworkX graph) – The graph in which to find a maximum locally (k, l)-connected
subgraph.
• k (integer) – The maximum length of paths to consider. A higher number means a looser
connectivity requirement.
• l (integer) – The number of edge-disjoint paths. A higher number means a stricter connec-
tivity requirement.
• low_memory (bool) – If this is True, this function uses an algorithm that uses slightly more
time but less memory.
• same_as_graph (bool) – If True then return a tuple of the form (H, is_same), where
H is the maximum locally (k, l)-connected subgraph and is_same is a Boolean repre-
senting whether G is locally (k, l)-connected (and hence, whether H is simply a copy of
the input graph G).
Returns If same_as_graph is True, then this function returns a two-tuple as described above.
Otherwise, it returns only the maximum locally (k, l)-connected subgraph.
Return type NetworkX graph or two-tuple
See also:
is_kl_connected()

References

3.31.2 networkx.algorithms.hybrid.is_kl_connected

is_kl_connected(G, k, l, low_memory=False)
Returns True if and only if G is locally (k, l)-connected.
A graph is locally (k, l)-connected if for each edge (u, v) in the graph there are at least l edge-disjoint
paths of length at most k joining u to v.
Parameters
• G (NetworkX graph) – The graph to test for local (k, l)-connectedness.
• k (integer) – The maximum length of paths to consider. A higher number means a looser
connectivity requirement.
• l (integer) – The number of edge-disjoint paths. A higher number means a stricter connec-
tivity requirement.
• low_memory (bool) – If this is True, this function uses an algorithm that uses slightly more
time but less memory.
Returns Whether the graph is locally (k, l)-connected subgraph.
Return type bool

3.31. Hybrid 345


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
kl_connected_subgraph()

References

3.32 Isolates

Functions for identifying isolate (degree zero) nodes.

is_isolate(G, n) Determines whether a node is an isolate.


isolates(G) Iterator over isolates in the graph.
number_of_isolates(G) Returns the number of isolates in the graph.

3.32.1 networkx.algorithms.isolate.is_isolate

is_isolate(G, n)
Determines whether a node is an isolate.
An isolate is a node with no neighbors (that is, with degree zero). For directed graphs, this means no in-neighbors
and no out-neighbors.
Parameters
• G (NetworkX graph)
• n (node) – A node in G.
Returns is_isolate – True if and only if n has no neighbors.
Return type bool

Examples

>>> G=nx.Graph()
>>> G.add_edge(1,2)
>>> G.add_node(3)
>>> nx.is_isolate(G,2)
False
>>> nx.is_isolate(G,3)
True

3.32.2 networkx.algorithms.isolate.isolates

isolates(G)
Iterator over isolates in the graph.
An isolate is a node with no neighbors (that is, with degree zero). For directed graphs, this means no in-neighbors
and no out-neighbors.
Parameters G (NetworkX graph)
Returns An iterator over the isolates of G.
Return type iterator

346 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

To get a list of all isolates of a graph, use the list constructor:

>>> G = nx.Graph()
>>> G.add_edge(1, 2)
>>> G.add_node(3)
>>> list(nx.isolates(G))
[3]

To remove all isolates in the graph, first create a list of the isolates, then use Graph.
remove_nodes_from():

>>> G.remove_nodes_from(list(nx.isolates(G)))
>>> list(G)
[1, 2]

For digraphs, isolates have zero in-degree and zero out_degre:

>>> G = nx.DiGraph([(0, 1), (1, 2)])


>>> G.add_node(3)
>>> list(nx.isolates(G))
[3]

3.32.3 networkx.algorithms.isolate.number_of_isolates

number_of_isolates(G)
Returns the number of isolates in the graph.
An isolate is a node with no neighbors (that is, with degree zero). For directed graphs, this means no in-neighbors
and no out-neighbors.
Parameters G (NetworkX graph)
Returns The number of degree zero nodes in the graph G.
Return type int

3.33 Isomorphism

is_isomorphic(G1, G2[, node_match, Returns True if the graphs G1 and G2 are isomorphic
edge_match]) and False otherwise.
could_be_isomorphic(G1, G2) Returns False if graphs are definitely not isomorphic.
fast_could_be_isomorphic(G1, G2) Returns False if graphs are definitely not isomorphic.
faster_could_be_isomorphic(G1, G2) Returns False if graphs are definitely not isomorphic.

3.33.1 networkx.algorithms.isomorphism.is_isomorphic

is_isomorphic(G1, G2, node_match=None, edge_match=None)


Returns True if the graphs G1 and G2 are isomorphic and False otherwise.
Parameters
• G1, G2 (graphs) – The two graphs G1 and G2 must be the same type.

3.33. Isomorphism 347


NetworkX Reference, Release 2.3rc1.dev20190222214247

• node_match (callable) – A function that returns True if node n1 in G1 and n2 in G2 should


be considered equal during the isomorphism test. If node_match is not specified then node
attributes are not considered.
The function will be called like
node_match(G1.nodes[n1], G2.nodes[n2]).
That is, the function will receive the node attribute dictionaries for n1 and n2 as inputs.
• edge_match (callable) – A function that returns True if the edge attribute dictionary for
the pair of nodes (u1, v1) in G1 and (u2, v2) in G2 should be considered equal during the
isomorphism test. If edge_match is not specified then edge attributes are not considered.
The function will be called like
edge_match(G1[u1][v1], G2[u2][v2]).
That is, the function will receive the edge attribute dictionaries of the edges under consider-
ation.

Notes

Uses the vf2 algorithm1 .

Examples

>>> import networkx.algorithms.isomorphism as iso

For digraphs G1 and G2, using ‘weight’ edge attribute (default: 1)

>>> G1 = nx.DiGraph()
>>> G2 = nx.DiGraph()
>>> nx.add_path(G1, [1,2,3,4], weight=1)
>>> nx.add_path(G2, [10,20,30,40], weight=2)
>>> em = iso.numerical_edge_match('weight', 1)
>>> nx.is_isomorphic(G1, G2) # no weights considered
True
>>> nx.is_isomorphic(G1, G2, edge_match=em) # match weights
False

For multidigraphs G1 and G2, using ‘fill’ node attribute (default: ‘’)

>>> G1 = nx.MultiDiGraph()
>>> G2 = nx.MultiDiGraph()
>>> G1.add_nodes_from([1,2,3], fill='red')
>>> G2.add_nodes_from([10,20,30,40], fill='red')
>>> nx.add_path(G1, [1,2,3,4], weight=3, linewidth=2.5)
>>> nx.add_path(G2, [10,20,30,40], weight=3)
>>> nm = iso.categorical_node_match('fill', 'red')
>>> nx.is_isomorphic(G1, G2, node_match=nm)
True

For multidigraphs G1 and G2, using ‘weight’ edge attribute (default: 7)


1 L. P. Cordella, P. Foggia, C. Sansone, M. Vento, “An Improved Algorithm for Matching Large Graphs”, 3rd IAPR-TC15 Workshop on Graph-

based Representations in Pattern Recognition, Cuen, pp. 149-159, 2001. http://amalfi.dis.unina.it/graph/db/papers/vf-algorithm.pdf

348 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G1.add_edge(1,2, weight=7)


1
>>> G2.add_edge(10,20)
1
>>> em = iso.numerical_multiedge_match('weight', 7, rtol=1e-6)
>>> nx.is_isomorphic(G1, G2, edge_match=em)
True

For multigraphs G1 and G2, using ‘weight’ and ‘linewidth’ edge attributes with default values 7 and 2.5. Also
using ‘fill’ node attribute with default value ‘red’.

>>> em = iso.numerical_multiedge_match(['weight', 'linewidth'], [7, 2.5])


>>> nm = iso.categorical_node_match('fill', 'red')
>>> nx.is_isomorphic(G1, G2, edge_match=em, node_match=nm)
True

See also:
numerical_node_match(), numerical_edge_match(), numerical_multiedge_match(),
categorical_node_match(), categorical_edge_match(), categorical_multiedge_match()

References

3.33.2 networkx.algorithms.isomorphism.could_be_isomorphic

could_be_isomorphic(G1, G2)
Returns False if graphs are definitely not isomorphic. True does NOT guarantee isomorphism.
Parameters G1, G2 (graphs) – The two graphs G1 and G2 must be the same type.

Notes

Checks for matching degree, triangle, and number of cliques sequences.

3.33.3 networkx.algorithms.isomorphism.fast_could_be_isomorphic

fast_could_be_isomorphic(G1, G2)
Returns False if graphs are definitely not isomorphic.
True does NOT guarantee isomorphism.
Parameters G1, G2 (graphs) – The two graphs G1 and G2 must be the same type.

Notes

Checks for matching degree and triangle sequences.

3.33.4 networkx.algorithms.isomorphism.faster_could_be_isomorphic

faster_could_be_isomorphic(G1, G2)
Returns False if graphs are definitely not isomorphic.
True does NOT guarantee isomorphism.

3.33. Isomorphism 349


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters G1, G2 (graphs) – The two graphs G1 and G2 must be the same type.

Notes

Checks for matching degree sequences.

3.33.5 Advanced Interface to VF2 Algorithm

VF2 Algorithm

VF2 Algorithm

An implementation of VF2 algorithm for graph ismorphism testing.


The simplest interface to use this module is to call networkx.is_isomorphic().

Introduction

The GraphMatcher and DiGraphMatcher are responsible for matching graphs or directed graphs in a predetermined
manner. This usually means a check for an isomorphism, though other checks are also possible. For example, a
subgraph of one graph can be checked for isomorphism to a second graph.
Matching is done via syntactic feasibility. It is also possible to check for semantic feasibility. Feasibility, then, is
defined as the logical AND of the two functions.
To include a semantic check, the (Di)GraphMatcher class should be subclassed, and the semantic_feasibility() function
should be redefined. By default, the semantic feasibility function always returns True. The effect of this is that
semantics are not considered in the matching of G1 and G2.

Examples

Suppose G1 and G2 are isomorphic graphs. Verification is as follows:

>>> from networkx.algorithms import isomorphism


>>> G1 = nx.path_graph(4)
>>> G2 = nx.path_graph(4)
>>> GM = isomorphism.GraphMatcher(G1,G2)
>>> GM.is_isomorphic()
True

GM.mapping stores the isomorphism mapping from G1 to G2.

>>> GM.mapping
{0: 0, 1: 1, 2: 2, 3: 3}

Suppose G1 and G2 are isomorphic directed graphs graphs. Verification is as follows:

>>> G1 = nx.path_graph(4, create_using=nx.DiGraph())


>>> G2 = nx.path_graph(4, create_using=nx.DiGraph())
>>> DiGM = isomorphism.DiGraphMatcher(G1,G2)
>>> DiGM.is_isomorphic()
True

350 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

DiGM.mapping stores the isomorphism mapping from G1 to G2.

>>> DiGM.mapping
{0: 0, 1: 1, 2: 2, 3: 3}

Subgraph Isomorphism

Graph theory literature can be ambiguious about the meaning of the above statement, and we seek to clarify it now.
In the VF2 literature, a mapping M is said to be a graph-subgraph isomorphism iff M is an isomorphism between G2
and a subgraph of G1. Thus, to say that G1 and G2 are graph-subgraph isomorphic is to say that a subgraph of G1 is
isomorphic to G2.
Other literature uses the phrase ‘subgraph isomorphic’ as in ‘G1 does not have a subgraph isomorphic to G2’. Another
use is as an in adverb for isomorphic. Thus, to say that G1 and G2 are subgraph isomorphic is to say that a subgraph
of G1 is isomorphic to G2.
Finally, the term ‘subgraph’ can have multiple meanings. In this context, ‘subgraph’ always means a ‘node-induced
subgraph’. Edge-induced subgraph isomorphisms are not directly supported, but one should be able to perform the
check by making use of nx.line_graph(). For subgraphs which are not induced, the term ‘monomorphism’ is preferred
over ‘isomorphism’. Currently, it is not possible to check for monomorphisms.
Let G=(N,E) be a graph with a set of nodes N and set of edges E.
If G’=(N’,E’) is a subgraph, then: N’ is a subset of N E’ is a subset of E
If G’=(N’,E’) is a node-induced subgraph, then: N’ is a subset of N E’ is the subset of edges in E relating nodes in
N’
If G’=(N’,E’) is an edge-induced subgraph, then: N’ is the subset of nodes in N related by edges in E’ E’ is a subset
of E

References

[1] Luigi P. Cordella, Pasquale Foggia, Carlo Sansone, Mario Vento, “A (Sub)Graph Isomorphism Algorithm for
Matching Large Graphs”, IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 26, no. 10, pp.
1367-1372, Oct., 2004. http://ieeexplore.ieee.org/iel5/34/29305/01323804.pdf
[2] L. P. Cordella, P. Foggia, C. Sansone, M. Vento, “An Improved Algorithm for Matching Large Graphs”, 3rd
IAPR-TC15 Workshop on Graph-based Representations in Pattern Recognition, Cuen, pp. 149-159, 2001.
http://amalfi.dis.unina.it/graph/db/papers/vf-algorithm.pdf
See also:
syntactic_feasibliity, semantic_feasibility

Notes

The implementation handles both directed and undirected graphs as well as multigraphs.
In general, the subgraph isomorphism problem is NP-complete whereas the graph isomorphism problem is most likely
not NP-complete (although no polynomial-time algorithm is known to exist).

Graph Matcher

3.33. Isomorphism 351


NetworkX Reference, Release 2.3rc1.dev20190222214247

GraphMatcher.__init__(G1, G2[, node_match, Initialize graph matcher.


. . . ])
GraphMatcher.initialize() Reinitializes the state of the algorithm.
GraphMatcher.is_isomorphic() Returns True if G1 and G2 are isomorphic graphs.
GraphMatcher.subgraph_is_isomorphic() Returns True if a subgraph of G1 is isomorphic to G2.
GraphMatcher.isomorphisms_iter() Generator over isomorphisms between G1 and G2.
GraphMatcher.subgraph_isomorphisms_iter() Generator over isomorphisms between a subgraph of G1
and G2.
GraphMatcher.candidate_pairs_iter() Iterator over candidate pairs of nodes in G1 and G2.
GraphMatcher.match() Extends the isomorphism mapping.
GraphMatcher.semantic_feasibility(G1_node, Returns True if mapping G1_node to G2_node is se-
...) mantically feasible.
GraphMatcher.syntactic_feasibility(G1_node,Returns True if adding (G1_node, G2_node) is syntac-
...) tically feasible.

networkx.algorithms.isomorphism.GraphMatcher.__init__

GraphMatcher.__init__(G1, G2, node_match=None, edge_match=None)


Initialize graph matcher.
Parameters
• G1, G2 (graph) – The graphs to be tested.
• node_match (callable) – A function that returns True iff node n1 in G1 and n2 in G2 should
be considered equal during the isomorphism test. The function will be called like:

node_match(G1.nodes[n1], G2.nodes[n2])

That is, the function will receive the node attribute dictionaries of the nodes under consider-
ation. If None, then no attributes are considered when testing for an isomorphism.
• edge_match (callable) – A function that returns True iff the edge attribute dictionary for
the pair of nodes (u1, v1) in G1 and (u2, v2) in G2 should be considered equal during the
isomorphism test. The function will be called like:

edge_match(G1[u1][v1], G2[u2][v2])

That is, the function will receive the edge attribute dictionaries of the edges under consider-
ation. If None, then no attributes are considered when testing for an isomorphism.

networkx.algorithms.isomorphism.GraphMatcher.initialize

GraphMatcher.initialize()
Reinitializes the state of the algorithm.
This method should be redefined if using something other than GMState. If only subclassing GraphMatcher, a
redefinition is not necessary.

networkx.algorithms.isomorphism.GraphMatcher.is_isomorphic

GraphMatcher.is_isomorphic()
Returns True if G1 and G2 are isomorphic graphs.

352 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.isomorphism.GraphMatcher.subgraph_is_isomorphic

GraphMatcher.subgraph_is_isomorphic()
Returns True if a subgraph of G1 is isomorphic to G2.

networkx.algorithms.isomorphism.GraphMatcher.isomorphisms_iter

GraphMatcher.isomorphisms_iter()
Generator over isomorphisms between G1 and G2.

networkx.algorithms.isomorphism.GraphMatcher.subgraph_isomorphisms_iter

GraphMatcher.subgraph_isomorphisms_iter()
Generator over isomorphisms between a subgraph of G1 and G2.

networkx.algorithms.isomorphism.GraphMatcher.candidate_pairs_iter

GraphMatcher.candidate_pairs_iter()
Iterator over candidate pairs of nodes in G1 and G2.

networkx.algorithms.isomorphism.GraphMatcher.match

GraphMatcher.match()
Extends the isomorphism mapping.
This function is called recursively to determine if a complete isomorphism can be found between G1 and G2. It
cleans up the class variables after each recursive call. If an isomorphism is found, we yield the mapping.

networkx.algorithms.isomorphism.GraphMatcher.semantic_feasibility

GraphMatcher.semantic_feasibility(G1_node, G2_node)
Returns True if mapping G1_node to G2_node is semantically feasible.

networkx.algorithms.isomorphism.GraphMatcher.syntactic_feasibility

GraphMatcher.syntactic_feasibility(G1_node, G2_node)
Returns True if adding (G1_node, G2_node) is syntactically feasible.
This function returns True if it is adding the candidate pair to the current partial isomorphism mapping is
allowable. The addition is allowable if the inclusion of the candidate pair does not make it impossible for an
isomorphism to be found.

DiGraph Matcher

DiGraphMatcher.__init__(G1, G2[, . . . ]) Initialize graph matcher.


DiGraphMatcher.initialize() Reinitializes the state of the algorithm.
Continued on next page

3.33. Isomorphism 353


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 102 – continued from previous page


DiGraphMatcher.is_isomorphic() Returns True if G1 and G2 are isomorphic graphs.
DiGraphMatcher.subgraph_is_isomorphic() Returns True if a subgraph of G1 is isomorphic to G2.
DiGraphMatcher.isomorphisms_iter() Generator over isomorphisms between G1 and G2.
DiGraphMatcher.subgraph_isomorphisms_iter() Generator over isomorphisms between a subgraph of G1
and G2.
DiGraphMatcher.candidate_pairs_iter() Iterator over candidate pairs of nodes in G1 and G2.
DiGraphMatcher.match() Extends the isomorphism mapping.
DiGraphMatcher.semantic_feasibility(G1_node,Returns True if mapping G1_node to G2_node is se-
...) mantically feasible.
DiGraphMatcher.syntactic_feasibility(. . . )Returns True if adding (G1_node, G2_node) is syntac-
tically feasible.

networkx.algorithms.isomorphism.DiGraphMatcher.__init__

DiGraphMatcher.__init__(G1, G2, node_match=None, edge_match=None)


Initialize graph matcher.
Parameters
• G1, G2 (graph) – The graphs to be tested.
• node_match (callable) – A function that returns True iff node n1 in G1 and n2 in G2 should
be considered equal during the isomorphism test. The function will be called like:

node_match(G1.nodes[n1], G2.nodes[n2])

That is, the function will receive the node attribute dictionaries of the nodes under consider-
ation. If None, then no attributes are considered when testing for an isomorphism.
• edge_match (callable) – A function that returns True iff the edge attribute dictionary for
the pair of nodes (u1, v1) in G1 and (u2, v2) in G2 should be considered equal during the
isomorphism test. The function will be called like:

edge_match(G1[u1][v1], G2[u2][v2])

That is, the function will receive the edge attribute dictionaries of the edges under consider-
ation. If None, then no attributes are considered when testing for an isomorphism.

networkx.algorithms.isomorphism.DiGraphMatcher.initialize

DiGraphMatcher.initialize()
Reinitializes the state of the algorithm.
This method should be redefined if using something other than DiGMState. If only subclassing GraphMatcher,
a redefinition is not necessary.

networkx.algorithms.isomorphism.DiGraphMatcher.is_isomorphic

DiGraphMatcher.is_isomorphic()
Returns True if G1 and G2 are isomorphic graphs.

354 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.isomorphism.DiGraphMatcher.subgraph_is_isomorphic

DiGraphMatcher.subgraph_is_isomorphic()
Returns True if a subgraph of G1 is isomorphic to G2.

networkx.algorithms.isomorphism.DiGraphMatcher.isomorphisms_iter

DiGraphMatcher.isomorphisms_iter()
Generator over isomorphisms between G1 and G2.

networkx.algorithms.isomorphism.DiGraphMatcher.subgraph_isomorphisms_iter

DiGraphMatcher.subgraph_isomorphisms_iter()
Generator over isomorphisms between a subgraph of G1 and G2.

networkx.algorithms.isomorphism.DiGraphMatcher.candidate_pairs_iter

DiGraphMatcher.candidate_pairs_iter()
Iterator over candidate pairs of nodes in G1 and G2.

networkx.algorithms.isomorphism.DiGraphMatcher.match

DiGraphMatcher.match()
Extends the isomorphism mapping.
This function is called recursively to determine if a complete isomorphism can be found between G1 and G2. It
cleans up the class variables after each recursive call. If an isomorphism is found, we yield the mapping.

networkx.algorithms.isomorphism.DiGraphMatcher.semantic_feasibility

DiGraphMatcher.semantic_feasibility(G1_node, G2_node)
Returns True if mapping G1_node to G2_node is semantically feasible.

networkx.algorithms.isomorphism.DiGraphMatcher.syntactic_feasibility

DiGraphMatcher.syntactic_feasibility(G1_node, G2_node)
Returns True if adding (G1_node, G2_node) is syntactically feasible.
This function returns True if it is adding the candidate pair to the current partial isomorphism mapping is
allowable. The addition is allowable if the inclusion of the candidate pair does not make it impossible for an
isomorphism to be found.

Match helpers

3.33. Isomorphism 355


NetworkX Reference, Release 2.3rc1.dev20190222214247

categorical_node_match(attr, default) Returns a comparison function for a categorical node


attribute.
categorical_edge_match(attr, default) Returns a comparison function for a categorical edge at-
tribute.
categorical_multiedge_match(attr, default) Returns a comparison function for a categorical edge at-
tribute.
numerical_node_match(attr, default[, rtol, atol]) Returns a comparison function for a numerical node at-
tribute.
numerical_edge_match(attr, default[, rtol, atol]) Returns a comparison function for a numerical edge at-
tribute.
numerical_multiedge_match(attr, default[, Returns a comparison function for a numerical edge at-
. . . ]) tribute.
generic_node_match(attr, default, op) Returns a comparison function for a generic attribute.
generic_edge_match(attr, default, op) Returns a comparison function for a generic attribute.
generic_multiedge_match(attr, default, op) Returns a comparison function for a generic attribute.

networkx.algorithms.isomorphism.categorical_node_match

categorical_node_match(attr, default)
Returns a comparison function for a categorical node attribute.
The value(s) of the attr(s) must be hashable and comparable via the == operator since they are placed into a
set([]) object. If the sets from G1 and G2 are the same, then the constructed function returns True.
Parameters
• attr (string | list) – The categorical node attribute to compare, or a list of categorical node
attributes to compare.
• default (value | list) – The default value for the categorical node attribute, or a list of default
values for the categorical node attributes.
Returns match – The customized, categorical node_match function.
Return type function

Examples

>>> import networkx.algorithms.isomorphism as iso


>>> nm = iso.categorical_node_match('size', 1)
>>> nm = iso.categorical_node_match(['color', 'size'], ['red', 2])

networkx.algorithms.isomorphism.categorical_edge_match

categorical_edge_match(attr, default)
Returns a comparison function for a categorical edge attribute.
The value(s) of the attr(s) must be hashable and comparable via the == operator since they are placed into a
set([]) object. If the sets from G1 and G2 are the same, then the constructed function returns True.
Parameters
• attr (string | list) – The categorical edge attribute to compare, or a list of categorical edge
attributes to compare.

356 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• default (value | list) – The default value for the categorical edge attribute, or a list of default
values for the categorical edge attributes.
Returns match – The customized, categorical edge_match function.
Return type function

Examples

>>> import networkx.algorithms.isomorphism as iso


>>> nm = iso.categorical_edge_match('size', 1)
>>> nm = iso.categorical_edge_match(['color', 'size'], ['red', 2])

networkx.algorithms.isomorphism.categorical_multiedge_match

categorical_multiedge_match(attr, default)
Returns a comparison function for a categorical edge attribute.
The value(s) of the attr(s) must be hashable and comparable via the == operator since they are placed into a
set([]) object. If the sets from G1 and G2 are the same, then the constructed function returns True.
Parameters
• attr (string | list) – The categorical edge attribute to compare, or a list of categorical edge
attributes to compare.
• default (value | list) – The default value for the categorical edge attribute, or a list of default
values for the categorical edge attributes.
Returns match – The customized, categorical edge_match function.
Return type function

Examples

>>> import networkx.algorithms.isomorphism as iso


>>> nm = iso.categorical_multiedge_match('size', 1)
>>> nm = iso.categorical_multiedge_match(['color', 'size'], ['red', 2])

networkx.algorithms.isomorphism.numerical_node_match

numerical_node_match(attr, default, rtol=1e-05, atol=1e-08)


Returns a comparison function for a numerical node attribute.
The value(s) of the attr(s) must be numerical and sortable. If the sorted list of values from G1 and G2 are the
same within some tolerance, then the constructed function returns True.
Parameters
• attr (string | list) – The numerical node attribute to compare, or a list of numerical node
attributes to compare.
• default (value | list) – The default value for the numerical node attribute, or a list of default
values for the numerical node attributes.
• rtol (float) – The relative error tolerance.

3.33. Isomorphism 357


NetworkX Reference, Release 2.3rc1.dev20190222214247

• atol (float) – The absolute error tolerance.


Returns match – The customized, numerical node_match function.
Return type function

Examples

>>> import networkx.algorithms.isomorphism as iso


>>> nm = iso.numerical_node_match('weight', 1.0)
>>> nm = iso.numerical_node_match(['weight', 'linewidth'], [.25, .5])

networkx.algorithms.isomorphism.numerical_edge_match

numerical_edge_match(attr, default, rtol=1e-05, atol=1e-08)


Returns a comparison function for a numerical edge attribute.
The value(s) of the attr(s) must be numerical and sortable. If the sorted list of values from G1 and G2 are the
same within some tolerance, then the constructed function returns True.
Parameters
• attr (string | list) – The numerical edge attribute to compare, or a list of numerical edge
attributes to compare.
• default (value | list) – The default value for the numerical edge attribute, or a list of default
values for the numerical edge attributes.
• rtol (float) – The relative error tolerance.
• atol (float) – The absolute error tolerance.
Returns match – The customized, numerical edge_match function.
Return type function

Examples

>>> import networkx.algorithms.isomorphism as iso


>>> nm = iso.numerical_edge_match('weight', 1.0)
>>> nm = iso.numerical_edge_match(['weight', 'linewidth'], [.25, .5])

networkx.algorithms.isomorphism.numerical_multiedge_match

numerical_multiedge_match(attr, default, rtol=1e-05, atol=1e-08)


Returns a comparison function for a numerical edge attribute.
The value(s) of the attr(s) must be numerical and sortable. If the sorted list of values from G1 and G2 are the
same within some tolerance, then the constructed function returns True.
Parameters
• attr (string | list) – The numerical edge attribute to compare, or a list of numerical edge
attributes to compare.
• default (value | list) – The default value for the numerical edge attribute, or a list of default
values for the numerical edge attributes.

358 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• rtol (float) – The relative error tolerance.


• atol (float) – The absolute error tolerance.
Returns match – The customized, numerical edge_match function.
Return type function

Examples

>>> import networkx.algorithms.isomorphism as iso


>>> nm = iso.numerical_multiedge_match('weight', 1.0)
>>> nm = iso.numerical_multiedge_match(['weight', 'linewidth'], [.25, .5])

networkx.algorithms.isomorphism.generic_node_match

generic_node_match(attr, default, op)


Returns a comparison function for a generic attribute.
The value(s) of the attr(s) are compared using the specified operators. If all the attributes are equal, then the
constructed function returns True.
Parameters
• attr (string | list) – The node attribute to compare, or a list of node attributes to compare.
• default (value | list) – The default value for the node attribute, or a list of default values for
the node attributes.
• op (callable | list) – The operator to use when comparing attribute values, or a list of opera-
tors to use when comparing values for each attribute.
Returns match – The customized, generic node_match function.
Return type function

Examples

>>> from operator import eq


>>> from networkx.algorithms.isomorphism.matchhelpers import close
>>> from networkx.algorithms.isomorphism import generic_node_match
>>> nm = generic_node_match('weight', 1.0, close)
>>> nm = generic_node_match('color', 'red', eq)
>>> nm = generic_node_match(['weight', 'color'], [1.0, 'red'], [close, eq])

networkx.algorithms.isomorphism.generic_edge_match

generic_edge_match(attr, default, op)


Returns a comparison function for a generic attribute.
The value(s) of the attr(s) are compared using the specified operators. If all the attributes are equal, then the
constructed function returns True.
Parameters
• attr (string | list) – The edge attribute to compare, or a list of edge attributes to compare.

3.33. Isomorphism 359


NetworkX Reference, Release 2.3rc1.dev20190222214247

• default (value | list) – The default value for the edge attribute, or a list of default values for
the edge attributes.
• op (callable | list) – The operator to use when comparing attribute values, or a list of opera-
tors to use when comparing values for each attribute.
Returns match – The customized, generic edge_match function.
Return type function

Examples

>>> from operator import eq


>>> from networkx.algorithms.isomorphism.matchhelpers import close
>>> from networkx.algorithms.isomorphism import generic_edge_match
>>> nm = generic_edge_match('weight', 1.0, close)
>>> nm = generic_edge_match('color', 'red', eq)
>>> nm = generic_edge_match(['weight', 'color'], [1.0, 'red'], [close, eq])

networkx.algorithms.isomorphism.generic_multiedge_match

generic_multiedge_match(attr, default, op)


Returns a comparison function for a generic attribute.
The value(s) of the attr(s) are compared using the specified operators. If all the attributes are equal, then the
constructed function returns True. Potentially, the constructed edge_match function can be slow since it must
verify that no isomorphism exists between the multiedges before it returns False.
Parameters
• attr (string | list) – The edge attribute to compare, or a list of node attributes to compare.
• default (value | list) – The default value for the edge attribute, or a list of default values for
the dgeattributes.
• op (callable | list) – The operator to use when comparing attribute values, or a list of opera-
tors to use when comparing values for each attribute.
Returns match – The customized, generic edge_match function.
Return type function

Examples

>>> from operator import eq


>>> from networkx.algorithms.isomorphism.matchhelpers import close
>>> from networkx.algorithms.isomorphism import generic_node_match
>>> nm = generic_node_match('weight', 1.0, close)
>>> nm = generic_node_match('color', 'red', eq)
>>> nm = generic_node_match(['weight', 'color'],
... [1.0, 'red'],
... [close, eq])
...

360 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.34 Link Analysis

3.34.1 PageRank

PageRank analysis of graph structure.

pagerank(G[, alpha, personalization, . . . ]) Returns the PageRank of the nodes in the graph.
pagerank_numpy(G[, alpha, personalization, . . . ]) Returns the PageRank of the nodes in the graph.
pagerank_scipy(G[, alpha, personalization, . . . ]) Returns the PageRank of the nodes in the graph.
google_matrix(G[, alpha, personalization, . . . ]) Returns the Google matrix of the graph.

networkx.algorithms.link_analysis.pagerank_alg.pagerank

pagerank(G, alpha=0.85, personalization=None, max_iter=100, tol=1e-06, nstart=None, weight=’weight’,


dangling=None)
Returns the PageRank of the nodes in the graph.
PageRank computes a ranking of the nodes in the graph G based on the structure of the incoming links. It was
originally designed as an algorithm to rank web pages.
Parameters
• G (graph) – A NetworkX graph. Undirected graphs will be converted to a directed graph
with two directed edges for each undirected edge.
• alpha (float, optional) – Damping parameter for PageRank, default=0.85.
• personalization (dict, optional) – The “personalization vector” consisting of a dictionary
with a key some subset of graph nodes and personalization value each of those. At least one
personalization value must be non-zero. If not specfiied, a nodes personalization value will
be zero. By default, a uniform distribution is used.
• max_iter (integer, optional) – Maximum number of iterations in power method eigenvalue
solver.
• tol (float, optional) – Error tolerance used to check convergence in power method solver.
• nstart (dictionary, optional) – Starting value of PageRank iteration for each node.
• weight (key, optional) – Edge data key to use as weight. If None weights are set to 1.
• dangling (dict, optional) – The outedges to be assigned to any “dangling” nodes, i.e., nodes
without any outedges. The dict key is the node the outedge points to and the dict value
is the weight of that outedge. By default, dangling nodes are given outedges according to
the personalization vector (uniform if not specified). This must be selected to result in an
irreducible transition matrix (see notes under google_matrix). It may be common to have
the dangling dict to be the same as the personalization dict.
Returns pagerank – Dictionary of nodes with PageRank as value
Return type dictionary

Examples

>>> G = nx.DiGraph(nx.path_graph(4))
>>> pr = nx.pagerank(G, alpha=0.9)

3.34. Link Analysis 361


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The eigenvector calculation is done by the power iteration method and has no guarantee of convergence. The
iteration will stop after an error tolerance of len(G) * tol has been reached. If the number of iterations
exceed max_iter, a networkx.exception.PowerIterationFailedConvergence exception is
raised.
The PageRank algorithm was designed for directed graphs but this algorithm does not check if the input graph
is directed and will execute on undirected graphs by converting each edge in the directed graph to two edges.
See also:
pagerank_numpy(), pagerank_scipy(), google_matrix()

Raises PowerIterationFailedConvergence – If the algorithm fails to converge to the


specified tolerance within the specified number of iterations of the power iteration method.

References

networkx.algorithms.link_analysis.pagerank_alg.pagerank_numpy

pagerank_numpy(G, alpha=0.85, personalization=None, weight=’weight’, dangling=None)


Returns the PageRank of the nodes in the graph.
PageRank computes a ranking of the nodes in the graph G based on the structure of the incoming links. It was
originally designed as an algorithm to rank web pages.
Parameters
• G (graph) – A NetworkX graph. Undirected graphs will be converted to a directed graph
with two directed edges for each undirected edge.
• alpha (float, optional) – Damping parameter for PageRank, default=0.85.
• personalization (dict, optional) – The “personalization vector” consisting of a dictionary
with a key some subset of graph nodes and personalization value each of those. At least one
personalization value must be non-zero. If not specfiied, a nodes personalization value will
be zero. By default, a uniform distribution is used.
• weight (key, optional) – Edge data key to use as weight. If None weights are set to 1.
• dangling (dict, optional) – The outedges to be assigned to any “dangling” nodes, i.e., nodes
without any outedges. The dict key is the node the outedge points to and the dict value
is the weight of that outedge. By default, dangling nodes are given outedges according to
the personalization vector (uniform if not specified) This must be selected to result in an
irreducible transition matrix (see notes under google_matrix). It may be common to have
the dangling dict to be the same as the personalization dict.
Returns pagerank – Dictionary of nodes with PageRank as value.
Return type dictionary

Examples

>>> G = nx.DiGraph(nx.path_graph(4))
>>> pr = nx.pagerank_numpy(G, alpha=0.9)

362 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The eigenvector calculation uses NumPy’s interface to the LAPACK eigenvalue solvers. This will be the fastest
and most accurate for small graphs.
This implementation works with Multi(Di)Graphs. For multigraphs the weight between two nodes is set to be
the sum of all edge weights between those nodes.
See also:
pagerank(), pagerank_scipy(), google_matrix()

References

networkx.algorithms.link_analysis.pagerank_alg.pagerank_scipy

pagerank_scipy(G, alpha=0.85, personalization=None, max_iter=100, tol=1e-06, weight=’weight’, dan-


gling=None)
Returns the PageRank of the nodes in the graph.
PageRank computes a ranking of the nodes in the graph G based on the structure of the incoming links. It was
originally designed as an algorithm to rank web pages.
Parameters
• G (graph) – A NetworkX graph. Undirected graphs will be converted to a directed graph
with two directed edges for each undirected edge.
• alpha (float, optional) – Damping parameter for PageRank, default=0.85.
• personalization (dict, optional) – The “personalization vector” consisting of a dictionary
with a key some subset of graph nodes and personalization value each of those. At least one
personalization value must be non-zero. If not specfiied, a nodes personalization value will
be zero. By default, a uniform distribution is used.
• max_iter (integer, optional) – Maximum number of iterations in power method eigenvalue
solver.
• tol (float, optional) – Error tolerance used to check convergence in power method solver.
• weight (key, optional) – Edge data key to use as weight. If None weights are set to 1.
• dangling (dict, optional) – The outedges to be assigned to any “dangling” nodes, i.e., nodes
without any outedges. The dict key is the node the outedge points to and the dict value
is the weight of that outedge. By default, dangling nodes are given outedges according to
the personalization vector (uniform if not specified) This must be selected to result in an
irreducible transition matrix (see notes under google_matrix). It may be common to have
the dangling dict to be the same as the personalization dict.
Returns pagerank – Dictionary of nodes with PageRank as value
Return type dictionary

Examples

>>> G = nx.DiGraph(nx.path_graph(4))
>>> pr = nx.pagerank_scipy(G, alpha=0.9)

3.34. Link Analysis 363


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The eigenvector calculation uses power iteration with a SciPy sparse matrix representation.
This implementation works with Multi(Di)Graphs. For multigraphs the weight between two nodes is set to be
the sum of all edge weights between those nodes.
See also:
pagerank(), pagerank_numpy(), google_matrix()

Raises PowerIterationFailedConvergence – If the algorithm fails to converge to the


specified tolerance within the specified number of iterations of the power iteration method.

References

networkx.algorithms.link_analysis.pagerank_alg.google_matrix

google_matrix(G, alpha=0.85, personalization=None, nodelist=None, weight=’weight’, dan-


gling=None)
Returns the Google matrix of the graph.
Parameters
• G (graph) – A NetworkX graph. Undirected graphs will be converted to a directed graph
with two directed edges for each undirected edge.
• alpha (float) – The damping factor.
• personalization (dict, optional) – The “personalization vector” consisting of a dictionary
with a key some subset of graph nodes and personalization value each of those. At least one
personalization value must be non-zero. If not specfiied, a nodes personalization value will
be zero. By default, a uniform distribution is used.
• nodelist (list, optional) – The rows and columns are ordered according to the nodes in
nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• weight (key, optional) – Edge data key to use as weight. If None weights are set to 1.
• dangling (dict, optional) – The outedges to be assigned to any “dangling” nodes, i.e., nodes
without any outedges. The dict key is the node the outedge points to and the dict value
is the weight of that outedge. By default, dangling nodes are given outedges according to
the personalization vector (uniform if not specified) This must be selected to result in an
irreducible transition matrix (see notes below). It may be common to have the dangling dict
to be the same as the personalization dict.
Returns A – Google matrix of the graph
Return type NumPy matrix

Notes

The matrix returned represents the transition matrix that describes the Markov chain used in PageRank. For
PageRank to converge to a unique solution (i.e., a unique stationary distribution in a Markov chain), the transition
matrix must be irreducible. In other words, it must be that there exists a path between every pair of nodes in the
graph, or else there is the potential of “rank sinks.”
This implementation works with Multi(Di)Graphs. For multigraphs the weight between two nodes is set to be
the sum of all edge weights between those nodes.

364 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
pagerank(), pagerank_numpy(), pagerank_scipy()

3.34.2 Hits

Hubs and authorities analysis of graph structure.

hits(G[, max_iter, tol, nstart, normalized]) Returns HITS hubs and authorities values for nodes.
hits_numpy(G[, normalized]) Returns HITS hubs and authorities values for nodes.
hits_scipy(G[, max_iter, tol, normalized]) Returns HITS hubs and authorities values for nodes.
hub_matrix(G[, nodelist]) Returns the HITS hub matrix.
authority_matrix(G[, nodelist]) Returns the HITS authority matrix.

networkx.algorithms.link_analysis.hits_alg.hits

hits(G, max_iter=100, tol=1e-08, nstart=None, normalized=True)


Returns HITS hubs and authorities values for nodes.
The HITS algorithm computes two numbers for a node. Authorities estimates the node value based on the
incoming links. Hubs estimates the node value based on outgoing links.
Parameters
• G (graph) – A NetworkX graph
• max_iter (integer, optional) – Maximum number of iterations in power method.
• tol (float, optional) – Error tolerance used to check convergence in power method iteration.
• nstart (dictionary, optional) – Starting value of each node for power method iteration.
• normalized (bool (default=True)) – Normalize results by the sum of all of the values.
Returns (hubs,authorities) – Two dictionaries keyed by node containing the hub and authority
values.
Return type two-tuple of dictionaries
Raises PowerIterationFailedConvergence – If the algorithm fails to converge to the
specified tolerance within the specified number of iterations of the power iteration method.

Examples

>>> G=nx.path_graph(4)
>>> h,a=nx.hits(G)

Notes

The eigenvector calculation is done by the power iteration method and has no guarantee of convergence. The
iteration will stop after max_iter iterations or an error tolerance of number_of_nodes(G)*tol has been reached.
The HITS algorithm was designed for directed graphs but this algorithm does not check if the input graph is
directed and will execute on undirected graphs.

3.34. Link Analysis 365


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

networkx.algorithms.link_analysis.hits_alg.hits_numpy

hits_numpy(G, normalized=True)
Returns HITS hubs and authorities values for nodes.
The HITS algorithm computes two numbers for a node. Authorities estimates the node value based on the
incoming links. Hubs estimates the node value based on outgoing links.
Parameters
• G (graph) – A NetworkX graph
• normalized (bool (default=True)) – Normalize results by the sum of all of the values.
Returns (hubs,authorities) – Two dictionaries keyed by node containing the hub and authority
values.
Return type two-tuple of dictionaries

Examples

>>> G=nx.path_graph(4)
>>> h,a=nx.hits(G)

Notes

The eigenvector calculation uses NumPy’s interface to LAPACK.


The HITS algorithm was designed for directed graphs but this algorithm does not check if the input graph is
directed and will execute on undirected graphs.

References

networkx.algorithms.link_analysis.hits_alg.hits_scipy

hits_scipy(G, max_iter=100, tol=1e-06, normalized=True)


Returns HITS hubs and authorities values for nodes.
The HITS algorithm computes two numbers for a node. Authorities estimates the node value based on the
incoming links. Hubs estimates the node value based on outgoing links.
Parameters
• G (graph) – A NetworkX graph
• max_iter (integer, optional) – Maximum number of iterations in power method.
• tol (float, optional) – Error tolerance used to check convergence in power method iteration.
• nstart (dictionary, optional) – Starting value of each node for power method iteration.
• normalized (bool (default=True)) – Normalize results by the sum of all of the values.
Returns (hubs,authorities) – Two dictionaries keyed by node containing the hub and authority
values.
Return type two-tuple of dictionaries

366 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G=nx.path_graph(4)
>>> h,a=nx.hits(G)

Notes

This implementation uses SciPy sparse matrices.


The eigenvector calculation is done by the power iteration method and has no guarantee of convergence. The
iteration will stop after max_iter iterations or an error tolerance of number_of_nodes(G)*tol has been reached.
The HITS algorithm was designed for directed graphs but this algorithm does not check if the input graph is
directed and will execute on undirected graphs.
Raises PowerIterationFailedConvergence – If the algorithm fails to converge to the
specified tolerance within the specified number of iterations of the power iteration method.

References

networkx.algorithms.link_analysis.hits_alg.hub_matrix

hub_matrix(G, nodelist=None)
Returns the HITS hub matrix.

networkx.algorithms.link_analysis.hits_alg.authority_matrix

authority_matrix(G, nodelist=None)
Returns the HITS authority matrix.

3.35 Link Prediction

Link prediction algorithms.

resource_allocation_index(G[, ebunch]) Compute the resource allocation index of all node pairs
in ebunch.
jaccard_coefficient(G[, ebunch]) Compute the Jaccard coefficient of all node pairs in
ebunch.
adamic_adar_index(G[, ebunch]) Compute the Adamic-Adar index of all node pairs in
ebunch.
preferential_attachment(G[, ebunch]) Compute the preferential attachment score of all node
pairs in ebunch.
cn_soundarajan_hopcroft(G[, ebunch, com- Count the number of common neighbors of all node
munity]) pairs in ebunch
ra_index_soundarajan_hopcroft(G[, Compute the resource allocation index of all node pairs
ebunch, . . . ]) in ebunch using community information.
within_inter_cluster(G[, ebunch, delta, . . . ]) Compute the ratio of within- and inter-cluster common
neighbors of all node pairs in ebunch.

3.35. Link Prediction 367


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.35.1 networkx.algorithms.link_prediction.resource_allocation_index

resource_allocation_index(G, ebunch=None)
Compute the resource allocation index of all node pairs in ebunch.
Resource allocation index of u and v is defined as
∑︁ 1
|Γ(𝑤)|
𝑤∈Γ(𝑢)∩Γ(𝑣)

where Γ(𝑢) denotes the set of neighbors of 𝑢.


Parameters
• G (graph) – A NetworkX undirected graph.
• ebunch (iterable of node pairs, optional (default = None)) – Resource allocation index will
be computed for each pair of nodes given in the iterable. The pairs must be given as 2-tuples
(u, v) where u and v are nodes in the graph. If ebunch is None then all non-existent edges in
the graph will be used. Default value: None.
Returns piter – An iterator of 3-tuples in the form (u, v, p) where (u, v) is a pair of nodes and p is
their resource allocation index.
Return type iterator

Examples

>>> import networkx as nx


>>> G = nx.complete_graph(5)
>>> preds = nx.resource_allocation_index(G, [(0, 1), (2, 3)])
>>> for u, v, p in preds:
... '(%d, %d) -> %.8f' % (u, v, p)
...
'(0, 1) -> 0.75000000'
'(2, 3) -> 0.75000000'

References

3.35.2 networkx.algorithms.link_prediction.jaccard_coefficient

jaccard_coefficient(G, ebunch=None)
Compute the Jaccard coefficient of all node pairs in ebunch.
Jaccard coefficient of nodes u and v is defined as
|Γ(𝑢) ∩ Γ(𝑣)|
|Γ(𝑢) ∪ Γ(𝑣)|
where Γ(𝑢) denotes the set of neighbors of 𝑢.
Parameters
• G (graph) – A NetworkX undirected graph.
• ebunch (iterable of node pairs, optional (default = None)) – Jaccard coefficient will be
computed for each pair of nodes given in the iterable. The pairs must be given as 2-tuples
(u, v) where u and v are nodes in the graph. If ebunch is None then all non-existent edges in
the graph will be used. Default value: None.

368 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns piter – An iterator of 3-tuples in the form (u, v, p) where (u, v) is a pair of nodes and p is
their Jaccard coefficient.
Return type iterator

Examples

>>> import networkx as nx


>>> G = nx.complete_graph(5)
>>> preds = nx.jaccard_coefficient(G, [(0, 1), (2, 3)])
>>> for u, v, p in preds:
... '(%d, %d) -> %.8f' % (u, v, p)
...
'(0, 1) -> 0.60000000'
'(2, 3) -> 0.60000000'

References

3.35.3 networkx.algorithms.link_prediction.adamic_adar_index

adamic_adar_index(G, ebunch=None)
Compute the Adamic-Adar index of all node pairs in ebunch.
Adamic-Adar index of u and v is defined as
∑︁ 1
log |Γ(𝑤)|
𝑤∈Γ(𝑢)∩Γ(𝑣)

where Γ(𝑢) denotes the set of neighbors of 𝑢.


Parameters
• G (graph) – NetworkX undirected graph.
• ebunch (iterable of node pairs, optional (default = None)) – Adamic-Adar index will be
computed for each pair of nodes given in the iterable. The pairs must be given as 2-tuples
(u, v) where u and v are nodes in the graph. If ebunch is None then all non-existent edges in
the graph will be used. Default value: None.
Returns piter – An iterator of 3-tuples in the form (u, v, p) where (u, v) is a pair of nodes and p is
their Adamic-Adar index.
Return type iterator

Examples

>>> import networkx as nx


>>> G = nx.complete_graph(5)
>>> preds = nx.adamic_adar_index(G, [(0, 1), (2, 3)])
>>> for u, v, p in preds:
... '(%d, %d) -> %.8f' % (u, v, p)
...
'(0, 1) -> 2.16404256'
'(2, 3) -> 2.16404256'

3.35. Link Prediction 369


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.35.4 networkx.algorithms.link_prediction.preferential_attachment

preferential_attachment(G, ebunch=None)
Compute the preferential attachment score of all node pairs in ebunch.
Preferential attachment score of u and v is defined as

|Γ(𝑢)||Γ(𝑣)|

where Γ(𝑢) denotes the set of neighbors of 𝑢.


Parameters
• G (graph) – NetworkX undirected graph.
• ebunch (iterable of node pairs, optional (default = None)) – Preferential attachment score
will be computed for each pair of nodes given in the iterable. The pairs must be given as
2-tuples (u, v) where u and v are nodes in the graph. If ebunch is None then all non-existent
edges in the graph will be used. Default value: None.
Returns piter – An iterator of 3-tuples in the form (u, v, p) where (u, v) is a pair of nodes and p is
their preferential attachment score.
Return type iterator

Examples

>>> import networkx as nx


>>> G = nx.complete_graph(5)
>>> preds = nx.preferential_attachment(G, [(0, 1), (2, 3)])
>>> for u, v, p in preds:
... '(%d, %d) -> %d' % (u, v, p)
...
'(0, 1) -> 16'
'(2, 3) -> 16'

References

3.35.5 networkx.algorithms.link_prediction.cn_soundarajan_hopcroft

cn_soundarajan_hopcroft(G, ebunch=None, community=’community’)


Count the number of common neighbors of all node pairs in ebunch using community information.
For two nodes 𝑢 and 𝑣, this function computes the number of common neighbors and bonus one for each
common neighbor belonging to the same community as 𝑢 and 𝑣. Mathematically,
∑︁
|Γ(𝑢) ∩ Γ(𝑣)| + 𝑓 (𝑤)
𝑤∈Γ(𝑢)∩Γ(𝑣)

where 𝑓 (𝑤) equals 1 if 𝑤 belongs to the same community as 𝑢 and 𝑣 or 0 otherwise and Γ(𝑢) denotes the set of
neighbors of 𝑢.
Parameters
• G (graph) – A NetworkX undirected graph.

370 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• ebunch (iterable of node pairs, optional (default = None)) – The score will be computed
for each pair of nodes given in the iterable. The pairs must be given as 2-tuples (u, v) where
u and v are nodes in the graph. If ebunch is None then all non-existent edges in the graph
will be used. Default value: None.
• community (string, optional (default = ‘community’)) – Nodes attribute name containing
the community information. G[u][community] identifies which community u belongs to.
Each node belongs to at most one community. Default value: ‘community’.
Returns piter – An iterator of 3-tuples in the form (u, v, p) where (u, v) is a pair of nodes and p is
their score.
Return type iterator

Examples

>>> import networkx as nx


>>> G = nx.path_graph(3)
>>> G.nodes[0]['community'] = 0
>>> G.nodes[1]['community'] = 0
>>> G.nodes[2]['community'] = 0
>>> preds = nx.cn_soundarajan_hopcroft(G, [(0, 2)])
>>> for u, v, p in preds:
... '(%d, %d) -> %d' % (u, v, p)
'(0, 2) -> 2'

References

3.35.6 networkx.algorithms.link_prediction.ra_index_soundarajan_hopcroft

ra_index_soundarajan_hopcroft(G, ebunch=None, community=’community’)


Compute the resource allocation index of all node pairs in ebunch using community information.
For two nodes 𝑢 and 𝑣, this function computes the resource allocation index considering only common neighbors
belonging to the same community as 𝑢 and 𝑣. Mathematically,
∑︁ 𝑓 (𝑤)
|Γ(𝑤)|
𝑤∈Γ(𝑢)∩Γ(𝑣)

where 𝑓 (𝑤) equals 1 if 𝑤 belongs to the same community as 𝑢 and 𝑣 or 0 otherwise and Γ(𝑢) denotes the set of
neighbors of 𝑢.
Parameters
• G (graph) – A NetworkX undirected graph.
• ebunch (iterable of node pairs, optional (default = None)) – The score will be computed
for each pair of nodes given in the iterable. The pairs must be given as 2-tuples (u, v) where
u and v are nodes in the graph. If ebunch is None then all non-existent edges in the graph
will be used. Default value: None.
• community (string, optional (default = ‘community’)) – Nodes attribute name containing
the community information. G[u][community] identifies which community u belongs to.
Each node belongs to at most one community. Default value: ‘community’.
Returns piter – An iterator of 3-tuples in the form (u, v, p) where (u, v) is a pair of nodes and p is
their score.

3.35. Link Prediction 371


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type iterator

Examples

>>> import networkx as nx


>>> G = nx.Graph()
>>> G.add_edges_from([(0, 1), (0, 2), (1, 3), (2, 3)])
>>> G.nodes[0]['community'] = 0
>>> G.nodes[1]['community'] = 0
>>> G.nodes[2]['community'] = 1
>>> G.nodes[3]['community'] = 0
>>> preds = nx.ra_index_soundarajan_hopcroft(G, [(0, 3)])
>>> for u, v, p in preds:
... '(%d, %d) -> %.8f' % (u, v, p)
'(0, 3) -> 0.50000000'

References

3.35.7 networkx.algorithms.link_prediction.within_inter_cluster

within_inter_cluster(G, ebunch=None, delta=0.001, community=’community’)


Compute the ratio of within- and inter-cluster common neighbors of all node pairs in ebunch.
For two nodes u and v, if a common neighbor w belongs to the same community as them, w is considered as
within-cluster common neighbor of u and v. Otherwise, it is considered as inter-cluster common neighbor of
u and v. The ratio between the size of the set of within- and inter-cluster common neighbors is defined as the
WIC measure.1
Parameters
• G (graph) – A NetworkX undirected graph.
• ebunch (iterable of node pairs, optional (default = None)) – The WIC measure will be
computed for each pair of nodes given in the iterable. The pairs must be given as 2-tuples
(u, v) where u and v are nodes in the graph. If ebunch is None then all non-existent edges in
the graph will be used. Default value: None.
• delta (float, optional (default = 0.001)) – Value to prevent division by zero in case there
is no inter-cluster common neighbor between two nodes. See1 for details. Default value:
0.001.
• community (string, optional (default = ‘community’)) – Nodes attribute name containing
the community information. G[u][community] identifies which community u belongs to.
Each node belongs to at most one community. Default value: ‘community’.
Returns piter – An iterator of 3-tuples in the form (u, v, p) where (u, v) is a pair of nodes and p is
their WIC measure.
Return type iterator
1 Jorge Carlos Valverde-Rebaza and Alneu de Andrade Lopes. Link prediction in complex networks based on cluster information. In Proceedings

of the 21st Brazilian conference on Advances in Artificial Intelligence (SBIA‘12) https://doi.org/10.1007/978-3-642-34459-6_10

372 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> import networkx as nx


>>> G = nx.Graph()
>>> G.add_edges_from([(0, 1), (0, 2), (0, 3), (1, 4), (2, 4), (3, 4)])
>>> G.nodes[0]['community'] = 0
>>> G.nodes[1]['community'] = 1
>>> G.nodes[2]['community'] = 0
>>> G.nodes[3]['community'] = 0
>>> G.nodes[4]['community'] = 0
>>> preds = nx.within_inter_cluster(G, [(0, 4)])
>>> for u, v, p in preds:
... '(%d, %d) -> %.8f' % (u, v, p)
...
'(0, 4) -> 1.99800200'
>>> preds = nx.within_inter_cluster(G, [(0, 4)], delta=0.5)
>>> for u, v, p in preds:
... '(%d, %d) -> %.8f' % (u, v, p)
...
'(0, 4) -> 1.33333333'

References

3.36 Lowest Common Ancestor

Algorithms for finding the lowest common ancestor of trees and DAGs.

all_pairs_lowest_common_ancestor(G[, Compute the lowest common ancestor for pairs of


pairs]) nodes.
tree_all_pairs_lowest_common_ancestor(G[,Yield the lowest common ancestor for sets of pairs in a
. . . ]) tree.
lowest_common_ancestor(G, node1, node2[, Compute the lowest common ancestor of the given pair
. . . ]) of nodes.

3.36.1 networkx.algorithms.lowest_common_ancestors.all_pairs_lowest_common_ancestor

all_pairs_lowest_common_ancestor(G, pairs=None)
Compute the lowest common ancestor for pairs of nodes.
Parameters
• G (NetworkX directed graph)
• pairs (iterable of pairs of nodes, optional (default: all pairs)) – The pairs of nodes of
interest. If None, will find the LCA of all pairs of nodes.
Returns
• An iterator over ((node1, node2), lca) where (node1, node2) are
• the pairs specified and lca is a lowest common ancestor of the pair.
• Note that for the default of all pairs in G, we consider
• unordered pairs, e.g. you will not get both (b, a) and (a, b).

3.36. Lowest Common Ancestor 373


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Only defined on non-null directed acyclic graphs.


Uses the 𝑂(𝑛3 ) ancestor-list algorithm from: M. A. Bender, M. Farach-Colton, G. Pemmasani, S. Skiena, P.
Sumazin. “Lowest common ancestors in trees and directed acyclic graphs.” Journal of Algorithms, 57(2): 75-94,
2005.
See also:
tree_all_pairs_lowest_common_ancestor(), lowest_common_ancestor()

3.36.2 networkx.algorithms.lowest_common_ancestors.tree_all_pairs_lowest_common_ancest

tree_all_pairs_lowest_common_ancestor(G, root=None, pairs=None)


Yield the lowest common ancestor for sets of pairs in a tree.
Parameters
• G (NetworkX directed graph (must be a tree))
• root (node, optional (default: None)) – The root of the subtree to operate on. If None,
assume the entire graph has exactly one source and use that.
• pairs (iterable or iterator of pairs of nodes, optional (default: None)) – The pairs of interest.
If None, Defaults to all pairs of nodes under root that have a lowest common ancestor.
Returns lcas – in pairs and lca is their lowest common ancestor.
Return type generator of tuples ((u, v), lca) where u and v are nodes

Notes

Only defined on non-null trees represented with directed edges from parents to children. Uses Tarjan’s off-line
lowest-common-ancestors algorithm. Runs in time 𝑂(4 × (𝑉 + 𝐸 + 𝑃 )) time, where 4 is the largest value of
the inverse Ackermann function likely to ever come up in actual use, and 𝑃 is the number of pairs requested (or
𝑉 2 if all are needed).
Tarjan, R. E. (1979), “Applications of path compression on balanced trees”, Journal of the ACM 26 (4): 690-715,
doi:10.1145/322154.322161.
See also:
all_pairs_lowest_common_ancestor(), lowest_common_ancestor()

3.36.3 networkx.algorithms.lowest_common_ancestors.lowest_common_ancestor

lowest_common_ancestor(G, node1, node2, default=None)


Compute the lowest common ancestor of the given pair of nodes.
Parameters
• G (NetworkX directed graph)
• node1, node2 (nodes in the graph.)
• default (object) – Returned if no common ancestor between node1 and node2
Returns

374 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• The lowest common ancestor of node1 and node2,


• or default if they have no common ancestors.

Notes

Only defined on non-null directed acyclic graphs. Takes n log(n) time in the size of the graph. See
all_pairs_lowest_common_ancestor when you have more than one pair of nodes of interest.
See also:
tree_all_pairs_lowest_common_ancestor(), all_pairs_lowest_common_ancestor()

3.37 Matching

Functions for computing and verifying matchings in a graph.

is_matching(G, matching) Decides whether the given set or dictionary represents a


valid matching in G.
is_maximal_matching(G, matching) Decides whether the given set or dictionary represents a
valid maximal matching in G.
is_perfect_matching(G, matching) Decides whether the given set represents a valid perfect
matching in G.
maximal_matching(G) Find a maximal matching in the graph.
max_weight_matching(G[, maxcardinality, Compute a maximum-weighted matching of G.
weight])

3.37.1 networkx.algorithms.matching.is_matching

is_matching(G, matching)
Decides whether the given set or dictionary represents a valid matching in G.
A matching in a graph is a set of edges in which no two distinct edges share a common endpoint.
Parameters
• G (NetworkX graph)
• matching (dict or set) – A dictionary or set representing a matching. If a dictionary, it
must have matching[u] == v and matching[v] == u for each edge (u, v) in
the matching. If a set, it must have elements of the form (u, v), where (u, v) is an
edge in the matching.
Returns Whether the given set or dictionary represents a valid matching in the graph.
Return type bool

3.37.2 networkx.algorithms.matching.is_maximal_matching

is_maximal_matching(G, matching)
Decides whether the given set or dictionary represents a valid maximal matching in G.
A maximal matching in a graph is a matching in which adding any edge would cause the set to no longer be a
valid matching.

3.37. Matching 375


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (NetworkX graph)
• matching (dict or set) – A dictionary or set representing a matching. If a dictionary, it
must have matching[u] == v and matching[v] == u for each edge (u, v) in
the matching. If a set, it must have elements of the form (u, v), where (u, v) is an
edge in the matching.
Returns Whether the given set or dictionary represents a valid maximal matching in the graph.
Return type bool

3.37.3 networkx.algorithms.matching.is_perfect_matching

is_perfect_matching(G, matching)
Decides whether the given set represents a valid perfect matching in G.
A perfect matching in a graph is a matching in which exactly one edge is incident upon each vertex.
Parameters
• G (NetworkX graph)
• matching (dict or set) – A dictionary or set representing a matching. If a dictionary, it
must have matching[u] == v and matching[v] == u for each edge (u, v) in
the matching. If a set, it must have elements of the form (u, v), where (u, v) is an
edge in the matching.
Returns Whether the given set or dictionary represents a valid perfect matching in the graph.
Return type bool

3.37.4 networkx.algorithms.matching.maximal_matching

maximal_matching(G)
Find a maximal matching in the graph.
A matching is a subset of edges in which no node occurs more than once. A maximal matching cannot add more
edges and still be a matching.
Parameters G (NetworkX graph) – Undirected graph
Returns matching – A maximal matching of the graph.
Return type set

Notes

The algorithm greedily selects a maximal matching M of the graph G (i.e. no superset of M exists). It runs in
𝑂(|𝐸|) time.

3.37.5 networkx.algorithms.matching.max_weight_matching

max_weight_matching(G, maxcardinality=False, weight=’weight’)


Compute a maximum-weighted matching of G.

376 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

A matching is a subset of edges in which no node occurs more than once. The weight of a matching is the sum
of the weights of its edges. A maximal matching cannot add more edges and still be a matching. The cardinality
of a matching is the number of matched edges.
Parameters
• G (NetworkX graph) – Undirected graph
• maxcardinality (bool, optional (default=False)) – If maxcardinality is True, compute the
maximum-cardinality matching with maximum weight among all maximum-cardinality
matchings.
• weight (string, optional (default=’weight’)) – Edge data key corresponding to the edge
weight. If key not found, uses 1 as weight.
Returns matching – A maximal matching of the graph.
Return type set

Notes

If G has edges with weight attributes the edge data are used as weight values else the weights are assumed to be
1.
This function takes time O(number_of_nodes ** 3).
If all edge weights are integers, the algorithm uses only integer computations. If floating point weights are used,
the algorithm could return a slightly suboptimal matching due to numeric precision errors.
This method is based on the “blossom” method for finding augmenting paths and the “primal-dual” method for
finding a matching of maximum weight, both methods invented by Jack Edmonds1 .
Bipartite graphs can also be matched using the functions present in networkx.algorithms.bipartite.
matching.

References

3.38 Minors

Provides functions for computing minors of a graph.

contracted_edge(G, edge[, self_loops]) Returns the graph that results from contracting the spec-
ified edge.
contracted_nodes(G, u, v[, self_loops]) Returns the graph that results from contracting u and v.
identified_nodes(G, u, v[, self_loops]) Returns the graph that results from contracting u and v.
quotient_graph(G, partition[, . . . ]) Returns the quotient graph of G under the specified
equivalence relation on nodes.

3.38.1 networkx.algorithms.minors.contracted_edge

contracted_edge(G, edge, self_loops=True)


Returns the graph that results from contracting the specified edge.
Edge contraction identifies the two endpoints of the edge as a single node incident to any edge that was incident
1 “Efficient Algorithms for Finding Maximum Matching in Graphs”, Zvi Galil, ACM Computing Surveys, 1986.

3.38. Minors 377


NetworkX Reference, Release 2.3rc1.dev20190222214247

to the original two nodes. A graph that results from edge contraction is called a minor of the original graph.
Parameters
• G (NetworkX graph) – The graph whose edge will be contracted.
• edge (tuple) – Must be a pair of nodes in G.
• self_loops (Boolean) – If this is True, any edges (including edge) joining the endpoints of
edge in G become self-loops on the new node in the returned graph.
Returns A new graph object of the same type as G (leaving G unmodified) with endpoints of edge
identified in a single node. The right node of edge will be merged into the left one, so only the
left one will appear in the returned graph.
Return type Networkx graph
Raises ValueError – If edge is not an edge in G.

Examples

Attempting to contract two nonadjacent nodes yields an error:

>>> import networkx as nx


>>> G = nx.cycle_graph(4)
>>> nx.contracted_edge(G, (1, 3))
Traceback (most recent call last):
...
ValueError: Edge (1, 3) does not exist in graph G; cannot contract it

Contracting two adjacent nodes in the cycle graph on n nodes yields the cycle graph on n - 1 nodes:

>>> import networkx as nx


>>> C5 = nx.cycle_graph(5)
>>> C4 = nx.cycle_graph(4)
>>> M = nx.contracted_edge(C5, (0, 1), self_loops=False)
>>> nx.is_isomorphic(M, C4)
True

See also:
contracted_nodes(), quotient_graph()

3.38.2 networkx.algorithms.minors.contracted_nodes

contracted_nodes(G, u, v, self_loops=True)
Returns the graph that results from contracting u and v.
Node contraction identifies the two nodes as a single node incident to any edge that was incident to the original
two nodes.
Parameters
• G (NetworkX graph) – The graph whose nodes will be contracted.
• u, v (nodes) – Must be nodes in G.
• self_loops (Boolean) – If this is True, any edges joining u and v in G become self-loops on
the new node in the returned graph.

378 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns A new graph object of the same type as G (leaving G unmodified) with u and v identified
in a single node. The right node v will be merged into the node u, so only u will appear in the
returned graph.
Return type Networkx graph

Notes

For multigraphs, the edge keys for the realigned edges may not be the same as the edge keys for the old edges.
This is natural because edge keys are unique only within each pair of nodes.

Examples

Contracting two nonadjacent nodes of the cycle graph on four nodes C_4 yields the path graph (ignoring parallel
edges):

>>> G = nx.cycle_graph(4)
>>> M = nx.contracted_nodes(G, 1, 3)
>>> P3 = nx.path_graph(3)
>>> nx.is_isomorphic(M, P3)
True

>>> G = nx.MultiGraph(P3)
>>> M = nx.contracted_nodes(G, 0, 2)
>>> M.edges
MultiEdgeView([(0, 1, 0), (0, 1, 1)])

>>> G = nx.Graph([(1,2), (2,2)])


>>> H = nx.contracted_nodes(G, 1, 2, self_loops=False)
>>> list(H.nodes())
[1]
>>> list(H.edges())
[(1, 1)]

See also:
contracted_edge(), quotient_graph()

Notes

This function is also available as identified_nodes.

3.38.3 networkx.algorithms.minors.identified_nodes

identified_nodes(G, u, v, self_loops=True)
Returns the graph that results from contracting u and v.
Node contraction identifies the two nodes as a single node incident to any edge that was incident to the original
two nodes.
Parameters
• G (NetworkX graph) – The graph whose nodes will be contracted.
• u, v (nodes) – Must be nodes in G.

3.38. Minors 379


NetworkX Reference, Release 2.3rc1.dev20190222214247

• self_loops (Boolean) – If this is True, any edges joining u and v in G become self-loops on
the new node in the returned graph.
Returns A new graph object of the same type as G (leaving G unmodified) with u and v identified
in a single node. The right node v will be merged into the node u, so only u will appear in the
returned graph.
Return type Networkx graph

Notes

For multigraphs, the edge keys for the realigned edges may not be the same as the edge keys for the old edges.
This is natural because edge keys are unique only within each pair of nodes.

Examples

Contracting two nonadjacent nodes of the cycle graph on four nodes C_4 yields the path graph (ignoring parallel
edges):

>>> G = nx.cycle_graph(4)
>>> M = nx.contracted_nodes(G, 1, 3)
>>> P3 = nx.path_graph(3)
>>> nx.is_isomorphic(M, P3)
True

>>> G = nx.MultiGraph(P3)
>>> M = nx.contracted_nodes(G, 0, 2)
>>> M.edges
MultiEdgeView([(0, 1, 0), (0, 1, 1)])

>>> G = nx.Graph([(1,2), (2,2)])


>>> H = nx.contracted_nodes(G, 1, 2, self_loops=False)
>>> list(H.nodes())
[1]
>>> list(H.edges())
[(1, 1)]

See also:
contracted_edge(), quotient_graph()

Notes

This function is also available as identified_nodes.

3.38.4 networkx.algorithms.minors.quotient_graph

quotient_graph(G, partition, edge_relation=None, node_data=None, edge_data=None, relabel=False,


create_using=None)
Returns the quotient graph of G under the specified equivalence relation on nodes.
Parameters
• G (NetworkX graph) – The graph for which to return the quotient graph with the specified
node relation.

380 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• partition (function or list of sets) – If a function, this function must represent an equivalence
relation on the nodes of G. It must take two arguments u and v and return True exactly when
u and v are in the same equivalence class. The equivalence classes form the nodes in the
returned graph.
If a list of sets, the list must form a valid partition of the nodes of the graph. That is, each
node must be in exactly one block of the partition.
• edge_relation (Boolean function with two arguments) – This function must represent an
edge relation on the blocks of G in the partition induced by node_relation. It must take
two arguments, B and C, each one a set of nodes, and return True exactly when there should
be an edge joining block B to block C in the returned graph.
If edge_relation is not specified, it is assumed to be the following relation. Block B is
related to block C if and only if some node in B is adjacent to some node in C, according to
the edge set of G.
• edge_data (function) – This function takes two arguments, B and C, each one a set of nodes,
and must return a dictionary representing the edge data attributes to set on the edge joining
B and C, should there be an edge joining B and C in the quotient graph (if no such edge
occurs in the quotient graph as determined by edge_relation, then the output of this
function is ignored).
If the quotient graph would be a multigraph, this function is not applied, since the edge data
from each edge in the graph G appears in the edges of the quotient graph.
• node_data (function) – This function takes one argument, B, a set of nodes in G, and must
return a dictionary representing the node data attributes to set on the node representing B in
the quotient graph. If None, the following node attributes will be set:
– ‘graph’, the subgraph of the graph G that this block represents,
– ‘nnodes’, the number of nodes in this block,
– ‘nedges’, the number of edges within this block,
– ‘density’, the density of the subgraph of G that this block represents.
• relabel (bool) – If True, relabel the nodes of the quotient graph to be nonnegative integers.
Otherwise, the nodes are identified with frozenset instances representing the blocks
given in partition.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
Returns The quotient graph of G under the equivalence relation specified by partition. If the
partition were given as a list of set instances and relabel is False, each node will be a
frozenset corresponding to the same set.
Return type NetworkX graph
Raises NetworkXException – If the given partition is not a valid partition of the nodes of G.

Examples

The quotient graph of the complete bipartite graph under the “same neighbors” equivalence relation is K_2.
Under this relation, two nodes are equivalent if they are not adjacent but have the same neighbor set:

3.38. Minors 381


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> import networkx as nx


>>> G = nx.complete_bipartite_graph(2, 3)
>>> same_neighbors = lambda u, v: (u not in G[v] and v not in G[u]
... and G[u] == G[v])
>>> Q = nx.quotient_graph(G, same_neighbors)
>>> K2 = nx.complete_graph(2)
>>> nx.is_isomorphic(Q, K2)
True

The quotient graph of a directed graph under the “same strongly connected component” equivalence relation
is the condensation of the graph (see condensation()). This example comes from the Wikipedia article
‘Strongly connected component‘_:

>>> import networkx as nx


>>> G = nx.DiGraph()
>>> edges = ['ab', 'be', 'bf', 'bc', 'cg', 'cd', 'dc', 'dh', 'ea',
... 'ef', 'fg', 'gf', 'hd', 'hf']
>>> G.add_edges_from(tuple(x) for x in edges)
>>> components = list(nx.strongly_connected_components(G))
>>> sorted(sorted(component) for component in components)
[['a', 'b', 'e'], ['c', 'd', 'h'], ['f', 'g']]
>>>
>>> C = nx.condensation(G, components)
>>> component_of = C.graph['mapping']
>>> same_component = lambda u, v: component_of[u] == component_of[v]
>>> Q = nx.quotient_graph(G, same_component)
>>> nx.is_isomorphic(C, Q)
True

Node identification can be represented as the quotient of a graph under the equivalence relation that places the
two nodes in one block and each other node in its own singleton block:

>>> import networkx as nx


>>> K24 = nx.complete_bipartite_graph(2, 4)
>>> K34 = nx.complete_bipartite_graph(3, 4)
>>> C = nx.contracted_nodes(K34, 1, 2)
>>> nodes = {1, 2}
>>> is_contracted = lambda u, v: u in nodes and v in nodes
>>> Q = nx.quotient_graph(K34, is_contracted)
>>> nx.is_isomorphic(Q, C)
True
>>> nx.is_isomorphic(Q, K24)
True

The blockmodeling technique described in1 can be implemented as a quotient graph:

>>> G = nx.path_graph(6)
>>> partition = [{0, 1}, {2, 3}, {4, 5}]
>>> M = nx.quotient_graph(G, partition, relabel=True)
>>> list(M.edges())
[(0, 1), (1, 2)]

1 Patrick Doreian, Vladimir Batagelj, and Anuska Ferligoj. Generalized Blockmodeling. Cambridge University Press, 2004.

382 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.39 Maximal independent set

Algorithm to find a maximal (not maximum) independent set.

maximal_independent_set(G[, nodes, seed]) Returns a random maximal independent set guaranteed


to contain a given set of nodes.

3.39.1 networkx.algorithms.mis.maximal_independent_set

maximal_independent_set(G, nodes=None, seed=None)


Returns a random maximal independent set guaranteed to contain a given set of nodes.
An independent set is a set of nodes such that the subgraph of G induced by these nodes contains no edges. A
maximal independent set is an independent set such that it is not possible to add a new node and still get an
independent set.
Parameters
• G (NetworkX graph)
• nodes (list or iterable) – Nodes that must be part of the independent set. This set of nodes
must be independent.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns indep_nodes – List of nodes that are part of a maximal independent set.
Return type list
Raises
• NetworkXUnfeasible – If the nodes in the provided list are not part of the graph or do
not form an independent set, an exception is raised.
• NetworkXNotImplemented – If G is directed.

Examples

>>> G = nx.path_graph(5)
>>> nx.maximal_independent_set(G) # doctest: +SKIP
[4, 0, 2]
>>> nx.maximal_independent_set(G, [1]) # doctest: +SKIP
[1, 3]

Notes

This algorithm does not solve the maximum independent set problem.

3.39. Maximal independent set 383


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.40 Node Classification

This module provides the functions for node classification problem.


The functions in this module are not imported into the top level networkx namespace. You can access these functions
by importing the networkx.algorithms.node_classification modules, then accessing the functions as
attributes of node_classification. For example:

>>> import networkx as nx


>>> from networkx.algorithms import node_classification
>>> G = nx.path_graph(4)
>>> G.edges()
EdgeView([(0, 1), (1, 2), (2, 3)])
>>> G.node[0]['label'] = 'A'
>>> G.node[3]['label'] = 'B'
>>> node_classification.harmonic_function(G) # doctest: +SKIP
['A', 'A', 'B', 'B']

3.40.1 Harmonic Function

Function for computing Harmonic function algorithm by Zhu et al.

References

Zhu, X., Ghahramani, Z., & Lafferty, J. (2003, August). Semi-supervised learning using gaussian fields and harmonic
functions. In ICML (Vol. 3, pp. 912-919).

harmonic_function(G[, max_iter, label_name]) Node classification by Harmonic function

networkx.algorithms.node_classification.hmn.harmonic_function

harmonic_function(G, max_iter=30, label_name=’label’)


Node classification by Harmonic function
Parameters
• G (NetworkX Graph)
• max_iter (int) – maximum number of iterations allowed
• label_name (string) – name of target labels to predict
Raises NetworkXError if no nodes on G has label_name.
Returns predicted – Array of predicted labels
Return type array, shape = [n_samples]

Examples

>>> from networkx.algorithms import node_classification


>>> G = nx.path_graph(4)
>>> G.node[0]['label'] = 'A'
(continues on next page)

384 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> G.node[3]['label'] = 'B'
>>> G.nodes(data=True)
NodeDataView({0: {'label': 'A'}, 1: {}, 2: {}, 3: {'label': 'B'}})
>>> G.edges()
EdgeView([(0, 1), (1, 2), (2, 3)])
>>> predicted = node_classification.harmonic_function(G)
>>> predicted
['A', 'A', 'B', 'B']

References

Zhu, X., Ghahramani, Z., & Lafferty, J. (2003, August). Semi-supervised learning using gaussian fields and
harmonic functions. In ICML (Vol. 3, pp. 912-919).

3.40.2 Local and Global Consistency

Function for computing Local and global consistency algorithm by Zhou et al.

References

Zhou, D., Bousquet, O., Lal, T. N., Weston, J., & Schölkopf, B. (2004). Learning with local and global consistency.
Advances in neural information processing systems, 16(16), 321-328.

local_and_global_consistency(G[, alpha, Node classification by Local and Global Consistency


. . . ])

networkx.algorithms.node_classification.lgc.local_and_global_consistency

local_and_global_consistency(G, alpha=0.99, max_iter=30, label_name=’label’)


Node classification by Local and Global Consistency
Parameters
• G (NetworkX Graph)
• alpha (float) – Clamping factor
• max_iter (int) – Maximum number of iterations allowed
• label_name (string) – Name of target labels to predict
Raises NetworkXError if no nodes on G has label_name.
Returns predicted – Array of predicted labels
Return type array, shape = [n_samples]

Examples

>>> from networkx.algorithms import node_classification


>>> G = nx.path_graph(4)
>>> G.node[0]['label'] = 'A'
(continues on next page)

3.40. Node Classification 385


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> G.node[3]['label'] = 'B'
>>> G.nodes(data=True)
NodeDataView({0: {'label': 'A'}, 1: {}, 2: {}, 3: {'label': 'B'}})
>>> G.edges()
EdgeView([(0, 1), (1, 2), (2, 3)])
>>> predicted = node_classification.local_and_global_consistency(G)
>>> predicted
['A', 'A', 'B', 'B']

References

Zhou, D., Bousquet, O., Lal, T. N., Weston, J., & Schölkopf, B. (2004). Learning with local and global consis-
tency. Advances in neural information processing systems, 16(16), 321-328.

3.41 Operators

Unary operations on graphs

complement(G) Returns the graph complement of G.


reverse(G[, copy]) Returns the reverse directed graph of G.

3.41.1 networkx.algorithms.operators.unary.complement

complement(G)
Returns the graph complement of G.
Parameters G (graph) – A NetworkX graph
Returns GC
Return type A new graph.

Notes

Note that complement() does not create self-loops and also does not produce parallel edges for MultiGraphs.
Graph, node, and edge data are not propagated to the new graph.

3.41.2 networkx.algorithms.operators.unary.reverse

reverse(G, copy=True)
Returns the reverse directed graph of G.
Parameters
• G (directed graph) – A NetworkX directed graph
• copy (bool) – If True, then a new graph is returned. If False, then the graph is reversed in
place.
Returns H – The reversed G.

386 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type directed graph


Operations on graphs including union, intersection, difference.

compose(G, H) Returns a new graph of G composed with H.


union(G, H[, rename, name]) Return the union of graphs G and H.
disjoint_union(G, H) Return the disjoint union of graphs G and H.
intersection(G, H) Returns a new graph that contains only the edges that
exist in both G and H.
difference(G, H) Returns a new graph that contains the edges that exist in
G but not in H.
symmetric_difference(G, H) Returns new graph with edges that exist in either G or
H but not both.

3.41.3 networkx.algorithms.operators.binary.compose

compose(G, H)
Returns a new graph of G composed with H.
Composition is the simple union of the node sets and edge sets. The node sets of G and H do not need to be
disjoint.
Parameters G, H (graph) – A NetworkX graph
Returns C
Return type A new graph with the same type as G

Notes

It is recommended that G and H be either both directed or both undirected. Attributes from H take precedent
over attributes from G.
For MultiGraphs, the edges are identified by incident nodes AND edge-key. This can cause surprises (i.e., edge
(1, 2) may or may not be the same in two graphs) if you use MultiGraph without keeping track of edge keys.

3.41.4 networkx.algorithms.operators.binary.union

union(G, H, rename=(None, None), name=None)


Return the union of graphs G and H.
Graphs G and H must be disjoint, otherwise an exception is raised.
Parameters
• G,H (graph) – A NetworkX graph
• rename (bool , default=(None, None)) – Node names of G and H can be changed by spec-
ifying the tuple rename=(‘G-‘,’H-‘) (for example). Node “u” in G is then renamed “G-u”
and “v” in H is renamed “H-v”.
• name (string) – Specify the name for the union graph
Returns U
Return type A union graph with the same type as G.

3.41. Operators 387


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

To force a disjoint union with node relabeling, use disjoint_union(G,H) or convert_node_labels_to integers().
Graph, edge, and node attributes are propagated from G and H to the union graph. If a graph attribute is present
in both G and H the value from H is used.
See also:
disjoint_union()

3.41.5 networkx.algorithms.operators.binary.disjoint_union

disjoint_union(G, H)
Return the disjoint union of graphs G and H.
This algorithm forces distinct integer node labels.
Parameters G,H (graph) – A NetworkX graph
Returns U
Return type A union graph with the same type as G.

Notes

A new graph is created, of the same class as G. It is recommended that G and H be either both directed or both
undirected.
The nodes of G are relabeled 0 to len(G)-1, and the nodes of H are relabeled len(G) to len(G)+len(H)-1.
Graph, edge, and node attributes are propagated from G and H to the union graph. If a graph attribute is present
in both G and H the value from H is used.

3.41.6 networkx.algorithms.operators.binary.intersection

intersection(G, H)
Returns a new graph that contains only the edges that exist in both G and H.
The node sets of H and G must be the same.
Parameters G,H (graph) – A NetworkX graph. G and H must have the same node sets.
Returns GH
Return type A new graph with the same type as G.

Notes

Attributes from the graph, nodes, and edges are not copied to the new graph. If you want a new graph of the
intersection of G and H with the attributes (including edge data) from G use remove_nodes_from() as follows

>>> G=nx.path_graph(3)
>>> H=nx.path_graph(5)
>>> R=G.copy()
>>> R.remove_nodes_from(n for n in G if n not in H)

388 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.41.7 networkx.algorithms.operators.binary.difference

difference(G, H)
Returns a new graph that contains the edges that exist in G but not in H.
The node sets of H and G must be the same.
Parameters G,H (graph) – A NetworkX graph. G and H must have the same node sets.
Returns D
Return type A new graph with the same type as G.

Notes

Attributes from the graph, nodes, and edges are not copied to the new graph. If you want a new graph of
the difference of G and H with with the attributes (including edge data) from G use remove_nodes_from() as
follows:

>>> G = nx.path_graph(3)
>>> H = nx.path_graph(5)
>>> R = G.copy()
>>> R.remove_nodes_from(n for n in G if n in H)

3.41.8 networkx.algorithms.operators.binary.symmetric_difference

symmetric_difference(G, H)
Returns new graph with edges that exist in either G or H but not both.
The node sets of H and G must be the same.
Parameters G,H (graph) – A NetworkX graph. G and H must have the same node sets.
Returns D
Return type A new graph with the same type as G.

Notes

Attributes from the graph, nodes, and edges are not copied to the new graph.
Operations on many graphs.

compose_all(graphs) Returns the composition of all graphs.


union_all(graphs[, rename]) Returns the union of all graphs.
disjoint_union_all(graphs) Returns the disjoint union of all graphs.
intersection_all(graphs) Returns a new graph that contains only the edges that
exist in all graphs.

3.41.9 networkx.algorithms.operators.all.compose_all

compose_all(graphs)
Returns the composition of all graphs.
Composition is the simple union of the node sets and edge sets. The node sets of the supplied graphs need not

3.41. Operators 389


NetworkX Reference, Release 2.3rc1.dev20190222214247

be disjoint.
Parameters graphs (list) – List of NetworkX graphs
Returns C
Return type A graph with the same type as the first graph in list
Raises ValueError – If graphs is an empty list.

Notes

It is recommended that the supplied graphs be either all directed or all undirected.
Graph, edge, and node attributes are propagated to the union graph. If a graph attribute is present in multiple
graphs, then the value from the last graph in the list with that attribute is used.

3.41.10 networkx.algorithms.operators.all.union_all

union_all(graphs, rename=(None, ))
Returns the union of all graphs.
The graphs must be disjoint, otherwise an exception is raised.
Parameters
• graphs (list of graphs) – List of NetworkX graphs
• rename (bool , default=(None, None)) – Node names of G and H can be changed by spec-
ifying the tuple rename=(‘G-‘,’H-‘) (for example). Node “u” in G is then renamed “G-u”
and “v” in H is renamed “H-v”.
Returns U
Return type a graph with the same type as the first graph in list
Raises ValueError – If graphs is an empty list.

Notes

To force a disjoint union with node relabeling, use disjoint_union_all(G,H) or convert_node_labels_to integers().
Graph, edge, and node attributes are propagated to the union graph. If a graph attribute is present in multiple
graphs, then the value from the last graph in the list with that attribute is used.
See also:
union(), disjoint_union_all()

3.41.11 networkx.algorithms.operators.all.disjoint_union_all

disjoint_union_all(graphs)
Returns the disjoint union of all graphs.
This operation forces distinct integer node labels starting with 0 for the first graph in the list and numbering
consecutively.
Parameters graphs (list) – List of NetworkX graphs
Returns U

390 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type A graph with the same type as the first graph in list
Raises ValueError – If graphs is an empty list.

Notes

It is recommended that the graphs be either all directed or all undirected.


Graph, edge, and node attributes are propagated to the union graph. If a graph attribute is present in multiple
graphs, then the value from the last graph in the list with that attribute is used.

3.41.12 networkx.algorithms.operators.all.intersection_all

intersection_all(graphs)
Returns a new graph that contains only the edges that exist in all graphs.
All supplied graphs must have the same node set.
Parameters graphs (list) – List of NetworkX graphs
Returns R
Return type A new graph with the same type as the first graph in list
Raises ValueError – If graphs is an empty list.

Notes

Attributes from the graph, nodes, and edges are not copied to the new graph.
Graph products.

cartesian_product(G, H) Returns the Cartesian product of G and H.


lexicographic_product(G, H) Returns the lexicographic product of G and H.
rooted_product(G, H, root) Return the rooted product of graphs G and H rooted at
root in H.
strong_product(G, H) Returns the strong product of G and H.
tensor_product(G, H) Returns the tensor product of G and H.
power(G, k) Returns the specified power of a graph.

3.41.13 networkx.algorithms.operators.product.cartesian_product

cartesian_product(G, H)
Returns the Cartesian product of G and H.
The Cartesian product 𝑃 of the graphs 𝐺 and 𝐻 has a node set that is the Cartesian product of the node sets,
𝑉 (𝑃 ) = 𝑉 (𝐺) × 𝑉 (𝐻). 𝑃 has an edge ((𝑢, 𝑣), (𝑥, 𝑦)) if and only if either 𝑢 is equal to 𝑥 and both 𝑣 and 𝑦 are
adjacent in 𝐻 or if 𝑣 is equal to 𝑦 and both 𝑢 and 𝑥 are adjacent in 𝐺.
Parameters G, H (graphs) – Networkx graphs.
Returns P – The Cartesian product of G and H. P will be a multi-graph if either G or H is a multi-
graph. Will be a directed if G and H are directed, and undirected if G and H are undirected.
Return type NetworkX graph
Raises NetworkXError – If G and H are not both directed or both undirected.

3.41. Operators 391


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Node attributes in P are two-tuple of the G and H node attributes. Missing attributes are assigned None.

Examples

>>> G = nx.Graph()
>>> H = nx.Graph()
>>> G.add_node(0, a1=True)
>>> H.add_node('a', a2='Spam')
>>> P = nx.cartesian_product(G, H)
>>> list(P)
[(0, 'a')]

Edge attributes and edge keys (for multigraphs) are also copied to the new product graph

3.41.14 networkx.algorithms.operators.product.lexicographic_product

lexicographic_product(G, H)
Returns the lexicographic product of G and H.
The lexicographical product 𝑃 of the graphs 𝐺 and 𝐻 has a node set that is the Cartesian product of the node
sets, 𝑉 (𝑃 ) = 𝑉 (𝐺) × 𝑉 (𝐻). 𝑃 has an edge ((𝑢, 𝑣), (𝑥, 𝑦)) if and only if (𝑢, 𝑣) is an edge in 𝐺 or 𝑢 == 𝑣 and
(𝑥, 𝑦) is an edge in 𝐻.
Parameters G, H (graphs) – Networkx graphs.
Returns P – The Cartesian product of G and H. P will be a multi-graph if either G or H is a multi-
graph. Will be a directed if G and H are directed, and undirected if G and H are undirected.
Return type NetworkX graph
Raises NetworkXError – If G and H are not both directed or both undirected.

Notes

Node attributes in P are two-tuple of the G and H node attributes. Missing attributes are assigned None.

Examples

>>> G = nx.Graph()
>>> H = nx.Graph()
>>> G.add_node(0, a1=True)
>>> H.add_node('a', a2='Spam')
>>> P = nx.lexicographic_product(G, H)
>>> list(P)
[(0, 'a')]

Edge attributes and edge keys (for multigraphs) are also copied to the new product graph

392 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.41.15 networkx.algorithms.operators.product.rooted_product

rooted_product(G, H, root)
Return the rooted product of graphs G and H rooted at root in H.
A new graph is constructed representing the rooted product of the inputted graphs, G and H, with a root in H. A
rooted product duplicates H for each nodes in G with the root of H corresponding to the node in G. Nodes are
renamed as the direct product of G and H. The result is a subgraph of the cartesian product.
Parameters
• G,H (graph) – A NetworkX graph
• root (node) – A node in H
Returns R
Return type The rooted product of G and H with a specified root in H

Notes

The nodes of R are the Cartesian Product of the nodes of G and H. The nodes of G and H are not relabeled.

3.41.16 networkx.algorithms.operators.product.strong_product

strong_product(G, H)
Returns the strong product of G and H.
The strong product 𝑃 of the graphs 𝐺 and 𝐻 has a node set that is the Cartesian product of the node sets,
𝑉 (𝑃 ) = 𝑉 (𝐺) × 𝑉 (𝐻). 𝑃 has an edge ((𝑢, 𝑣), (𝑥, 𝑦)) if and only if 𝑢 == 𝑣 and (𝑥, 𝑦) is an edge in 𝐻, or
𝑥 == 𝑦 and (𝑢, 𝑣) is an edge in 𝐺, or (𝑢, 𝑣) is an edge in 𝐺 and (𝑥, 𝑦) is an edge in 𝐻.
Parameters G, H (graphs) – Networkx graphs.
Returns P – The Cartesian product of G and H. P will be a multi-graph if either G or H is a multi-
graph. Will be a directed if G and H are directed, and undirected if G and H are undirected.
Return type NetworkX graph
Raises NetworkXError – If G and H are not both directed or both undirected.

Notes

Node attributes in P are two-tuple of the G and H node attributes. Missing attributes are assigned None.

Examples

>>> G = nx.Graph()
>>> H = nx.Graph()
>>> G.add_node(0, a1=True)
>>> H.add_node('a', a2='Spam')
>>> P = nx.strong_product(G, H)
>>> list(P)
[(0, 'a')]

Edge attributes and edge keys (for multigraphs) are also copied to the new product graph

3.41. Operators 393


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.41.17 networkx.algorithms.operators.product.tensor_product

tensor_product(G, H)
Returns the tensor product of G and H.
The tensor product 𝑃 of the graphs 𝐺 and 𝐻 has a node set that is the tensor product of the node sets, 𝑉 (𝑃 ) =
𝑉 (𝐺) × 𝑉 (𝐻). 𝑃 has an edge ((𝑢, 𝑣), (𝑥, 𝑦)) if and only if (𝑢, 𝑥) is an edge in 𝐺 and (𝑣, 𝑦) is an edge in 𝐻.
Tensor product is sometimes also referred to as the categorical product, direct product, cardinal product or
conjunction.
Parameters G, H (graphs) – Networkx graphs.
Returns P – The tensor product of G and H. P will be a multi-graph if either G or H is a multi-graph,
will be a directed if G and H are directed, and undirected if G and H are undirected.
Return type NetworkX graph
Raises NetworkXError – If G and H are not both directed or both undirected.

Notes

Node attributes in P are two-tuple of the G and H node attributes. Missing attributes are assigned None.

Examples

>>> G = nx.Graph()
>>> H = nx.Graph()
>>> G.add_node(0, a1=True)
>>> H.add_node('a', a2='Spam')
>>> P = nx.tensor_product(G, H)
>>> list(P)
[(0, 'a')]

Edge attributes and edge keys (for multigraphs) are also copied to the new product graph

3.41.18 networkx.algorithms.operators.product.power

power(G, k)
Returns the specified power of a graph.
The 𝑘, denoted 𝐺𝑘 , is a graph on the same set of nodes in which two distinct nodes 𝑢 and 𝑣 are adjacent in 𝐺𝑘
if and only if the shortest path distance between 𝑢 and 𝑣 in 𝐺 is at most 𝑘.
Parameters
• G (graph) – A NetworkX simple graph object.
• k (positive integer) – The power to which to raise the graph G.
Returns G to the power k.
Return type NetworkX simple graph
Raises
• ValueError – If the exponent k is not positive.
• NetworkXNotImplemented – If G is not a simple graph.

394 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

The number of edges will never decrease when taking successive powers:

>>> G = nx.path_graph(4)
>>> list(nx.power(G, 2).edges)
[(0, 1), (0, 2), (1, 2), (1, 3), (2, 3)]
>>> list(nx.power(G, 3).edges)
[(0, 1), (0, 2), (0, 3), (1, 2), (1, 3), (2, 3)]

The k`th power of a cycle graph on *n* nodes is the complete graph on *n*
nodes, if `k is at least n // 2:

>>> G = nx.cycle_graph(5)
>>> H = nx.complete_graph(5)
>>> nx.is_isomorphic(nx.power(G, 2), H)
True
>>> G = nx.cycle_graph(8)
>>> H = nx.complete_graph(8)
>>> nx.is_isomorphic(nx.power(G, 4), H)
True

References

Notes

This definition of “power graph” comes from Exercise 3.1.6 of Graph Theory by Bondy and Murty1 .

3.42 Planarity

check_planarity(G[, counterexample]) Check if a graph is planar and return a counterexample


or an embedding.

3.42.1 networkx.algorithms.planarity.check_planarity

check_planarity(G, counterexample=False)
Check if a graph is planar and return a counterexample or an embedding.
A graph is planar iff it can be drawn in a plane without any edge intersections.
Parameters
• G (NetworkX graph)
• counterexample (bool) – A Kuratowski subgraph (to proof non planarity) is only returned
if set to true.
Returns (is_planar, certificate) – is_planar is true if the graph is planar. If the graph is planar
certificate is a PlanarEmbedding otherwise it is a Kuratowski subgraph.
Return type (bool, NetworkX graph) tuple
1

J. A. Bondy, U. S. R. Murty, Graph Theory. Springer, 2008.

3.42. Planarity 395


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

A (combinatorial) embedding consists of cyclic orderings of the incident edges at each vertex. Given such
an embedding there are multiple approaches discussed in literature to drawing the graph (subject to various
constraints, e.g. integer coordinates), see e.g. [2].
The planarity check algorithm and extraction of the combinatorial embedding is based on the Left-Right Pla-
narity Test [1].
A counterexample is only generated if the corresponding parameter is set, because the complexity of the coun-
terexample generation is higher.

References

class PlanarEmbedding(incoming_graph_data=None, **attr)


Represents a planar graph with its planar embedding.
The planar embedding is given by a combinatorial embedding.
Neighbor ordering:
In comparison to a usual graph structure, the embedding also stores the order of all neighbors for every vertex.
The order of the neighbors can be given in clockwise (cw) direction or counterclockwise (ccw) direction. This
order is stored as edge attributes in the underlying directed graph. For the edge (u, v) the edge attribute ‘cw’ is
set to the neighbor of u that follows immediately after v in clockwise direction.
In order for a PlanarEmbedding to be valid it must fulfill multiple conditions. It is possible to check if these
conditions are fulfilled with the method check_structure(). The conditions are:
• Edges must go in both directions (because the edge attributes differ)
• Every edge must have a ‘cw’ and ‘ccw’ attribute which corresponds to a correct planar embedding.
• A node with non zero degree must have a node attribute ‘first_nbr’.
As long as a PlanarEmbedding is invalid only the following methods should be called:
• add_half_edge_ccw()
• add_half_edge_cw()
• connect_components()
• add_half_edge_first()
Even though the graph is a subclass of nx.DiGraph, it can still be used for algorithms that require undirected
graphs, because the method is_directed() is overridden. This is possible, because a valid PlanarGraph
must have edges in both directions.
Half edges:
In methods like add_half_edge_ccw the term “half-edge” is used, which is a term that is used in doubly
connected edge lists. It is used to emphasize that the edge is only in one direction and there exists another
half-edge in the opposite direction. While conventional edges always have two faces (including outer face) next
to them, it is possible to assign each half-edge exactly one face. For a half-edge (u, v) that is orientated such that
u is below v then the face that belongs to (u, v) is to the right of this half-edge.

Examples

Create an embedding of a star graph (compare nx.star_graph(3)):

396 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.PlanarEmbedding()
>>> G.add_half_edge_cw(0, 1, None)
>>> G.add_half_edge_cw(0, 2, 1)
>>> G.add_half_edge_cw(0, 3, 2)
>>> G.add_half_edge_cw(1, 0, None)
>>> G.add_half_edge_cw(2, 0, None)
>>> G.add_half_edge_cw(3, 0, None)

Alternatively the same embedding can also be defined in counterclockwise orientation. The following results in
exactly the same PlanarEmbedding:

>>> G = nx.PlanarEmbedding()
>>> G.add_half_edge_ccw(0, 1, None)
>>> G.add_half_edge_ccw(0, 3, 1)
>>> G.add_half_edge_ccw(0, 2, 3)
>>> G.add_half_edge_ccw(1, 0, None)
>>> G.add_half_edge_ccw(2, 0, None)
>>> G.add_half_edge_ccw(3, 0, None)

After creating a graph, it is possible to validate that the PlanarEmbedding object is correct:

>>> G.check_structure()

add_half_edge_ccw(start_node, end_node, reference_neighbor)


Adds a half-edge from start_node to end_node.
The half-edge is added counter clockwise next to the existing half-edge (start_node, reference_neighbor).
Parameters
• start_node (node) – Start node of inserted edge.
• end_node (node) – End node of inserted edge.
• reference_neighbor (node) – End node of reference edge.
Raises nx.NetworkXException – If the reference_neighbor does not exist.
See also:
add_half_edge_cw(), connect_components(), add_half_edge_first()
add_half_edge_cw(start_node, end_node, reference_neighbor)
Adds a half-edge from start_node to end_node.
The half-edge is added clockwise next to the existing half-edge (start_node, reference_neighbor).
Parameters
• start_node (node) – Start node of inserted edge.
• end_node (node) – End node of inserted edge.
• reference_neighbor (node) – End node of reference edge.
Raises nx.NetworkXException – If the reference_neighbor does not exist.
See also:
add_half_edge_ccw(), connect_components(), add_half_edge_first()
add_half_edge_first(start_node, end_node)
The added half-edge is inserted at the first position in the order.

3.42. Planarity 397


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• start_node (node)
• end_node (node)
See also:
add_half_edge_ccw(), add_half_edge_cw(), connect_components()
check_structure()
Runs without exceptions if this object is valid.
Checks that the following properties are fulfilled:
• Edges go in both directions (because the edge attributes differ).
• Every edge has a ‘cw’ and ‘ccw’ attribute which corresponds to a correct planar embedding.
• A node with a degree larger than 0 has a node attribute ‘first_nbr’.
Running this method verifies that the underlying Graph must be planar.
Raises nx.NetworkXException – This exception is raised with a short explanation if the
PlanarEmbedding is invalid.
connect_components(v, w)
Adds half-edges for (v, w) and (w, v) at some position.
This method should only be called if v and w are in different components, or it might break the embed-
ding. This especially means that if connect_components(v, w) is called it is not allowed to call
connect_components(w, v) afterwards. The neighbor orientations in both directions are all set
correctly after the first call.
Parameters
• v (node)
• w (node)
See also:
add_half_edge_ccw(), add_half_edge_cw(), add_half_edge_first()
get_data()
Converts the adjacency structure into a better readable structure.
Returns embedding – A dict mapping all nodes to a list of neighbors sorted in clockwise order.
Return type dict
is_directed()
A valid PlanarEmbedding is undirected.
All reverse edges are contained, i.e. for every existing half-edge (v, w) the half-edge in the opposite
direction (w, v) is also contained.
neighbors_cw_order(v)
Generator for the neighbors of v in clockwise order.
Parameters v (node)
Yields node
next_face_half_edge(v, w)
Returns the following half-edge left of a face.
Parameters

398 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• v (node)
• w (node)
Returns half-edge
Return type tuple
traverse_face(v, w, mark_half_edges=None)
Returns nodes on the face that belong to the half-edge (v, w).
The face that is traversed lies to the right of the half-edge (in an orientation where v is below w).
Optionally it is possible to pass a set to which all encountered half edges are added. Before calling this
method, this set must not include any half-edges that belong to the face.
Parameters
• v (node) – Start node of half-edge.
• w (node) – End node of half-edge.
• mark_half_edges (set, optional) – Set to which all encountered half-edges are added.
Returns face – A list of nodes that lie on this face.
Return type list

3.43 Reciprocity

Algorithms to calculate reciprocity in a directed graph.

reciprocity(G[, nodes]) Compute the reciprocity in a directed graph.


overall_reciprocity(G) Compute the reciprocity for the whole graph.

3.43.1 networkx.algorithms.reciprocity.reciprocity

reciprocity(G, nodes=None)
Compute the reciprocity in a directed graph.
The reciprocity of a directed graph is defined as the ratio of the number of edges pointing in both directions to
the total number of edges in the graph. Formally, 𝑟 = |(𝑢, 𝑣) ∈ 𝐺|(𝑣, 𝑢) ∈ 𝐺|/|(𝑢, 𝑣) ∈ 𝐺|.
The reciprocity of a single node u is defined similarly, it is the ratio of the number of edges in both directions to
the total number of edges attached to node u.
Parameters
• G (graph) – A networkx directed graph
• nodes (container of nodes, optional (default=whole graph)) – Compute reciprocity for
nodes in this container.
Returns out – Reciprocity keyed by node label.
Return type dictionary

3.43. Reciprocity 399


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The reciprocity is not defined for isolated nodes. In such cases this function will return None.

3.43.2 networkx.algorithms.reciprocity.overall_reciprocity

overall_reciprocity(G)
Compute the reciprocity for the whole graph.
See the doc of reciprocity for the definition.
Parameters G (graph) – A networkx graph

3.44 Rich Club

Functions for computing rich-club coefficients.

rich_club_coefficient(G[, normalized, Q, Returns the rich-club coefficient of the graph G.


seed])

3.44.1 networkx.algorithms.richclub.rich_club_coefficient

rich_club_coefficient(G, normalized=True, Q=100, seed=None)


Returns the rich-club coefficient of the graph G.
For each degree k, the rich-club coefficient is the ratio of the number of actual to the number of potential edges
for nodes with degree greater than k:
2𝐸𝑘
𝜑(𝑘) =
𝑁𝑘 (𝑁𝑘 − 1)

where N_k is the number of nodes with degree larger than k, and E_k is the number of edges among those
nodes.
Parameters
• G (NetworkX graph) – Undirected graph with neither parallel edges nor self-loops.
• normalized (bool (optional)) – Normalize using randomized network as in1
• Q (float (optional, default=100)) – If normalized is True, perform Q * m double-edge
swaps, where m is the number of edges in G, to use as a null-model for normalization.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns rc – A dictionary, keyed by degree, with rich-club coefficient values.
Return type dictionary
1 Julian J. McAuley, Luciano da Fontoura Costa, and Tibério S. Caetano, “The rich-club phenomenon across complex network hierarchies”,

Applied Physics Letters Vol 91 Issue 8, August 2007. https://arxiv.org/abs/physics/0701290

400 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.Graph([(0, 1), (0, 2), (1, 2), (1, 3), (1, 4), (4, 5)])
>>> rc = nx.rich_club_coefficient(G, normalized=False)
>>> rc[0] # doctest: +SKIP
0.4

Notes

The rich club definition and algorithm are found in1 . This algorithm ignores any edge weights and is not defined
for directed graphs or graphs with parallel edges or self loops.
Estimates for appropriate values of Q are found in2 .

References

3.45 Shortest Paths

Compute the shortest paths and path lengths between nodes in the graph.
These algorithms work with undirected and directed graphs.

shortest_path(G[, source, target, weight, . . . ]) Compute shortest paths in the graph.


all_shortest_paths(G, source, target[, . . . ]) Compute all shortest paths in the graph.
shortest_path_length(G[, source, target, . . . ]) Compute shortest path lengths in the graph.
average_shortest_path_length(G[, weight, Returns the average shortest path length.
method])
has_path(G, source, target) Returns True if G has a path from source to target.

3.45.1 networkx.algorithms.shortest_paths.generic.shortest_path

shortest_path(G, source=None, target=None, weight=None, method=’dijkstra’)


Compute shortest paths in the graph.
Parameters
• G (NetworkX graph)
• source (node, optional) – Starting node for path. If not specified, compute shortest paths for
each possible starting node.
• target (node, optional) – Ending node for path. If not specified, compute shortest paths to
all possible nodes.
• weight (None or string, optional (default = None)) – If None, every edge has
weight/distance/cost 1. If a string, use this edge attribute as the edge weight. Any edge
attribute not present defaults to 1.
• method (string, optional (default = ‘dijkstra’)) – The algorithm to use to compute the
path. Supported options: ‘dijkstra’, ‘bellman-ford’. Other inputs produce a ValueError.
If weight is None, unweighted graph methods are used, and this suggestion is ignored.
2 R. Milo, N. Kashtan, S. Itzkovitz, M. E. J. Newman, U. Alon, “Uniform generation of random graphs with arbitrary degree sequences”, 2006.
https://arxiv.org/abs/cond-mat/0312028

3.45. Shortest Paths 401


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns
path – All returned paths include both the source and target in the path.
If the source and target are both specified, return a single list of nodes in a shortest path from the
source to the target.
If only the source is specified, return a dictionary keyed by targets with a list of nodes in a
shortest path from the source to one of the targets.
If only the target is specified, return a dictionary keyed by sources with a list of nodes in a
shortest path from one of the sources to the target.
If neither the source nor target are specified return a dictionary of dictionaries with
path[source][target]=[list of nodes in path].
Return type list or dictionary
Raises
• NodeNotFound – If source is not in G.
• ValueError – If method is not among the supported options.

Examples

>>> G = nx.path_graph(5)
>>> print(nx.shortest_path(G, source=0, target=4))
[0, 1, 2, 3, 4]
>>> p = nx.shortest_path(G, source=0) # target not specified
>>> p[4]
[0, 1, 2, 3, 4]
>>> p = nx.shortest_path(G, target=4) # source not specified
>>> p[0]
[0, 1, 2, 3, 4]
>>> p = nx.shortest_path(G) # source, target not specified
>>> p[0][4]
[0, 1, 2, 3, 4]

Notes

There may be more than one shortest path between a source and target. This returns only one of them.
See also:
all_pairs_shortest_path(), all_pairs_dijkstra_path(),
all_pairs_bellman_ford_path(), single_source_shortest_path(),
single_source_dijkstra_path(), single_source_bellman_ford_path()

3.45.2 networkx.algorithms.shortest_paths.generic.all_shortest_paths

all_shortest_paths(G, source, target, weight=None, method=’dijkstra’)


Compute all shortest paths in the graph.
Parameters
• G (NetworkX graph)
• source (node) – Starting node for path.

402 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• target (node) – Ending node for path.


• weight (None or string, optional (default = None)) – If None, every edge has
weight/distance/cost 1. If a string, use this edge attribute as the edge weight. Any edge
attribute not present defaults to 1.
• method (string, optional (default = ‘dijkstra’)) – The algorithm to use to compute the path
lengths. Supported options: ‘dijkstra’, ‘bellman-ford’. Other inputs produce a ValueError.
If weight is None, unweighted graph methods are used, and this suggestion is ignored.
Returns paths – A generator of all paths between source and target.
Return type generator of lists
Raises
• ValueError – If method is not among the supported options.
• NetworkXNoPath – If target cannot be reached from source.

Examples

>>> G = nx.Graph()
>>> nx.add_path(G, [0, 1, 2])
>>> nx.add_path(G, [0, 10, 2])
>>> print([p for p in nx.all_shortest_paths(G, source=0, target=2)])
[[0, 1, 2], [0, 10, 2]]

Notes

There may be many shortest paths between the source and target.
See also:
shortest_path(), single_source_shortest_path(), all_pairs_shortest_path()

3.45.3 networkx.algorithms.shortest_paths.generic.shortest_path_length

shortest_path_length(G, source=None, target=None, weight=None, method=’dijkstra’)


Compute shortest path lengths in the graph.
Parameters
• G (NetworkX graph)
• source (node, optional) – Starting node for path. If not specified, compute shortest path
lengths using all nodes as source nodes.
• target (node, optional) – Ending node for path. If not specified, compute shortest path
lengths using all nodes as target nodes.
• weight (None or string, optional (default = None)) – If None, every edge has
weight/distance/cost 1. If a string, use this edge attribute as the edge weight. Any edge
attribute not present defaults to 1.
• method (string, optional (default = ‘dijkstra’)) – The algorithm to use to compute the path
length. Supported options: ‘dijkstra’, ‘bellman-ford’. Other inputs produce a ValueError. If
weight is None, unweighted graph methods are used, and this suggestion is ignored.

3.45. Shortest Paths 403


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns
length – If the source and target are both specified, return the length of the shortest path from
the source to the target.
If only the source is specified, return a dict keyed by target to the shortest path length from the
source to that target.
If only the target is specified, return a dict keyed by source to the shortest path length from that
source to the target.
If neither the source nor target are specified, return an iterator over (source, dictionary) where
dictionary is keyed by target to shortest path length from source to that target.
Return type int or iterator
Raises
• NodeNotFound – If source is not in G.
• NetworkXNoPath – If no path exists between source and target.
• ValueError – If method is not among the supported options.

Examples

>>> G = nx.path_graph(5)
>>> nx.shortest_path_length(G, source=0, target=4)
4
>>> p = nx.shortest_path_length(G, source=0) # target not specified
>>> p[4]
4
>>> p = nx.shortest_path_length(G, target=4) # source not specified
>>> p[0]
4
>>> p = dict(nx.shortest_path_length(G)) # source,target not specified
>>> p[0][4]
4

Notes

The length of the path is always 1 less than the number of nodes involved in the path since the length measures
the number of edges followed.
For digraphs this returns the shortest directed path length. To find path lengths in the reverse direction use
G.reverse(copy=False) first to flip the edge orientation.
See also:
all_pairs_shortest_path_length(), all_pairs_dijkstra_path_length(),
all_pairs_bellman_ford_path_length(), single_source_shortest_path_length(),
single_source_dijkstra_path_length(), single_source_bellman_ford_path_length()

3.45.4 networkx.algorithms.shortest_paths.generic.average_shortest_path_length

average_shortest_path_length(G, weight=None, method=’dijkstra’)


Returns the average shortest path length.

404 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

The average shortest path length is


∑︁ 𝑑(𝑠, 𝑡)
𝑎=
𝑛(𝑛 − 1)
𝑠,𝑡∈𝑉

where V is the set of nodes in G, d(s, t) is the shortest path from s to t, and n is the number of nodes in G.
Parameters
• G (NetworkX graph)
• weight (None or string, optional (default = None)) – If None, every edge has
weight/distance/cost 1. If a string, use this edge attribute as the edge weight. Any edge
attribute not present defaults to 1.
• method (string, optional (default = ‘dijkstra’)) – The algorithm to use to compute the path
lengths. Supported options: ‘dijkstra’, ‘bellman-ford’. Other inputs produce a ValueError.
If weight is None, unweighted graph methods are used, and this suggestion is ignored.
Raises
• NetworkXPointlessConcept – If G is the null graph (that is, the graph on zero nodes).
• NetworkXError – If G is not connected (or not weakly connected, in the case of a directed
graph).
• ValueError – If method is not among the supported options.

Examples

>>> G = nx.path_graph(5)
>>> nx.average_shortest_path_length(G)
2.0

For disconnected graphs, you can compute the average shortest path length for each component

>>> G = nx.Graph([(1, 2), (3, 4)])


>>> for C in nx.connected_component_subgraphs(G):
... print(nx.average_shortest_path_length(C))
1.0
1.0

3.45.5 networkx.algorithms.shortest_paths.generic.has_path

has_path(G, source, target)


Returns True if G has a path from source to target.
Parameters
• G (NetworkX graph)
• source (node) – Starting node for path
• target (node) – Ending node for path

3.45. Shortest Paths 405


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.45.6 Advanced Interface

Shortest path algorithms for unweighted graphs.

single_source_shortest_path(G, source[, Compute shortest path between source and all other
cutoff]) nodes reachable from source.
single_source_shortest_path_length(G, Compute the shortest path lengths from source to all
source) reachable nodes.
single_target_shortest_path(G, target[, Compute shortest path to target from all nodes that reach
cutoff]) target.
single_target_shortest_path_length(G, Compute the shortest path lengths to target from all
target) reachable nodes.
bidirectional_shortest_path(G, source, tar- Returns a list of nodes in a shortest path between source
get) and target.
all_pairs_shortest_path(G[, cutoff]) Compute shortest paths between all nodes.
all_pairs_shortest_path_length(G[, cut- Computes the shortest path lengths between all nodes in
off]) G.
predecessor(G, source[, target, cutoff, . . . ]) Returns dict of predecessors for the path from source to
all nodes in G

networkx.algorithms.shortest_paths.unweighted.single_source_shortest_path

single_source_shortest_path(G, source, cutoff=None)


Compute shortest path between source and all other nodes reachable from source.
Parameters
• G (NetworkX graph)
• source (node label) – Starting node for path
• cutoff (integer, optional) – Depth to stop the search. Only paths of length <= cutoff are
returned.
Returns lengths – Dictionary, keyed by target, of shortest paths.
Return type dictionary

Examples

>>> G = nx.path_graph(5)
>>> path = nx.single_source_shortest_path(G, 0)
>>> path[4]
[0, 1, 2, 3, 4]

Notes

The shortest path is not necessarily unique. So there can be multiple paths between the source and each target
node, all of which have the same ‘shortest’ length. For each target node, this function returns only one of those
paths.
See also:
shortest_path()

406 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.shortest_paths.unweighted.single_source_shortest_path_length

single_source_shortest_path_length(G, source, cutoff=None)


Compute the shortest path lengths from source to all reachable nodes.
Parameters
• G (NetworkX graph)
• source (node) – Starting node for path
• cutoff (integer, optional) – Depth to stop the search. Only paths of length <= cutoff are
returned.
Returns lengths – Dict keyed by node to shortest path length to source.
Return type dict

Examples

>>> G = nx.path_graph(5)
>>> length = nx.single_source_shortest_path_length(G, 0)
>>> length[4]
4
>>> for node in length:
... print('{}: {}'.format(node, length[node]))
0: 0
1: 1
2: 2
3: 3
4: 4

See also:
shortest_path_length()

networkx.algorithms.shortest_paths.unweighted.single_target_shortest_path

single_target_shortest_path(G, target, cutoff=None)


Compute shortest path to target from all nodes that reach target.
Parameters
• G (NetworkX graph)
• target (node label) – Target node for path
• cutoff (integer, optional) – Depth to stop the search. Only paths of length <= cutoff are
returned.
Returns lengths – Dictionary, keyed by target, of shortest paths.
Return type dictionary

Examples

3.45. Shortest Paths 407


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.path_graph(5, create_using=nx.DiGraph())


>>> path = nx.single_target_shortest_path(G, 4)
>>> path[0]
[0, 1, 2, 3, 4]

Notes

The shortest path is not necessarily unique. So there can be multiple paths between the source and each target
node, all of which have the same ‘shortest’ length. For each target node, this function returns only one of those
paths.
See also:
shortest_path(), single_source_shortest_path()

networkx.algorithms.shortest_paths.unweighted.single_target_shortest_path_length

single_target_shortest_path_length(G, target, cutoff=None)


Compute the shortest path lengths to target from all reachable nodes.
Parameters
• G (NetworkX graph)
• target (node) – Target node for path
• cutoff (integer, optional) – Depth to stop the search. Only paths of length <= cutoff are
returned.
Returns lengths – (source, shortest path length) iterator
Return type iterator

Examples

>>> G = nx.path_graph(5, create_using=nx.DiGraph())


>>> length = dict(nx.single_target_shortest_path_length(G, 4))
>>> length[0]
4
>>> for node in range(5):
... print('{}: {}'.format(node, length[node]))
0: 4
1: 3
2: 2
3: 1
4: 0

See also:
single_source_shortest_path_length(), shortest_path_length()

networkx.algorithms.shortest_paths.unweighted.bidirectional_shortest_path

bidirectional_shortest_path(G, source, target)


Returns a list of nodes in a shortest path between source and target.

408 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (NetworkX graph)
• source (node label) – starting node for path
• target (node label) – ending node for path
Returns path – List of nodes in a path from source to target.
Return type list
Raises NetworkXNoPath – If no path exists between source and target.
See also:
shortest_path()

Notes

This algorithm is used by shortest_path(G, source, target).

networkx.algorithms.shortest_paths.unweighted.all_pairs_shortest_path

all_pairs_shortest_path(G, cutoff=None)
Compute shortest paths between all nodes.
Parameters
• G (NetworkX graph)
• cutoff (integer, optional) – Depth at which to stop the search. Only paths of length at most
cutoff are returned.
Returns lengths – Dictionary, keyed by source and target, of shortest paths.
Return type dictionary

Examples

>>> G = nx.path_graph(5)
>>> path = dict(nx.all_pairs_shortest_path(G))
>>> print(path[0][4])
[0, 1, 2, 3, 4]

See also:
floyd_warshall()

networkx.algorithms.shortest_paths.unweighted.all_pairs_shortest_path_length

all_pairs_shortest_path_length(G, cutoff=None)
Computes the shortest path lengths between all nodes in G.
Parameters
• G (NetworkX graph)
• cutoff (integer, optional) – Depth at which to stop the search. Only paths of length at most
cutoff are returned.

3.45. Shortest Paths 409


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns lengths – (source, dictionary) iterator with dictionary keyed by target and shortest path
length as the key value.
Return type iterator

Notes

The iterator returned only has reachable node pairs.

Examples

>>> G = nx.path_graph(5)
>>> length = dict(nx.all_pairs_shortest_path_length(G))
>>> for node in [0, 1, 2, 3, 4]:
... print('1 - {}: {}'.format(node, length[1][node]))
1 - 0: 1
1 - 1: 0
1 - 2: 1
1 - 3: 2
1 - 4: 3
>>> length[3][2]
1
>>> length[2][2]
0

networkx.algorithms.shortest_paths.unweighted.predecessor

predecessor(G, source, target=None, cutoff=None, return_seen=None)


Returns dict of predecessors for the path from source to all nodes in G
Parameters
• G (NetworkX graph)
• source (node label) – Starting node for path
• target (node label, optional) – Ending node for path. If provided only predecessors between
source and target are returned
• cutoff (integer, optional) – Depth to stop the search. Only paths of length <= cutoff are
returned.
Returns pred – Dictionary, keyed by node, of predecessors in the shortest path.
Return type dictionary

Examples

>>> G = nx.path_graph(4)
>>> list(G)
[0, 1, 2, 3]
>>> nx.predecessor(G, 0)
{0: [], 1: [0], 2: [1], 3: [2]}

410 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Shortest path algorithms for weighed graphs.

dijkstra_predecessor_and_distance(G, Compute weighted shortest path length and predeces-


source) sors.
dijkstra_path(G, source, target[, weight]) Returns the shortest weighted path from source to target
in G.
dijkstra_path_length(G, source, target[, Returns the shortest weighted path length in G from
weight]) source to target.
single_source_dijkstra(G, source[, target, Find shortest weighted paths and lengths from a source
. . . ]) node.
single_source_dijkstra_path(G, source[, Find shortest weighted paths in G from a source node.
. . . ])
single_source_dijkstra_path_length(G, Find shortest weighted path lengths in G from a source
source) node.
multi_source_dijkstra(G, sources[, target, Find shortest weighted paths and lengths from a given
. . . ]) set of source nodes.
multi_source_dijkstra_path(G, sources[, Find shortest weighted paths in G from a given set of
. . . ]) source nodes.
multi_source_dijkstra_path_length(G, Find shortest weighted path lengths in G from a given
sources) set of source nodes.
all_pairs_dijkstra(G[, cutoff, weight]) Find shortest weighted paths and lengths between all
nodes.
all_pairs_dijkstra_path(G[, cutoff, weight]) Compute shortest paths between all nodes in a weighted
graph.
all_pairs_dijkstra_path_length(G[, cut- Compute shortest path lengths between all nodes in a
off, . . . ]) weighted graph.
bidirectional_dijkstra(G, source, target[, Dijkstra’s algorithm for shortest paths using bidirec-
. . . ]) tional search.
bellman_ford_path(G, source, target[, weight]) Returns the shortest path from source to target in a
weighted graph G.
bellman_ford_path_length(G, source, target) Returns the shortest path length from source to target in
a weighted graph.
single_source_bellman_ford(G, source[, Compute shortest paths and lengths in a weighted graph
. . . ]) G.
single_source_bellman_ford_path(G, Compute shortest path between source and all other
source[, . . . ]) reachable nodes for a weighted graph.
single_source_bellman_ford_path_length(G, Compute the shortest path length between source and all
source) other reachable nodes for a weighted graph.
all_pairs_bellman_ford_path(G[, weight]) Compute shortest paths between all nodes in a weighted
graph.
all_pairs_bellman_ford_path_length(G[, Compute shortest path lengths between all nodes in a
weight]) weighted graph.
bellman_ford_predecessor_and_distance(G,Compute shortest path lengths and predecessors on
source) shortest paths in weighted graphs.
negative_edge_cycle(G[, weight]) Returns True if there exists a negative edge cycle any-
where in G.
goldberg_radzik(G, source[, weight]) Compute shortest path lengths and predecessors on
shortest paths in weighted graphs.
johnson(G[, weight]) Uses Johnson’s Algorithm to compute shortest paths.

3.45. Shortest Paths 411


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.shortest_paths.weighted.dijkstra_predecessor_and_distance

dijkstra_predecessor_and_distance(G, source, cutoff=None, weight=’weight’)


Compute weighted shortest path length and predecessors.
Uses Dijkstra’s Method to obtain the shortest weighted paths and return dictionaries of predecessors for each
node and distance for each node from the source.
Parameters
• G (NetworkX graph)
• source (node label) – Starting node for path
• cutoff (integer or float, optional) – Depth to stop the search. Only return paths with length
<= cutoff.
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns pred, distance – Returns two dictionaries representing a list of predecessors of a node and
the distance to each node. Warning: If target is specified, the dicts are incomplete as they only
contain information for the nodes along a path to target.
Return type dictionaries
Raises NodeNotFound – If source is not in G.

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The list of predecessors contains more than one element only when there are more than one shortest paths to the
key node.

Examples

>>> import networkx as nx


>>> G = nx.path_graph(5, create_using = nx.DiGraph())
>>> pred, dist = nx.dijkstra_predecessor_and_distance(G, 0)
>>> sorted(pred.items())
[(0, []), (1, [0]), (2, [1]), (3, [2]), (4, [3])]
>>> sorted(dist.items())
[(0, 0), (1, 1), (2, 2), (3, 3), (4, 4)]

>>> pred, dist = nx.dijkstra_predecessor_and_distance(G, 0, 1)


>>> sorted(pred.items())
[(0, []), (1, [0])]
>>> sorted(dist.items())
[(0, 0), (1, 1)]

412 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.shortest_paths.weighted.dijkstra_path

dijkstra_path(G, source, target, weight=’weight’)


Returns the shortest weighted path from source to target in G.
Uses Dijkstra’s Method to compute the shortest weighted path between two nodes in a graph.
Parameters
• G (NetworkX graph)
• source (node) – Starting node
• target (node) – Ending node
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns path – List of nodes in a shortest path.
Return type list
Raises
• NodeNotFound – If source is not in G.
• NetworkXNoPath – If no path exists between source and target.

Examples

>>> G=nx.path_graph(5)
>>> print(nx.dijkstra_path(G,0,4))
[0, 1, 2, 3, 4]

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The weight function can be used to hide edges by returning None. So weight = lambda u, v, d: 1
if d['color']=="red" else None will find the shortest red path.
The weight function can be used to include node weights.
>>> def func(u, v, d):
... node_u_wt = G.nodes[u].get('node_weight', 1)
... node_v_wt = G.nodes[v].get('node_weight', 1)
... edge_wt = d.get('weight', 1)
... return node_u_wt/2 + node_v_wt/2 + edge_wt

In this example we take the average of start and end node weights of an edge and add it to the weight of the
edge.
See also:
bidirectional_dijkstra(), bellman_ford_path()

3.45. Shortest Paths 413


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.shortest_paths.weighted.dijkstra_path_length

dijkstra_path_length(G, source, target, weight=’weight’)


Returns the shortest weighted path length in G from source to target.
Uses Dijkstra’s Method to compute the shortest weighted path length between two nodes in a graph.
Parameters
• G (NetworkX graph)
• source (node label) – starting node for path
• target (node label) – ending node for path
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns length – Shortest path length.
Return type number
Raises
• NodeNotFound – If source is not in G.
• NetworkXNoPath – If no path exists between source and target.

Examples

>>> G=nx.path_graph(5)
>>> print(nx.dijkstra_path_length(G,0,4))
4

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The weight function can be used to hide edges by returning None. So weight = lambda u, v, d: 1
if d['color']=="red" else None will find the shortest red path.
See also:
bidirectional_dijkstra(), bellman_ford_path_length()

networkx.algorithms.shortest_paths.weighted.single_source_dijkstra

single_source_dijkstra(G, source, target=None, cutoff=None, weight=’weight’)


Find shortest weighted paths and lengths from a source node.
Compute the shortest path length between source and all other reachable nodes for a weighted graph.
Uses Dijkstra’s algorithm to compute shortest paths and lengths between a source and all other reachable nodes
in a weighted graph.

414 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (NetworkX graph)
• source (node label) – Starting node for path
• target (node label, optional) – Ending node for path
• cutoff (integer or float, optional) – Depth to stop the search. Only return paths with length
<= cutoff.
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns distance, path – If target is None, paths and lengths to all nodes are computed. The return
value is a tuple of two dictionaries keyed by target nodes. The first dictionary stores distance to
each target node. The second stores the path to each target node. If target is not None, returns
a tuple (distance, path), where distance is the distance from source to target and path is a list
representing the path from source to target.
Return type pair of dictionaries, or numeric and list.
Raises NodeNotFound – If source is not in G.

Examples

>>> G = nx.path_graph(5)
>>> length, path = nx.single_source_dijkstra(G, 0)
>>> print(length[4])
4
>>> for node in [0, 1, 2, 3, 4]:
... print('{}: {}'.format(node, length[node]))
0: 0
1: 1
2: 2
3: 3
4: 4
>>> path[4]
[0, 1, 2, 3, 4]
>>> length, path = nx.single_source_dijkstra(G, 0, 1)
>>> length
1
>>> path
[0, 1]

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The weight function can be used to hide edges by returning None. So weight = lambda u, v, d: 1
if d['color']=="red" else None will find the shortest red path.

3.45. Shortest Paths 415


NetworkX Reference, Release 2.3rc1.dev20190222214247

Based on the Python cookbook recipe (119466) at http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/


119466
This algorithm is not guaranteed to work if edge weights are negative or are floating point numbers (overflows
and roundoff errors can cause problems).
See also:
single_source_dijkstra_path(), single_source_dijkstra_path_length(),
single_source_bellman_ford()

networkx.algorithms.shortest_paths.weighted.single_source_dijkstra_path

single_source_dijkstra_path(G, source, cutoff=None, weight=’weight’)


Find shortest weighted paths in G from a source node.
Compute shortest path between source and all other reachable nodes for a weighted graph.
Parameters
• G (NetworkX graph)
• source (node) – Starting node for path.
• cutoff (integer or float, optional) – Depth to stop the search. Only return paths with length
<= cutoff.
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns paths – Dictionary of shortest path lengths keyed by target.
Return type dictionary
Raises NodeNotFound – If source is not in G.

Examples

>>> G=nx.path_graph(5)
>>> path=nx.single_source_dijkstra_path(G,0)
>>> path[4]
[0, 1, 2, 3, 4]

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The weight function can be used to hide edges by returning None. So weight = lambda u, v, d: 1
if d['color']=="red" else None will find the shortest red path.
See also:
single_source_dijkstra(), single_source_bellman_ford()

416 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.shortest_paths.weighted.single_source_dijkstra_path_length

single_source_dijkstra_path_length(G, source, cutoff=None, weight=’weight’)


Find shortest weighted path lengths in G from a source node.
Compute the shortest path length between source and all other reachable nodes for a weighted graph.
Parameters
• G (NetworkX graph)
• source (node label) – Starting node for path
• cutoff (integer or float, optional) – Depth to stop the search. Only return paths with length
<= cutoff.
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns length – Dict keyed by node to shortest path length from source.
Return type dict
Raises NodeNotFound – If source is not in G.

Examples

>>> G = nx.path_graph(5)
>>> length = nx.single_source_dijkstra_path_length(G, 0)
>>> length[4]
4
>>> for node in [0, 1, 2, 3, 4]:
... print('{}: {}'.format(node, length[node]))
0: 0
1: 1
2: 2
3: 3
4: 4

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The weight function can be used to hide edges by returning None. So weight = lambda u, v, d: 1
if d['color']=="red" else None will find the shortest red path.
See also:
single_source_dijkstra(), single_source_bellman_ford_path_length()

3.45. Shortest Paths 417


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.shortest_paths.weighted.multi_source_dijkstra

multi_source_dijkstra(G, sources, target=None, cutoff=None, weight=’weight’)


Find shortest weighted paths and lengths from a given set of source nodes.
Uses Dijkstra’s algorithm to compute the shortest paths and lengths between one of the source nodes and the
given target, or all other reachable nodes if not specified, for a weighted graph.
Parameters
• G (NetworkX graph)
• sources (non-empty set of nodes) – Starting nodes for paths. If this is just a set containing a
single node, then all paths computed by this function will start from that node. If there are
two or more nodes in the set, the computed paths may begin from any one of the start nodes.
• target (node label, optional) – Ending node for path
• cutoff (integer or float, optional) – Depth to stop the search. Only return paths with length
<= cutoff.
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns distance, path – If target is None, returns a tuple of two dictionaries keyed by node. The
first dictionary stores distance from one of the source nodes. The second stores the path from
one of the sources to that node. If target is not None, returns a tuple of (distance, path) where
distance is the distance from source to target and path is a list representing the path from source
to target.
Return type pair of dictionaries, or numeric and list

Examples

>>> G = nx.path_graph(5)
>>> length, path = nx.multi_source_dijkstra(G, {0, 4})
>>> for node in [0, 1, 2, 3, 4]:
... print('{}: {}'.format(node, length[node]))
0: 0
1: 1
2: 2
3: 1
4: 0
>>> path[1]
[0, 1]
>>> path[3]
[4, 3]

>>> length, path = nx.multi_source_dijkstra(G, {0, 4}, 1)


>>> length
1
>>> path
[0, 1]

418 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The weight function can be used to hide edges by returning None. So weight = lambda u, v, d: 1
if d['color']=="red" else None will find the shortest red path.
Based on the Python cookbook recipe (119466) at http://aspn.activestate.com/ASPN/Cookbook/Python/Recipe/
119466
This algorithm is not guaranteed to work if edge weights are negative or are floating point numbers (overflows
and roundoff errors can cause problems).
Raises
• ValueError – If sources is empty.
• NodeNotFound – If any of sources is not in G.
See also:
multi_source_dijkstra_path(), multi_source_dijkstra_path_length()

networkx.algorithms.shortest_paths.weighted.multi_source_dijkstra_path

multi_source_dijkstra_path(G, sources, cutoff=None, weight=’weight’)


Find shortest weighted paths in G from a given set of source nodes.
Compute shortest path between any of the source nodes and all other reachable nodes for a weighted graph.
Parameters
• G (NetworkX graph)
• sources (non-empty set of nodes) – Starting nodes for paths. If this is just a set containing a
single node, then all paths computed by this function will start from that node. If there are
two or more nodes in the set, the computed paths may begin from any one of the start nodes.
• cutoff (integer or float, optional) – Depth to stop the search. Only return paths with length
<= cutoff.
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns paths – Dictionary of shortest paths keyed by target.
Return type dictionary

Examples

>>> G = nx.path_graph(5)
>>> path = nx.multi_source_dijkstra_path(G, {0, 4})
>>> path[1]
[0, 1]
(continues on next page)

3.45. Shortest Paths 419


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> path[3]
[4, 3]

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The weight function can be used to hide edges by returning None. So weight = lambda u, v, d: 1
if d['color']=="red" else None will find the shortest red path.
Raises
• ValueError – If sources is empty.
• NodeNotFound – If any of sources is not in G.
See also:
multi_source_dijkstra(), multi_source_bellman_ford()

networkx.algorithms.shortest_paths.weighted.multi_source_dijkstra_path_length

multi_source_dijkstra_path_length(G, sources, cutoff=None, weight=’weight’)


Find shortest weighted path lengths in G from a given set of source nodes.
Compute the shortest path length between any of the source nodes and all other reachable nodes for a weighted
graph.
Parameters
• G (NetworkX graph)
• sources (non-empty set of nodes) – Starting nodes for paths. If this is just a set containing a
single node, then all paths computed by this function will start from that node. If there are
two or more nodes in the set, the computed paths may begin from any one of the start nodes.
• cutoff (integer or float, optional) – Depth to stop the search. Only return paths with length
<= cutoff.
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns length – Dict keyed by node to shortest path length to nearest source.
Return type dict

Examples

420 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.path_graph(5)
>>> length = nx.multi_source_dijkstra_path_length(G, {0, 4})
>>> for node in [0, 1, 2, 3, 4]:
... print('{}: {}'.format(node, length[node]))
0: 0
1: 1
2: 2
3: 1
4: 0

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The weight function can be used to hide edges by returning None. So weight = lambda u, v, d: 1
if d['color']=="red" else None will find the shortest red path.
Raises
• ValueError – If sources is empty.
• NodeNotFound – If any of sources is not in G.
See also:
multi_source_dijkstra()

networkx.algorithms.shortest_paths.weighted.all_pairs_dijkstra

all_pairs_dijkstra(G, cutoff=None, weight=’weight’)


Find shortest weighted paths and lengths between all nodes.
Parameters
• G (NetworkX graph)
• cutoff (integer or float, optional) – Depth to stop the search. Only return paths with length
<= cutoff.
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be
G.edge[u][v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Yields (node, (distance, path)) ((node obj, (dict, dict))) – Each source node has two associated
dicts. The first holds distance keyed by target and the second holds paths keyed by target.
(See single_source_dijkstra for the source/target node terminology.) If desired you can apply
dict() to this function to create a dict keyed by source node to the two dicts.

Examples

3.45. Shortest Paths 421


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.path_graph(5)
>>> len_path = dict(nx.all_pairs_dijkstra(G))
>>> print(len_path[3][0][1])
2
>>> for node in [0, 1, 2, 3, 4]:
... print('3 - {}: {}'.format(node, len_path[3][0][node]))
3 - 0: 3
3 - 1: 2
3 - 2: 1
3 - 3: 0
3 - 4: 1
>>> len_path[3][1][1]
[3, 2, 1]
>>> for n, (dist, path) in nx.all_pairs_dijkstra(G):
... print(path[1])
[0, 1]
[1]
[2, 1]
[3, 2, 1]
[4, 3, 2, 1]

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The yielded dicts only have keys for reachable nodes.

networkx.algorithms.shortest_paths.weighted.all_pairs_dijkstra_path

all_pairs_dijkstra_path(G, cutoff=None, weight=’weight’)


Compute shortest paths between all nodes in a weighted graph.
Parameters
• G (NetworkX graph)
• cutoff (integer or float, optional) – Depth to stop the search. Only return paths with length
<= cutoff.
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns distance – Dictionary, keyed by source and target, of shortest paths.
Return type dictionary

Examples

422 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.path_graph(5)
>>> path = dict(nx.all_pairs_dijkstra_path(G))
>>> print(path[0][4])
[0, 1, 2, 3, 4]

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
See also:
floyd_warshall(), all_pairs_bellman_ford_path()

networkx.algorithms.shortest_paths.weighted.all_pairs_dijkstra_path_length

all_pairs_dijkstra_path_length(G, cutoff=None, weight=’weight’)


Compute shortest path lengths between all nodes in a weighted graph.
Parameters
• G (NetworkX graph)
• cutoff (integer or float, optional) – Depth to stop the search. Only return paths with length
<= cutoff.
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns distance – (source, dictionary) iterator with dictionary keyed by target and shortest path
length as the key value.
Return type iterator

Examples

>>> G = nx.path_graph(5)
>>> length = dict(nx.all_pairs_dijkstra_path_length(G))
>>> for node in [0, 1, 2, 3, 4]:
... print('1 - {}: {}'.format(node, length[1][node]))
1 - 0: 1
1 - 1: 0
1 - 2: 1
1 - 3: 2
1 - 4: 3
>>> length[3][2]
1
>>> length[2][2]
0

3.45. Shortest Paths 423


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The dictionary returned only has keys for reachable node pairs.

networkx.algorithms.shortest_paths.weighted.bidirectional_dijkstra

bidirectional_dijkstra(G, source, target, weight=’weight’)


Dijkstra’s algorithm for shortest paths using bidirectional search.
Parameters
• G (NetworkX graph)
• source (node) – Starting node.
• target (node) – Ending node.
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns length, path – length is the distance from source to target. path is a list of nodes on a path
from source to target.
Return type number and list
Raises
• NodeNotFound – If either source or target is not in G.
• NetworkXNoPath – If no path exists between source and target.

Examples

>>> G = nx.path_graph(5)
>>> length, path = nx.bidirectional_dijkstra(G, 0, 4)
>>> print(length)
4
>>> print(path)
[0, 1, 2, 3, 4]

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
In practice bidirectional Dijkstra is much more than twice as fast as ordinary Dijkstra.
Ordinary Dijkstra expands nodes in a sphere-like manner from the source. The radius of this sphere will even-
tually be the length of the shortest path. Bidirectional Dijkstra will expand nodes from both the source and
the target, making two spheres of half this radius. Volume of the first sphere is pi*r*r while the others are
2*pi*r/2*r/2, making up half the volume.

424 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

This algorithm is not guaranteed to work if edge weights are negative or are floating point numbers (overflows
and roundoff errors can cause problems).
See also:
shortest_path(), shortest_path_length()

networkx.algorithms.shortest_paths.weighted.bellman_ford_path

bellman_ford_path(G, source, target, weight=’weight’)


Returns the shortest path from source to target in a weighted graph G.
Parameters
• G (NetworkX graph)
• source (node) – Starting node
• target (node) – Ending node
• weight (string, optional (default=’weight’)) – Edge data key corresponding to the edge
weight
Returns path – List of nodes in a shortest path.
Return type list
Raises
• NodeNotFound – If source is not in G.
• NetworkXNoPath – If no path exists between source and target.

Examples

>>> G=nx.path_graph(5)
>>> print(nx.bellman_ford_path(G, 0, 4))
[0, 1, 2, 3, 4]

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
See also:
dijkstra_path(), bellman_ford_path_length()

networkx.algorithms.shortest_paths.weighted.bellman_ford_path_length

bellman_ford_path_length(G, source, target, weight=’weight’)


Returns the shortest path length from source to target in a weighted graph.
Parameters
• G (NetworkX graph)
• source (node label) – starting node for path
• target (node label) – ending node for path

3.45. Shortest Paths 425


NetworkX Reference, Release 2.3rc1.dev20190222214247

• weight (string, optional (default=’weight’)) – Edge data key corresponding to the edge
weight
Returns length – Shortest path length.
Return type number
Raises
• NodeNotFound – If source is not in G.
• NetworkXNoPath – If no path exists between source and target.

Examples

>>> G=nx.path_graph(5)
>>> print(nx.bellman_ford_path_length(G,0,4))
4

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
See also:
dijkstra_path_length(), bellman_ford_path()

networkx.algorithms.shortest_paths.weighted.single_source_bellman_ford

single_source_bellman_ford(G, source, target=None, weight=’weight’)


Compute shortest paths and lengths in a weighted graph G.
Uses Bellman-Ford algorithm for shortest paths.
Parameters
• G (NetworkX graph)
• source (node label) – Starting node for path
• target (node label, optional) – Ending node for path
Returns distance, path – If target is None, returns a tuple of two dictionaries keyed by node. The
first dictionary stores distance from one of the source nodes. The second stores the path from
one of the sources to that node. If target is not None, returns a tuple of (distance, path) where
distance is the distance from source to target and path is a list representing the path from source
to target.
Return type pair of dictionaries, or numeric and list
Raises NodeNotFound – If source is not in G.

Examples

426 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.path_graph(5)
>>> length, path = nx.single_source_bellman_ford(G, 0)
>>> print(length[4])
4
>>> for node in [0, 1, 2, 3, 4]:
... print('{}: {}'.format(node, length[node]))
0: 0
1: 1
2: 2
3: 3
4: 4
>>> path[4]
[0, 1, 2, 3, 4]
>>> length, path = nx.single_source_bellman_ford(G, 0, 1)
>>> length
1
>>> path
[0, 1]

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
See also:
single_source_dijkstra(), single_source_bellman_ford_path(),
single_source_bellman_ford_path_length()

networkx.algorithms.shortest_paths.weighted.single_source_bellman_ford_path

single_source_bellman_ford_path(G, source, weight=’weight’)


Compute shortest path between source and all other reachable nodes for a weighted graph.
Parameters
• G (NetworkX graph)
• source (node) – Starting node for path.
• weight (string, optional (default=’weight’)) – Edge data key corresponding to the edge
weight
Returns paths – Dictionary of shortest path lengths keyed by target.
Return type dictionary
Raises NodeNotFound – If source is not in G.

Examples

>>> G=nx.path_graph(5)
>>> path=nx.single_source_bellman_ford_path(G,0)
>>> path[4]
[0, 1, 2, 3, 4]

3.45. Shortest Paths 427


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
See also:
single_source_dijkstra(), single_source_bellman_ford()

networkx.algorithms.shortest_paths.weighted.single_source_bellman_ford_path_length

single_source_bellman_ford_path_length(G, source, weight=’weight’)


Compute the shortest path length between source and all other reachable nodes for a weighted graph.
Parameters
• G (NetworkX graph)
• source (node label) – Starting node for path
• weight (string, optional (default=’weight’)) – Edge data key corresponding to the edge
weight.
Returns length – (target, shortest path length) iterator
Return type iterator
Raises NodeNotFound – If source is not in G.

Examples

>>> G = nx.path_graph(5)
>>> length = dict(nx.single_source_bellman_ford_path_length(G, 0))
>>> length[4]
4
>>> for node in [0, 1, 2, 3, 4]:
... print('{}: {}'.format(node, length[node]))
0: 0
1: 1
2: 2
3: 3
4: 4

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
See also:
single_source_dijkstra(), single_source_bellman_ford()

networkx.algorithms.shortest_paths.weighted.all_pairs_bellman_ford_path

all_pairs_bellman_ford_path(G, weight=’weight’)
Compute shortest paths between all nodes in a weighted graph.
Parameters
• G (NetworkX graph)

428 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• weight (string, optional (default=’weight’)) – Edge data key corresponding to the edge
weight
Returns distance – Dictionary, keyed by source and target, of shortest paths.
Return type dictionary

Examples

>>> G = nx.path_graph(5)
>>> path = dict(nx.all_pairs_bellman_ford_path(G))
>>> print(path[0][4])
[0, 1, 2, 3, 4]

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
See also:
floyd_warshall(), all_pairs_dijkstra_path()

networkx.algorithms.shortest_paths.weighted.all_pairs_bellman_ford_path_length

all_pairs_bellman_ford_path_length(G, weight=’weight’)
Compute shortest path lengths between all nodes in a weighted graph.
Parameters
• G (NetworkX graph)
• weight (string, optional (default=’weight’)) – Edge data key corresponding to the edge
weight
Returns distance – (source, dictionary) iterator with dictionary keyed by target and shortest path
length as the key value.
Return type iterator

Examples

>>> G = nx.path_graph(5)
>>> length = dict(nx.all_pairs_bellman_ford_path_length(G))
>>> for node in [0, 1, 2, 3, 4]:
... print('1 - {}: {}'.format(node, length[1][node]))
1 - 0: 1
1 - 1: 0
1 - 2: 1
1 - 3: 2
1 - 4: 3
>>> length[3][2]
1
>>> length[2][2]
0

3.45. Shortest Paths 429


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The dictionary returned only has keys for reachable node pairs.

networkx.algorithms.shortest_paths.weighted.bellman_ford_predecessor_and_distance

bellman_ford_predecessor_and_distance(G, source, target=None, weight=’weight’)


Compute shortest path lengths and predecessors on shortest paths in weighted graphs.
The algorithm has a running time of 𝑂(𝑚𝑛) where 𝑛 is the number of nodes and 𝑚 is the number of edges. It
is slower than Dijkstra but can handle negative edge weights.
Parameters
• G (NetworkX graph) – The algorithm works for all types of graphs, including directed
graphs and multigraphs.
• source (node label) – Starting node for path
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns pred, dist – Returns two dictionaries keyed by node to predecessor in the path and to the
distance from the source respectively.
Return type dictionaries
Raises
• NodeNotFound – If source is not in G.
• NetworkXUnbounded – If the (di)graph contains a negative cost (di)cycle, the algorithm
raises an exception to indicate the presence of the negative cost (di)cycle. Note: any negative
weight edge in an undirected graph is a negative cost cycle.

Examples

>>> import networkx as nx


>>> G = nx.path_graph(5, create_using = nx.DiGraph())
>>> pred, dist = nx.bellman_ford_predecessor_and_distance(G, 0)
>>> sorted(pred.items())
[(0, []), (1, [0]), (2, [1]), (3, [2]), (4, [3])]
>>> sorted(dist.items())
[(0, 0), (1, 1), (2, 2), (3, 3), (4, 4)]

>>> pred, dist = nx.bellman_ford_predecessor_and_distance(G, 0, 1)


>>> sorted(pred.items())
[(0, []), (1, [0]), (2, [1]), (3, [2]), (4, [3])]
>>> sorted(dist.items())
[(0, 0), (1, 1), (2, 2), (3, 3), (4, 4)]

430 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> from nose.tools import assert_raises


>>> G = nx.cycle_graph(5, create_using = nx.DiGraph())
>>> G[1][2]['weight'] = -7
>>> assert_raises(nx.NetworkXUnbounded, nx.bellman_ford_
˓→predecessor_and_distance, G, 0)

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The dictionaries returned only have keys for nodes reachable from the source.
In the case where the (di)graph is not connected, if a component not containing the source contains a negative
cost (di)cycle, it will not be detected.
In NetworkX v2.1 and prior, the source node had predecessor [None]. In NetworkX v2.2 this changed to the
source node having predecessor []

networkx.algorithms.shortest_paths.weighted.negative_edge_cycle

negative_edge_cycle(G, weight=’weight’)
Returns True if there exists a negative edge cycle anywhere in G.
Parameters
• G (NetworkX graph)
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns negative_cycle – True if a negative edge cycle exists, otherwise False.
Return type bool

Examples

>>> import networkx as nx


>>> G = nx.cycle_graph(5, create_using = nx.DiGraph())
>>> print(nx.negative_edge_cycle(G))
False
>>> G[1][2]['weight'] = -7
>>> print(nx.negative_edge_cycle(G))
True

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.

3.45. Shortest Paths 431


NetworkX Reference, Release 2.3rc1.dev20190222214247

This algorithm uses bellman_ford_predecessor_and_distance() but finds negative cycles on any component by
first adding a new node connected to every node, and starting bellman_ford_predecessor_and_distance on that
node. It then removes that extra node.

networkx.algorithms.shortest_paths.weighted.goldberg_radzik

goldberg_radzik(G, source, weight=’weight’)


Compute shortest path lengths and predecessors on shortest paths in weighted graphs.
The algorithm has a running time of 𝑂(𝑚𝑛) where 𝑛 is the number of nodes and 𝑚 is the number of edges. It
is slower than Dijkstra but can handle negative edge weights.
Parameters
• G (NetworkX graph) – The algorithm works for all types of graphs, including directed
graphs and multigraphs.
• source (node label) – Starting node for path
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns pred, dist – Returns two dictionaries keyed by node to predecessor in the path and to the
distance from the source respectively.
Return type dictionaries
Raises
• NodeNotFound – If source is not in G.
• NetworkXUnbounded – If the (di)graph contains a negative cost (di)cycle, the algorithm
raises an exception to indicate the presence of the negative cost (di)cycle. Note: any negative
weight edge in an undirected graph is a negative cost cycle.

Examples

>>> import networkx as nx


>>> G = nx.path_graph(5, create_using = nx.DiGraph())
>>> pred, dist = nx.goldberg_radzik(G, 0)
>>> sorted(pred.items())
[(0, None), (1, 0), (2, 1), (3, 2), (4, 3)]
>>> sorted(dist.items())
[(0, 0), (1, 1), (2, 2), (3, 3), (4, 4)]

>>> from nose.tools import assert_raises


>>> G = nx.cycle_graph(5, create_using = nx.DiGraph())
>>> G[1][2]['weight'] = -7
>>> assert_raises(nx.NetworkXUnbounded, nx.goldberg_radzik, G, 0)

432 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Edge weight attributes must be numerical. Distances are calculated as sums of weighted edges traversed.
The dictionaries returned only have keys for nodes reachable from the source.
In the case where the (di)graph is not connected, if a component not containing the source contains a negative
cost (di)cycle, it will not be detected.

networkx.algorithms.shortest_paths.weighted.johnson

johnson(G, weight=’weight’)
Uses Johnson’s Algorithm to compute shortest paths.
Johnson’s Algorithm finds a shortest path between each pair of nodes in a weighted graph even if negative
weights are present.
Parameters
• G (NetworkX graph)
• weight (string or function) – If this is a string, then edge weights will be accessed via
the edge attribute with this key (that is, the weight of the edge joining u to v will be G.
edges[u, v][weight]). If no such edge attribute exists, the weight of the edge is
assumed to be one.
If this is a function, the weight of an edge is the value returned by the function. The func-
tion must accept exactly three positional arguments: the two endpoints of an edge and the
dictionary of edge attributes for that edge. The function must return a number.
Returns distance – Dictionary, keyed by source and target, of shortest paths.
Return type dictionary
Raises NetworkXError – If given graph is not weighted.

Examples

>>> import networkx as nx


>>> graph = nx.DiGraph()
>>> graph.add_weighted_edges_from([('0', '3', 3), ('0', '1', -5),
... ('0', '2', 2), ('1', '2', 4), ('2', '3', 1)])
>>> paths = nx.johnson(graph, weight='weight')
>>> paths['0']['2']
['0', '1', '2']

Notes

Johnson’s algorithm is suitable even for graphs with negative weights. It works by using the Bellman–Ford
algorithm to compute a transformation of the input graph that removes all negative weights, allowing Dijkstra’s
algorithm to be used on the transformed graph.
The time complexity of this algorithm is 𝑂(𝑛2 log 𝑛 + 𝑛𝑚), where 𝑛 is the number of nodes and 𝑚 the number
of edges in the graph. For dense graphs, this may be faster than the Floyd–Warshall algorithm.
See also:

3.45. Shortest Paths 433


NetworkX Reference, Release 2.3rc1.dev20190222214247

floyd_warshall_predecessor_and_distance(), floyd_warshall_numpy(),
all_pairs_shortest_path(), all_pairs_shortest_path_length(),
all_pairs_dijkstra_path(), bellman_ford_predecessor_and_distance(),
all_pairs_bellman_ford_path(), all_pairs_bellman_ford_path_length()

3.45.7 Dense Graphs

Floyd-Warshall algorithm for shortest paths.

floyd_warshall(G[, weight]) Find all-pairs shortest path lengths using Floyd’s algo-
rithm.
floyd_warshall_predecessor_and_distance(G[, Find all-pairs shortest path lengths using Floyd’s algo-
. . . ]) rithm.
floyd_warshall_numpy(G[, nodelist, weight]) Find all-pairs shortest path lengths using Floyd’s algo-
rithm.
reconstruct_path(source, target, predecessors) Reconstruct a path from source to target us-
ing the predecessors dict as returned by
floyd_warshall_predecessor_and_distance

networkx.algorithms.shortest_paths.dense.floyd_warshall

floyd_warshall(G, weight=’weight’)
Find all-pairs shortest path lengths using Floyd’s algorithm.
Parameters
• G (NetworkX graph)
• weight (string, optional (default= ‘weight’)) – Edge data key corresponding to the edge
weight.
Returns distance – A dictionary, keyed by source and target, of shortest paths distances between
nodes.
Return type dict

Notes

Floyd’s algorithm is appropriate for finding shortest paths in dense graphs or graphs with negative weights when
Dijkstra’s algorithm fails. This algorithm can still fail if there are negative cycles. It has running time 𝑂(𝑛3 )
with running space of 𝑂(𝑛2 ).
See also:
floyd_warshall_predecessor_and_distance(), floyd_warshall_numpy(),
all_pairs_shortest_path(), all_pairs_shortest_path_length()

networkx.algorithms.shortest_paths.dense.floyd_warshall_predecessor_and_distance

floyd_warshall_predecessor_and_distance(G, weight=’weight’)
Find all-pairs shortest path lengths using Floyd’s algorithm.
Parameters
• G (NetworkX graph)

434 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• weight (string, optional (default= ‘weight’)) – Edge data key corresponding to the edge
weight.
Returns predecessor,distance – Dictionaries, keyed by source and target, of predecessors and dis-
tances in the shortest path.
Return type dictionaries

Examples

>>> G = nx.DiGraph()
>>> G.add_weighted_edges_from([('s', 'u', 10), ('s', 'x', 5),
... ('u', 'v', 1), ('u', 'x', 2), ('v', 'y', 1), ('x', 'u', 3),
... ('x', 'v', 5), ('x', 'y', 2), ('y', 's', 7), ('y', 'v', 6)])
>>> predecessors, _ = nx.floyd_warshall_predecessor_and_distance(G)
>>> print(nx.reconstruct_path('s', 'v', predecessors))
['s', 'x', 'u', 'v']

Notes

Floyd’s algorithm is appropriate for finding shortest paths in dense graphs or graphs with negative weights when
Dijkstra’s algorithm fails. This algorithm can still fail if there are negative cycles. It has running time 𝑂(𝑛3 )
with running space of 𝑂(𝑛2 ).
See also:
floyd_warshall(), floyd_warshall_numpy(), all_pairs_shortest_path(),
all_pairs_shortest_path_length()

networkx.algorithms.shortest_paths.dense.floyd_warshall_numpy

floyd_warshall_numpy(G, nodelist=None, weight=’weight’)


Find all-pairs shortest path lengths using Floyd’s algorithm.
Parameters
• G (NetworkX graph)
• nodelist (list, optional) – The rows and columns are ordered by the nodes in nodelist. If
nodelist is None then the ordering is produced by G.nodes().
• weight (string, optional (default= ‘weight’)) – Edge data key corresponding to the edge
weight.
Returns distance – A matrix of shortest path distances between nodes. If there is no path between
to nodes the corresponding matrix entry will be Inf.
Return type NumPy matrix

Notes

Floyd’s algorithm is appropriate for finding shortest paths in dense graphs or graphs with negative weights when
Dijkstra’s algorithm fails. This algorithm can still fail if there are negative cycles. It has running time 𝑂(𝑛3 )
with running space of 𝑂(𝑛2 ).

3.45. Shortest Paths 435


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.shortest_paths.dense.reconstruct_path

reconstruct_path(source, target, predecessors)


Reconstruct a path from source to target using the predecessors dict as returned by
floyd_warshall_predecessor_and_distance
Parameters
• source (node) – Starting node for path
• target (node) – Ending node for path
• predecessors (dictionary) – Dictionary, keyed by source and target, of predecessors in the
shortest path, as returned by floyd_warshall_predecessor_and_distance
Returns
path – A list of nodes containing the shortest path from source to target
If source and target are the same, an empty list is returned
Return type list

Notes

This function is meant to give more applicability to the floyd_warshall_predecessor_and_distance function


See also:
floyd_warshall_predecessor_and_distance()

3.45.8 A* Algorithm

Shortest paths and path lengths using the A* (“A star”) algorithm.

astar_path(G, source, target[, heuristic, . . . ]) Returns a list of nodes in a shortest path between source
and target using the A* (“A-star”) algorithm.
astar_path_length(G, source, target[, . . . ]) Returns the length of the shortest path between source
and target using the A* (“A-star”) algorithm.

networkx.algorithms.shortest_paths.astar.astar_path

astar_path(G, source, target, heuristic=None, weight=’weight’)


Returns a list of nodes in a shortest path between source and target using the A* (“A-star”) algorithm.
There may be more than one shortest path. This returns only one.
Parameters
• G (NetworkX graph)
• source (node) – Starting node for path
• target (node) – Ending node for path
• heuristic (function) – A function to evaluate the estimate of the distance from the a node to
the target. The function takes two nodes arguments and must return a number.

436 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• weight (string, optional (default=’weight’)) – Edge data key corresponding to the edge
weight.
Raises NetworkXNoPath – If no path exists between source and target.

Examples

>>> G = nx.path_graph(5)
>>> print(nx.astar_path(G, 0, 4))
[0, 1, 2, 3, 4]
>>> G = nx.grid_graph(dim=[3, 3]) # nodes are two-tuples (x,y)
>>> nx.set_edge_attributes(G, {e: e[1][0]*2 for e in G.edges()}, 'cost')
>>> def dist(a, b):
... (x1, y1) = a
... (x2, y2) = b
... return ((x1 - x2) ** 2 + (y1 - y2) ** 2) ** 0.5
>>> print(nx.astar_path(G, (0, 0), (2, 2), heuristic=dist, weight='cost'))
[(0, 0), (0, 1), (0, 2), (1, 2), (2, 2)]

See also:
shortest_path(), dijkstra_path()

networkx.algorithms.shortest_paths.astar.astar_path_length

astar_path_length(G, source, target, heuristic=None, weight=’weight’)


Returns the length of the shortest path between source and target using the A* (“A-star”) algorithm.
Parameters
• G (NetworkX graph)
• source (node) – Starting node for path
• target (node) – Ending node for path
• heuristic (function) – A function to evaluate the estimate of the distance from the a node to
the target. The function takes two nodes arguments and must return a number.
Raises NetworkXNoPath – If no path exists between source and target.
See also:
astar_path()

3.46 Similarity Measures

Functions measuring similarity using graph edit distance.


The graph edit distance is the number of edge/node changes needed to make two graphs isomorphic.
The default algorithm/implementation is sub-optimal for some graphs. The problem of finding the exact Graph Edit
Distance (GED) is NP-hard so it is often slow. If the simple interface graph_edit_distance takes too long for
your graph, try optimize_graph_edit_distance and/or optimize_edit_paths.
At the same time, I encourage capable people to investigate alternative GED algorithms, in order to improve the
choices available.

3.46. Similarity Measures 437


NetworkX Reference, Release 2.3rc1.dev20190222214247

graph_edit_distance(G1, G2[, node_match, Returns GED (graph edit distance) between graphs G1
. . . ]) and G2.
optimal_edit_paths(G1, G2[, node_match, . . . ]) Returns all minimum-cost edit paths transforming G1 to
G2.
optimize_graph_edit_distance(G1, G2[, Returns consecutive approximations of GED (graph edit
. . . ]) distance) between graphs G1 and G2.
optimize_edit_paths(G1, G2[, node_match, GED (graph edit distance) calculation: advanced inter-
. . . ]) face.

3.46.1 networkx.algorithms.similarity.graph_edit_distance

graph_edit_distance(G1, G2, node_match=None, edge_match=None, node_subst_cost=None,


node_del_cost=None, node_ins_cost=None, edge_subst_cost=None,
edge_del_cost=None, edge_ins_cost=None, upper_bound=None)
Returns GED (graph edit distance) between graphs G1 and G2.
Graph edit distance is a graph similarity measure analogous to Levenshtein distance for strings. It is defined
as minimum cost of edit path (sequence of node and edge edit operations) transforming graph G1 to graph
isomorphic to G2.
Parameters
• G1, G2 (graphs) – The two graphs G1 and G2 must be of the same type.
• node_match (callable) – A function that returns True if node n1 in G1 and n2 in G2 should
be considered equal during matching.
The function will be called like
node_match(G1.nodes[n1], G2.nodes[n2]).
That is, the function will receive the node attribute dictionaries for n1 and n2 as inputs.
Ignored if node_subst_cost is specified. If neither node_match nor node_subst_cost are
specified then node attributes are not considered.
• edge_match (callable) – A function that returns True if the edge attribute dictionaries for
the pair of nodes (u1, v1) in G1 and (u2, v2) in G2 should be considered equal during
matching.
The function will be called like
edge_match(G1[u1][v1], G2[u2][v2]).
That is, the function will receive the edge attribute dictionaries of the edges under consider-
ation.
Ignored if edge_subst_cost is specified. If neither edge_match nor edge_subst_cost are spec-
ified then edge attributes are not considered.
• node_subst_cost, node_del_cost, node_ins_cost (callable) – Functions that return the
costs of node substitution, node deletion, and node insertion, respectively.
The functions will be called like
node_subst_cost(G1.nodes[n1], G2.nodes[n2]), node_del_cost(G1.nodes[n1]),
node_ins_cost(G2.nodes[n2]).
That is, the functions will receive the node attribute dictionaries as inputs. The functions are
expected to return positive numeric values.

438 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Function node_subst_cost overrides node_match if specified. If neither node_match nor


node_subst_cost are specified then default node substitution cost of 0 is used (node attributes
are not considered during matching).
If node_del_cost is not specified then default node deletion cost of 1 is used. If
node_ins_cost is not specified then default node insertion cost of 1 is used.
• edge_subst_cost, edge_del_cost, edge_ins_cost (callable) – Functions that return the costs
of edge substitution, edge deletion, and edge insertion, respectively.
The functions will be called like
edge_subst_cost(G1[u1][v1], G2[u2][v2]), edge_del_cost(G1[u1][v1]),
edge_ins_cost(G2[u2][v2]).
That is, the functions will receive the edge attribute dictionaries as inputs. The functions are
expected to return positive numeric values.
Function edge_subst_cost overrides edge_match if specified. If neither edge_match nor
edge_subst_cost are specified then default edge substitution cost of 0 is used (edge attributes
are not considered during matching).
If edge_del_cost is not specified then default edge deletion cost of 1 is used. If edge_ins_cost
is not specified then default edge insertion cost of 1 is used.
• upper_bound (numeric) – Maximum edit distance to consider. Return None if no edit
distance under or equal to upper_bound exists.

Examples

>>> G1 = nx.cycle_graph(6)
>>> G2 = nx.wheel_graph(7)
>>> nx.graph_edit_distance(G1, G2)
7.0

See also:
optimal_edit_paths(), optimize_graph_edit_distance(), is_isomorphic()

References

3.46.2 networkx.algorithms.similarity.optimal_edit_paths

optimal_edit_paths(G1, G2, node_match=None, edge_match=None, node_subst_cost=None,


node_del_cost=None, node_ins_cost=None, edge_subst_cost=None,
edge_del_cost=None, edge_ins_cost=None, upper_bound=None)
Returns all minimum-cost edit paths transforming G1 to G2.
Graph edit path is a sequence of node and edge edit operations transforming graph G1 to graph isomorphic to
G2. Edit operations include substitutions, deletions, and insertions.
Parameters
• G1, G2 (graphs) – The two graphs G1 and G2 must be of the same type.
• node_match (callable) – A function that returns True if node n1 in G1 and n2 in G2 should
be considered equal during matching.
The function will be called like

3.46. Similarity Measures 439


NetworkX Reference, Release 2.3rc1.dev20190222214247

node_match(G1.nodes[n1], G2.nodes[n2]).
That is, the function will receive the node attribute dictionaries for n1 and n2 as inputs.
Ignored if node_subst_cost is specified. If neither node_match nor node_subst_cost are
specified then node attributes are not considered.
• edge_match (callable) – A function that returns True if the edge attribute dictionaries for
the pair of nodes (u1, v1) in G1 and (u2, v2) in G2 should be considered equal during
matching.
The function will be called like
edge_match(G1[u1][v1], G2[u2][v2]).
That is, the function will receive the edge attribute dictionaries of the edges under consider-
ation.
Ignored if edge_subst_cost is specified. If neither edge_match nor edge_subst_cost are spec-
ified then edge attributes are not considered.
• node_subst_cost, node_del_cost, node_ins_cost (callable) – Functions that return the
costs of node substitution, node deletion, and node insertion, respectively.
The functions will be called like
node_subst_cost(G1.nodes[n1], G2.nodes[n2]), node_del_cost(G1.nodes[n1]),
node_ins_cost(G2.nodes[n2]).
That is, the functions will receive the node attribute dictionaries as inputs. The functions are
expected to return positive numeric values.
Function node_subst_cost overrides node_match if specified. If neither node_match nor
node_subst_cost are specified then default node substitution cost of 0 is used (node attributes
are not considered during matching).
If node_del_cost is not specified then default node deletion cost of 1 is used. If
node_ins_cost is not specified then default node insertion cost of 1 is used.
• edge_subst_cost, edge_del_cost, edge_ins_cost (callable) – Functions that return the costs
of edge substitution, edge deletion, and edge insertion, respectively.
The functions will be called like
edge_subst_cost(G1[u1][v1], G2[u2][v2]), edge_del_cost(G1[u1][v1]),
edge_ins_cost(G2[u2][v2]).
That is, the functions will receive the edge attribute dictionaries as inputs. The functions are
expected to return positive numeric values.
Function edge_subst_cost overrides edge_match if specified. If neither edge_match nor
edge_subst_cost are specified then default edge substitution cost of 0 is used (edge attributes
are not considered during matching).
If edge_del_cost is not specified then default edge deletion cost of 1 is used. If edge_ins_cost
is not specified then default edge insertion cost of 1 is used.
• upper_bound (numeric) – Maximum edit distance to consider.
Returns
• edit_paths (list of tuples (node_edit_path, edge_edit_path)) – node_edit_path : list of tuples
(u, v) edge_edit_path : list of tuples ((u1, v1), (u2, v2))
• cost (numeric) – Optimal edit path cost (graph edit distance).

440 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G1 = nx.cycle_graph(6)
>>> G2 = nx.wheel_graph(7)
>>> paths, cost = nx.optimal_edit_paths(G1, G2)
>>> len(paths)
84
>>> cost
7.0

See also:
graph_edit_distance(), optimize_edit_paths()

References

3.46.3 networkx.algorithms.similarity.optimize_graph_edit_distance

optimize_graph_edit_distance(G1, G2, node_match=None, edge_match=None,


node_subst_cost=None, node_del_cost=None,
node_ins_cost=None, edge_subst_cost=None,
edge_del_cost=None, edge_ins_cost=None, upper_bound=None)
Returns consecutive approximations of GED (graph edit distance) between graphs G1 and G2.
Graph edit distance is a graph similarity measure analogous to Levenshtein distance for strings. It is defined
as minimum cost of edit path (sequence of node and edge edit operations) transforming graph G1 to graph
isomorphic to G2.
Parameters
• G1, G2 (graphs) – The two graphs G1 and G2 must be of the same type.
• node_match (callable) – A function that returns True if node n1 in G1 and n2 in G2 should
be considered equal during matching.
The function will be called like
node_match(G1.nodes[n1], G2.nodes[n2]).
That is, the function will receive the node attribute dictionaries for n1 and n2 as inputs.
Ignored if node_subst_cost is specified. If neither node_match nor node_subst_cost are
specified then node attributes are not considered.
• edge_match (callable) – A function that returns True if the edge attribute dictionaries for
the pair of nodes (u1, v1) in G1 and (u2, v2) in G2 should be considered equal during
matching.
The function will be called like
edge_match(G1[u1][v1], G2[u2][v2]).
That is, the function will receive the edge attribute dictionaries of the edges under consider-
ation.
Ignored if edge_subst_cost is specified. If neither edge_match nor edge_subst_cost are spec-
ified then edge attributes are not considered.
• node_subst_cost, node_del_cost, node_ins_cost (callable) – Functions that return the
costs of node substitution, node deletion, and node insertion, respectively.

3.46. Similarity Measures 441


NetworkX Reference, Release 2.3rc1.dev20190222214247

The functions will be called like


node_subst_cost(G1.nodes[n1], G2.nodes[n2]), node_del_cost(G1.nodes[n1]),
node_ins_cost(G2.nodes[n2]).
That is, the functions will receive the node attribute dictionaries as inputs. The functions are
expected to return positive numeric values.
Function node_subst_cost overrides node_match if specified. If neither node_match nor
node_subst_cost are specified then default node substitution cost of 0 is used (node attributes
are not considered during matching).
If node_del_cost is not specified then default node deletion cost of 1 is used. If
node_ins_cost is not specified then default node insertion cost of 1 is used.
• edge_subst_cost, edge_del_cost, edge_ins_cost (callable) – Functions that return the costs
of edge substitution, edge deletion, and edge insertion, respectively.
The functions will be called like
edge_subst_cost(G1[u1][v1], G2[u2][v2]), edge_del_cost(G1[u1][v1]),
edge_ins_cost(G2[u2][v2]).
That is, the functions will receive the edge attribute dictionaries as inputs. The functions are
expected to return positive numeric values.
Function edge_subst_cost overrides edge_match if specified. If neither edge_match nor
edge_subst_cost are specified then default edge substitution cost of 0 is used (edge attributes
are not considered during matching).
If edge_del_cost is not specified then default edge deletion cost of 1 is used. If edge_ins_cost
is not specified then default edge insertion cost of 1 is used.
• upper_bound (numeric) – Maximum edit distance to consider.
Returns
Return type Generator of consecutive approximations of graph edit distance.

Examples

>>> G1 = nx.cycle_graph(6)
>>> G2 = nx.wheel_graph(7)
>>> for v in nx.optimize_graph_edit_distance(G1, G2):
... minv = v
>>> minv
7.0

See also:
graph_edit_distance(), optimize_edit_paths()

442 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.46.4 networkx.algorithms.similarity.optimize_edit_paths

optimize_edit_paths(G1, G2, node_match=None, edge_match=None, node_subst_cost=None,


node_del_cost=None, node_ins_cost=None, edge_subst_cost=None,
edge_del_cost=None, edge_ins_cost=None, upper_bound=None,
strictly_decreasing=True)
GED (graph edit distance) calculation: advanced interface.
Graph edit path is a sequence of node and edge edit operations transforming graph G1 to graph isomorphic to
G2. Edit operations include substitutions, deletions, and insertions.
Graph edit distance is defined as minimum cost of edit path.
Parameters
• G1, G2 (graphs) – The two graphs G1 and G2 must be of the same type.
• node_match (callable) – A function that returns True if node n1 in G1 and n2 in G2 should
be considered equal during matching.
The function will be called like
node_match(G1.nodes[n1], G2.nodes[n2]).
That is, the function will receive the node attribute dictionaries for n1 and n2 as inputs.
Ignored if node_subst_cost is specified. If neither node_match nor node_subst_cost are
specified then node attributes are not considered.
• edge_match (callable) – A function that returns True if the edge attribute dictionaries for
the pair of nodes (u1, v1) in G1 and (u2, v2) in G2 should be considered equal during
matching.
The function will be called like
edge_match(G1[u1][v1], G2[u2][v2]).
That is, the function will receive the edge attribute dictionaries of the edges under consider-
ation.
Ignored if edge_subst_cost is specified. If neither edge_match nor edge_subst_cost are spec-
ified then edge attributes are not considered.
• node_subst_cost, node_del_cost, node_ins_cost (callable) – Functions that return the
costs of node substitution, node deletion, and node insertion, respectively.
The functions will be called like
node_subst_cost(G1.nodes[n1], G2.nodes[n2]), node_del_cost(G1.nodes[n1]),
node_ins_cost(G2.nodes[n2]).
That is, the functions will receive the node attribute dictionaries as inputs. The functions are
expected to return positive numeric values.
Function node_subst_cost overrides node_match if specified. If neither node_match nor
node_subst_cost are specified then default node substitution cost of 0 is used (node attributes
are not considered during matching).
If node_del_cost is not specified then default node deletion cost of 1 is used. If
node_ins_cost is not specified then default node insertion cost of 1 is used.

3.46. Similarity Measures 443


NetworkX Reference, Release 2.3rc1.dev20190222214247

• edge_subst_cost, edge_del_cost, edge_ins_cost (callable) – Functions that return the costs


of edge substitution, edge deletion, and edge insertion, respectively.
The functions will be called like
edge_subst_cost(G1[u1][v1], G2[u2][v2]), edge_del_cost(G1[u1][v1]),
edge_ins_cost(G2[u2][v2]).
That is, the functions will receive the edge attribute dictionaries as inputs. The functions are
expected to return positive numeric values.
Function edge_subst_cost overrides edge_match if specified. If neither edge_match nor
edge_subst_cost are specified then default edge substitution cost of 0 is used (edge attributes
are not considered during matching).
If edge_del_cost is not specified then default edge deletion cost of 1 is used. If edge_ins_cost
is not specified then default edge insertion cost of 1 is used.
• upper_bound (numeric) – Maximum edit distance to consider.
• strictly_decreasing (bool) – If True, return consecutive approximations of strictly decreas-
ing cost. Otherwise, return all edit paths of cost less than or equal to the previous minimum
cost.
Returns node_edit_path : list of tuples (u, v) edge_edit_path : list of tuples ((u1, v1), (u2, v2)) cost
: numeric
Return type Generator of tuples (node_edit_path, edge_edit_path, cost)
See also:
graph_edit_distance(), optimize_graph_edit_distance(), optimal_edit_paths()

References

3.47 Simple Paths

all_simple_paths(G, source, target[, cutoff]) Generate all simple paths in the graph G from source to
target.
is_simple_path(G, nodes) Returns True if and only if the given nodes form a sim-
ple path in G.
shortest_simple_paths(G, source, target[, . . . ]) Generate all simple paths in the graph G from source to
target,

3.47.1 networkx.algorithms.simple_paths.all_simple_paths

all_simple_paths(G, source, target, cutoff=None)


Generate all simple paths in the graph G from source to target.
A simple path is a path with no repeated nodes.
Parameters
• G (NetworkX graph)
• source (node) – Starting node for path
• target (nodes) – Single node or iterable of nodes at which to end path

444 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• cutoff (integer, optional) – Depth to stop the search. Only paths of length <= cutoff are
returned.
Returns path_generator – A generator that produces lists of simple paths. If there are no paths
between the source and target within the given cutoff the generator produces no output.
Return type generator

Examples

This iterator generates lists of nodes:

>>> G = nx.complete_graph(4)
>>> for path in nx.all_simple_paths(G, source=0, target=3):
... print(path)
...
[0, 1, 2, 3]
[0, 1, 3]
[0, 2, 1, 3]
[0, 2, 3]
[0, 3]

You can generate only those paths that are shorter than a certain length by using the cutoff keyword argument:

>>> paths = nx.all_simple_paths(G, source=0, target=3, cutoff=2)


>>> print(list(paths))
[[0, 1, 3], [0, 2, 3], [0, 3]]

To get each path as the corresponding list of edges, you can use the networkx.utils.pairwise() helper
function:

>>> paths = nx.all_simple_paths(G, source=0, target=3)


>>> for path in map(nx.utils.pairwise, paths):
... print(list(path))
[(0, 1), (1, 2), (2, 3)]
[(0, 1), (1, 3)]
[(0, 2), (2, 1), (1, 3)]
[(0, 2), (2, 3)]
[(0, 3)]

Pass an iterable of nodes as target to generate all paths ending in any of several nodes:

>>> G = nx.complete_graph(4)
>>> for path in nx.all_simple_paths(G, source=0, target=[3, 2]):
... print(path)
...
[0, 1, 2]
[0, 1, 2, 3]
[0, 1, 3]
[0, 1, 3, 2]
[0, 2]
[0, 2, 1, 3]
[0, 2, 3]
[0, 3]
[0, 3, 1, 2]
[0, 3, 2]

3.47. Simple Paths 445


NetworkX Reference, Release 2.3rc1.dev20190222214247

Iterate over each path from the root nodes to the leaf nodes in a directed acyclic graph using a functional
programming approach:

>>> from itertools import chain


>>> from itertools import product
>>> from itertools import starmap
>>> from functools import partial
>>>
>>> chaini = chain.from_iterable
>>>
>>> G = nx.DiGraph([(0, 1), (1, 2), (0, 3), (3, 2)])
>>> roots = (v for v, d in G.in_degree() if d == 0)
>>> leaves = (v for v, d in G.out_degree() if d == 0)
>>> all_paths = partial(nx.all_simple_paths, G)
>>> list(chaini(starmap(all_paths, product(roots, leaves))))
[[0, 1, 2], [0, 3, 2]]

The same list computed using an iterative approach:

>>> G = nx.DiGraph([(0, 1), (1, 2), (0, 3), (3, 2)])


>>> roots = (v for v, d in G.in_degree() if d == 0)
>>> leaves = (v for v, d in G.out_degree() if d == 0)
>>> all_paths = []
>>> for root in roots:
... for leaf in leaves:
... paths = nx.all_simple_paths(G, root, leaf)
... all_paths.extend(paths)
>>> all_paths
[[0, 1, 2], [0, 3, 2]]

Iterate over each path from the root nodes to the leaf nodes in a directed acyclic graph passing all leaves together
to avoid unnecessary compute:

>>> G = nx.DiGraph([(0, 1), (2, 1), (1, 3), (1, 4)])


>>> roots = (v for v, d in G.in_degree() if d == 0)
>>> leaves = [v for v, d in G.out_degree() if d == 0]
>>> all_paths = []
>>> for root in roots:
... paths = nx.all_simple_paths(G, root, leaves)
... all_paths.extend(paths)
>>> all_paths
[[0, 1, 3], [0, 1, 4], [2, 1, 3], [2, 1, 4]]

Notes

This algorithm uses a modified depth-first search to generate the paths1 . A single path can be found in 𝑂(𝑉 +𝐸)
time but the number of simple paths in a graph can be very large, e.g. 𝑂(𝑛!) in the complete graph of order 𝑛.

References

See also:
all_shortest_paths(), shortest_path()
1 R. Sedgewick, “Algorithms in C, Part 5: Graph Algorithms”, Addison Wesley Professional, 3rd ed., 2001.

446 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.47.2 networkx.algorithms.simple_paths.is_simple_path

is_simple_path(G, nodes)
Returns True if and only if the given nodes form a simple path in G.
A simple path in a graph is a nonempty sequence of nodes in which no node appears more than once in the
sequence, and each adjacent pair of nodes in the sequence is adjacent in the graph.
Parameters nodes (list) – A list of one or more nodes in the graph G.
Returns Whether the given list of nodes represents a simple path in G.
Return type bool

Notes

A list of zero nodes is not a path and a list of one node is a path. Here’s an explanation why.
This function operates on node paths. One could also consider edge paths. There is a bijection between node
paths and edge paths.
The length of a path is the number of edges in the path, so a list of nodes of length n corresponds to a path of
length n - 1. Thus the smallest edge path would be a list of zero edges, the empty path. This corresponds to a
list of one node.
To convert between a node path and an edge path, you can use code like the following:

>>> from networkx.utils import pairwise


>>> nodes = [0, 1, 2, 3]
>>> edges = list(pairwise(nodes))
>>> edges
[(0, 1), (1, 2), (2, 3)]
>>> nodes = [edges[0][0]] + [v for u, v in edges]
>>> nodes
[0, 1, 2, 3]

Examples

>>> G = nx.cycle_graph(4)
>>> nx.is_simple_path(G, [2, 3, 0])
True
>>> nx.is_simple_path(G, [0, 2])
False

3.47.3 networkx.algorithms.simple_paths.shortest_simple_paths

shortest_simple_paths(G, source, target, weight=None)


Generate all simple paths in the graph G from source to target, starting from shortest ones.
A simple path is a path with no repeated nodes.
If a weighted shortest path search is to be used, no negative weights are allowed.
Parameters
• G (NetworkX graph)

3.47. Simple Paths 447


NetworkX Reference, Release 2.3rc1.dev20190222214247

• source (node) – Starting node for path


• target (node) – Ending node for path
• weight (string) – Name of the edge attribute to be used as a weight. If None all edges are
considered to have unit weight. Default value None.
Returns path_generator – A generator that produces lists of simple paths, in order from shortest to
longest.
Return type generator
Raises
• NetworkXNoPath – If no path exists between source and target.
• NetworkXError – If source or target nodes are not in the input graph.
• NetworkXNotImplemented – If the input graph is a Multi[Di]Graph.

Examples

>>> G = nx.cycle_graph(7)
>>> paths = list(nx.shortest_simple_paths(G, 0, 3))
>>> print(paths)
[[0, 1, 2, 3], [0, 6, 5, 4, 3]]

You can use this function to efficiently compute the k shortest/best paths between two nodes.

>>> from itertools import islice


>>> def k_shortest_paths(G, source, target, k, weight=None):
... return list(islice(nx.shortest_simple_paths(G, source, target,
˓→weight=weight), k))

>>> for path in k_shortest_paths(G, 0, 3, 2):


... print(path)
[0, 1, 2, 3]
[0, 6, 5, 4, 3]

Notes

This procedure is based on algorithm by Jin Y. Yen1 . Finding the first 𝐾 paths requires 𝑂(𝐾𝑁 3 ) operations.
See also:
all_shortest_paths(), shortest_path(), all_simple_paths()

References

3.48 Similarity Measures

Functions for estimating the small-world-ness of graphs.


A small world network is characterized by a small average shortest path length, and a large clustering coefficient.
Small-worldness is commonly measured with the coefficient sigma or omega.
1 Jin Y. Yen, “Finding the K Shortest Loopless Paths in a Network”, Management Science, Vol. 17, No. 11, Theory Series (Jul., 1971), pp.

712-716.

448 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Both coefficients compare the average clustering coefficient and shortest path length of a given graph against the same
quantities for an equivalent random or lattice graph.
For more information, see the Wikipedia article on small-world network1 .

random_reference(G[, niter, connectivity, seed]) Compute a random graph by swapping edges of a given
graph.
lattice_reference(G[, niter, D, . . . ]) Latticize the given graph by swapping edges.
sigma(G[, niter, nrand, seed]) Returns the small-world coefficient (sigma) of the given
graph.
omega(G[, niter, nrand, seed]) Returns the small-world coefficient (omega) of a graph

3.48.1 networkx.algorithms.smallworld.random_reference

random_reference(G, niter=1, connectivity=True, seed=None)


Compute a random graph by swapping edges of a given graph.
Parameters
• G (graph) – An undirected graph with 4 or more nodes.
• niter (integer (optional, default=1)) – An edge is rewired approximately niter times.
• connectivity (boolean (optional, default=True)) – When True, ensure connectivity for the
randomized graph.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G – The randomized graph.
Return type graph

Notes

The implementation is adapted from the algorithm by Maslov and Sneppen (2002)1 .

References

3.48.2 networkx.algorithms.smallworld.lattice_reference

lattice_reference(G, niter=1, D=None, connectivity=True, seed=None)


Latticize the given graph by swapping edges.
Parameters
• G (graph) – An undirected graph with 4 or more nodes.
• niter (integer (optional, default=1)) – An edge is rewired approximatively niter times.
• D (numpy.array (optional, default=None)) – Distance to the diagonal matrix.
• connectivity (boolean (optional, default=True)) – Ensure connectivity for the latticized
graph when set to True.
1 Small-world network:: https://en.wikipedia.org/wiki/Small-world_network
1 Maslov, Sergei, and Kim Sneppen. “Specificity and stability in topology of protein networks.” Science 296.5569 (2002): 910-913.

3.48. Similarity Measures 449


NetworkX Reference, Release 2.3rc1.dev20190222214247

• seed (integer, random_state, or None (default)) – Indicator of random number generation


state. See Randomness.
Returns G – The latticized graph.
Return type graph

Notes

The implementation is adapted from the algorithm by Sporns et al.1 . which is inspired from the original work
by Maslov and Sneppen(2002)2 .

References

3.48.3 networkx.algorithms.smallworld.sigma

sigma(G, niter=100, nrand=10, seed=None)


Returns the small-world coefficient (sigma) of the given graph.
The small-world coefficient is defined as: sigma = C/Cr / L/Lr where C and L are respectively the average
clustering coefficient and average shortest path length of G. Cr and Lr are respectively the average clustering
coefficient and average shortest path length of an equivalent random graph.
A graph is commonly classified as small-world if sigma>1.
Parameters
• G (NetworkX graph) – An undirected graph.
• niter (integer (optional, default=100)) – Approximate number of rewiring per edge to com-
pute the equivalent random graph.
• nrand (integer (optional, default=10)) – Number of random graphs generated to compute
the average clustering coefficient (Cr) and average shortest path length (Lr).
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns sigma – The small-world coefficient of G.
Return type float

Notes

The implementation is adapted from Humphries et al.12 .


1 Sporns, Olaf, and Jonathan D. Zwi. “The small world of the cerebral cortex.” Neuroinformatics 2.2 (2004): 145-162.
2 Maslov, Sergei, and Kim Sneppen. “Specificity and stability in topology of protein networks.” Science 296.5569 (2002): 910-913.
1 The brainstem reticular formation is a small-world, not scale-free, network M. D. Humphries, K. Gurney and T. J. Prescott, Proc. Roy. Soc. B

2006 273, 503-511, doi:10.1098/rspb.2005.3354.


2 Humphries and Gurney (2008). “Network ‘Small-World-Ness’: A Quantitative Method for Determining Canonical Network Equivalence”.

PLoS One. 3 (4). PMID 18446219. doi:10.1371/journal.pone.0002051.

450 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.48.4 networkx.algorithms.smallworld.omega

omega(G, niter=100, nrand=10, seed=None)


Returns the small-world coefficient (omega) of a graph
The small-world coefficient of a graph G is:
omega = Lr/L - C/Cl
where C and L are respectively the average clustering coefficient and average shortest path length of G. Lr is
the average shortest path length of an equivalent random graph and Cl is the average clustering coefficient of an
equivalent lattice graph.
The small-world coefficient (omega) ranges between -1 and 1. Values close to 0 means the G features small-
world characteristics. Values close to -1 means G has a lattice shape whereas values close to 1 means G is a
random graph.
Parameters
• G (NetworkX graph) – An undirected graph.
• niter (integer (optional, default=100)) – Approximate number of rewiring per edge to com-
pute the equivalent random graph.
• nrand (integer (optional, default=10)) – Number of random graphs generated to compute
the average clustering coefficient (Cr) and average shortest path length (Lr).
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns omega – The small-work coefficient (omega)
Return type float

Notes

The implementation is adapted from the algorithm by Telesford et al.1 .

References

3.49 s metric

s_metric(G[, normalized]) Returns the s-metric of graph.

3.49.1 networkx.algorithms.smetric.s_metric

s_metric(G, normalized=True)
Returns the s-metric of graph.
The s-metric is defined as the sum of the products deg(u)*deg(v) for every edge (u,v) in G. If norm is provided
construct the s-max graph and compute it’s s_metric, and return the normalized s value
Parameters
1
Telesford, Joyce, Hayasaka, Burdette, and Laurienti (2011). “The Ubiquity of Small-World Networks”. Brain Connectivity. 1 (0038): 367-75.
PMC 3604768. PMID 22432451. doi:10.1089/brain.2011.0038.

3.49. s metric 451


NetworkX Reference, Release 2.3rc1.dev20190222214247

• G (graph) – The graph used to compute the s-metric.


• normalized (bool (optional)) – Normalize the value.
Returns s – The s-metric of the graph.
Return type float

References

3.50 Sparsifiers

Functions for computing sparsifiers of graphs.

spanner(G, stretch[, weight, seed]) Returns a spanner of the given graph with the given
stretch.

3.50.1 networkx.algorithms.sparsifiers.spanner

spanner(G, stretch, weight=None, seed=None)


Returns a spanner of the given graph with the given stretch.
A spanner of a graph G = (V, E) with stretch t is a subgraph H = (V, E_S) such that E_S is a subset of E and the
distance between any pair of nodes in H is at most t times the distance between the nodes in G.
Parameters
• G (NetworkX graph) – An undirected simple graph.
• stretch (float) – The stretch of the spanner.
• weight (object) – The edge attribute to use as distance.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns A spanner of the given graph with the given stretch.
Return type NetworkX graph
Raises ValueError – If a stretch less than 1 is given.

Notes

This function implements the spanner algorithm by Baswana and Sen, see [1].
This algorithm is a randomized las vegas algorithm: The expected running time is O(km) where k = (stretch +
1) // 2 and m is the number of edges in G. The returned graph is always a spanner of the given graph with the
specified stretch. For weighted graphs the number of edges in the spanner is O(k * n^(1 + 1 / k)) where k is
defined as above and n is the number of nodes in G. For unweighted graphs the number of edges is O(n^(1 + 1 /
k) + kn).

References

[1] S. Baswana, S. Sen. A Simple and Linear Time Randomized Algorithm for Computing Sparse Spanners in
Weighted Graphs. Random Struct. Algorithms 30(4): 532-563 (2007).

452 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.51 Structural holes

Functions for computing measures of structural holes.

constraint(G[, nodes, weight]) Returns the constraint on all nodes in the graph G.
effective_size(G[, nodes, weight]) Returns the effective size of all nodes in the graph G.
local_constraint(G, u, v[, weight]) Returns the local constraint on the node u with respect
to the node v in the graph G.

3.51.1 networkx.algorithms.structuralholes.constraint

constraint(G, nodes=None, weight=None)


Returns the constraint on all nodes in the graph G.
The constraint is a measure of the extent to which a node v is invested in those nodes that are themselves invested
in the neighbors of v. Formally, the constraint on v, denoted c(v), is defined by
∑︁
𝑐(𝑣) = ℓ(𝑣, 𝑤)
𝑤∈𝑁 (𝑣)∖{𝑣}

where N(v) is the subset of the neighbors of v that are either predecessors or successors of v and
ell(v, w) is the local constraint on v with respect to w1 . For the definition of local constraint, see
local_constraint().
Parameters
• G (NetworkX graph) – The graph containing v. This can be either directed or undirected.
• nodes (container, optional) – Container of nodes in the graph G to compute the constraint.
If None, the constraint of every node is computed.
• weight (None or string, optional) – If None, all edge weights are considered equal. Other-
wise holds the name of the edge attribute used as weight.
Returns Dictionary with nodes as keys and the constraint on the node as values.
Return type dict
See also:
local_constraint()

References

3.51.2 networkx.algorithms.structuralholes.effective_size

effective_size(G, nodes=None, weight=None)


Returns the effective size of all nodes in the graph G.
The effective size of a node’s ego network is based on the concept of redundancy. A person’s ego network
has redundancy to the extent that her contacts are connected to each other as well. The nonredundant part of a
person’s relationships it’s the effective size of her ego network1 . Formally, the effective size of a node 𝑢, denoted
1 Burt, Ronald S. “Structural holes and good ideas”. American Journal of Sociology (110): 349–399.
1 Burt, Ronald S. Structural Holes: The Social Structure of Competition. Cambridge: Harvard University Press, 1995.

3.51. Structural holes 453


NetworkX Reference, Release 2.3rc1.dev20190222214247

𝑒(𝑢), is defined by
⎛ ⎞
∑︁ ∑︁
𝑒(𝑢) = ⎝1 − 𝑝𝑢𝑤 𝑚𝑣𝑤 ⎠
𝑣∈𝑁 (𝑢)∖{𝑢} 𝑤∈𝑁 (𝑣)

where 𝑁 (𝑢) is the set of neighbors of 𝑢 and 𝑝𝑢𝑤 is the normalized mutual weight of the (directed or undirected)
edges joining 𝑢 and 𝑣, for each vertex 𝑢 and 𝑣 1 . And 𝑚𝑣𝑤 is the mutual weight of 𝑣 and 𝑤 divided by 𝑣 highest
mutual weight with any of its neighbors. The mutual weight of 𝑢 and 𝑣 is the sum of the weights of edges joining
them (edge weights are assumed to be one if the graph is unweighted).
For the case of unweighted and undirected graphs, Borgatti proposed a simplified formula to compute effective
size2
2𝑡
𝑒(𝑢) = 𝑛 −
𝑛
where t is the number of ties in the ego network (not including ties to ego) and n is the number of nodes
(excluding ego).
Parameters
• G (NetworkX graph) – The graph containing v. Directed graphs are treated like undirected
graphs when computing neighbors of v.
• nodes (container, optional) – Container of nodes in the graph G to compute the effective
size. If None, the effective size of every node is computed.
• weight (None or string, optional) – If None, all edge weights are considered equal. Other-
wise holds the name of the edge attribute used as weight.
Returns Dictionary with nodes as keys and the constraint on the node as values.
Return type dict

Notes

Burt also defined the related concept of efficiency of a node’s ego network, which is its effective size divided by
the degree of that node1 . So you can easily compute efficiency:

>>> G = nx.DiGraph()
>>> G.add_edges_from([(0, 1), (0, 2), (1, 0), (2, 1)])
>>> esize = nx.effective_size(G)
>>> efficiency = {n: v / G.degree(n) for n, v in esize.items()}

See also:
constraint()

References

3.51.3 networkx.algorithms.structuralholes.local_constraint

local_constraint(G, u, v, weight=None)
Returns the local constraint on the node u with respect to the node v in the graph G.
2 Borgatti, S. “Structural Holes: Unpacking Burt’s Redundancy Measures” CONNECTIONS 20(1):35-38. http://www.analytictech.com/

connections/v20(1)/holes.htm

454 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Formally, the local constraint on u with respect to v, denoted ℓ(𝑣), is defined by


⎛ ⎞2
∑︁
ℓ(𝑢, 𝑣) = ⎝𝑝𝑢𝑣 + 𝑝𝑢𝑤 𝑝𝑤𝑣 ⎠ ,
𝑤∈𝑁 (𝑣)

where 𝑁 (𝑣) is the set of neighbors of 𝑣 and 𝑝𝑢𝑣 is the normalized mutual weight of the (directed or undirected)
edges joining 𝑢 and 𝑣, for each vertex 𝑢 and 𝑣 1 . The mutual weight of 𝑢 and 𝑣 is the sum of the weights of edges
joining them (edge weights are assumed to be one if the graph is unweighted).
Parameters
• G (NetworkX graph) – The graph containing u and v. This can be either directed or undi-
rected.
• u (node) – A node in the graph G.
• v (node) – A node in the graph G.
• weight (None or string, optional) – If None, all edge weights are considered equal. Other-
wise holds the name of the edge attribute used as weight.
Returns The constraint of the node v in the graph G.
Return type float
See also:
constraint()

References

3.52 Swap

Swap edges in a graph.

double_edge_swap(G[, nswap, max_tries, seed]) Swap two edges in the graph while keeping the node
degrees fixed.
connected_double_edge_swap(G[, nswap, Attempts the specified number of double-edge swaps in
. . . ]) the graph G.

3.52.1 networkx.algorithms.swap.double_edge_swap

double_edge_swap(G, nswap=1, max_tries=100, seed=None)


Swap two edges in the graph while keeping the node degrees fixed.
A double-edge swap removes two randomly chosen edges u-v and x-y and creates the new edges u-x and v-y:

u--v u v
becomes | |
x--y x y

If either the edge u-x or v-y already exist no swap is performed and another attempt is made to find a suitable
edge pair.
Parameters
1 Burt, Ronald S. “Structural holes and good ideas”. American Journal of Sociology (110): 349–399.

3.52. Swap 455


NetworkX Reference, Release 2.3rc1.dev20190222214247

• G (graph) – An undirected graph


• nswap (integer (optional, default=1)) – Number of double-edge swaps to perform
• max_tries (integer (optional)) – Maximum number of attempts to swap edges
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G – The graph after double edge swaps.
Return type graph

Notes

Does not enforce any connectivity constraints.


The graph G is modified in place.

3.52.2 networkx.algorithms.swap.connected_double_edge_swap

connected_double_edge_swap(G, nswap=1, _window_threshold=3, seed=None)


Attempts the specified number of double-edge swaps in the graph G.
A double-edge swap removes two randomly chosen edges (u, v) and (x, y) and creates the new edges
(u, x) and (v, y):

u--v u v
becomes | |
x--y x y

If either (u, x) or (v, y) already exist, then no swap is performed so the actual number of swapped edges
is always at most nswap.
Parameters
• G (graph) – An undirected graph
• nswap (integer (optional, default=1)) – Number of double-edge swaps to perform
• _window_threshold (integer) – The window size below which connectedness of the graph
will be checked after each swap.
The “window” in this function is a dynamically updated integer that represents the number of
swap attempts to make before checking if the graph remains connected. It is an optimization
used to decrease the running time of the algorithm in exchange for increased complexity of
implementation.
If the window size is below this threshold, then the algorithm checks after each swap if the
graph remains connected by checking if there is a path joining the two nodes whose edge
was just removed. If the window size is above this threshold, then the algorithm performs
do all the swaps in the window and only then check if the graph is still connected.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns The number of successful swaps
Return type int

456 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Raises NetworkXError – If the input graph is not connected, or if the graph has fewer than four
nodes.

Notes

The initial graph G must be connected, and the resulting graph is connected. The graph G is modified in place.

References

3.53 Tournament

Functions concerning tournament graphs.


A tournament graph is a complete oriented graph. In other words, it is a directed graph in which there is exactly one
directed edge joining each pair of distinct nodes. For each function in this module that accepts a graph as input, you
must provide a tournament graph. The responsibility is on the caller to ensure that the graph is a tournament graph.
To access the functions in this module, you must access them through the networkx.algorithms.tournament
module:

>>> import networkx as nx


>>> from networkx.algorithms import tournament
>>> G = nx.DiGraph([(0, 1), (1, 2), (2, 0)])
>>> tournament.is_tournament(G)
True

hamiltonian_path(G) Returns a Hamiltonian path in the given tournament


graph.
is_reachable(G, s, t) Decides whether there is a path from s to t in the tour-
nament.
is_strongly_connected(G) Decides whether the given tournament is strongly con-
nected.
is_tournament(G) Returns True if and only if G is a tournament.
random_tournament(n[, seed]) Returns a random tournament graph on n nodes.
score_sequence(G) Returns the score sequence for the given tournament
graph.

3.53.1 networkx.algorithms.tournament.hamiltonian_path

hamiltonian_path(G)
Returns a Hamiltonian path in the given tournament graph.
Each tournament has a Hamiltonian path. If furthermore, the tournament is strongly connected, then the returned
Hamiltonian path is a Hamiltonian cycle (by joining the endpoints of the path).
Parameters G (NetworkX graph) – A directed graph representing a tournament.
Returns Whether the given graph is a tournament graph.
Return type bool

3.53. Tournament 457


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This is a recursive implementation with an asymptotic running time of 𝑂(𝑛2 ), ignoring multiplicative polylog-
arithmic factors, where 𝑛 is the number of nodes in the graph.

3.53.2 networkx.algorithms.tournament.is_reachable

is_reachable(G, s, t)
Decides whether there is a path from s to t in the tournament.
This function is more theoretically efficient than the reachability checks than the shortest path algorithms in
networkx.algorithms.shortest_paths.
The given graph must be a tournament, otherwise this function’s behavior is undefined.
Parameters
• G (NetworkX graph) – A directed graph representing a tournament.
• s (node) – A node in the graph.
• t (node) – A node in the graph.
Returns Whether there is a path from s to t in G.
Return type bool

Notes

Although this function is more theoretically efficient than the generic shortest path functions, a speedup requires
the use of parallelism. Though it may in the future, the current implementation does not use parallelism, thus
you may not see much of a speedup.
This algorithm comes from [1].

References

3.53.3 networkx.algorithms.tournament.is_strongly_connected

is_strongly_connected(G)
Decides whether the given tournament is strongly connected.
This function is more theoretically efficient than the is_strongly_connected() function.
The given graph must be a tournament, otherwise this function’s behavior is undefined.
Parameters G (NetworkX graph) – A directed graph representing a tournament.
Returns Whether the tournament is strongly connected.
Return type bool

Notes

Although this function is more theoretically efficient than the generic strong connectivity function, a speedup
requires the use of parallelism. Though it may in the future, the current implementation does not use parallelism,
thus you may not see much of a speedup.

458 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

This algorithm comes from [1].

References

3.53.4 networkx.algorithms.tournament.is_tournament

is_tournament(G)
Returns True if and only if G is a tournament.
A tournament is a directed graph, with neither self-loops nor multi-edges, in which there is exactly one directed
edge joining each pair of distinct nodes.
Parameters G (NetworkX graph) – A directed graph representing a tournament.
Returns Whether the given graph is a tournament graph.
Return type bool

Notes

Some definitions require a self-loop on each node, but that is not the convention used here.

3.53.5 networkx.algorithms.tournament.random_tournament

random_tournament(n, seed=None)
Returns a random tournament graph on n nodes.
Parameters
• n (int) – The number of nodes in the returned graph.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns Whether the given graph is a tournament graph.
Return type bool

Notes

This algorithm adds, for each pair of distinct nodes, an edge with uniformly random orientation. In other words,
binom{n}{2} flips of an unbiased coin decide the orientations of the edges in the graph.

3.53.6 networkx.algorithms.tournament.score_sequence

score_sequence(G)
Returns the score sequence for the given tournament graph.
The score sequence is the sorted list of the out-degrees of the nodes of the graph.
Parameters G (NetworkX graph) – A directed graph representing a tournament.
Returns A sorted list of the out-degrees of the nodes of G.
Return type list

3.53. Tournament 459


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.54 Traversal

3.54.1 Depth First Search

Basic algorithms for depth-first searching the nodes of a graph.

dfs_edges(G[, source, depth_limit]) Iterate over edges in a depth-first-search (DFS).


dfs_tree(G[, source, depth_limit]) Returns oriented tree constructed from a depth-first-
search from source.
dfs_predecessors(G[, source, depth_limit]) Returns dictionary of predecessors in depth-first-search
from source.
dfs_successors(G[, source, depth_limit]) Returns dictionary of successors in depth-first-search
from source.
dfs_preorder_nodes(G[, source, depth_limit]) Generate nodes in a depth-first-search pre-ordering
starting at source.
dfs_postorder_nodes(G[, source, depth_limit]) Generate nodes in a depth-first-search post-ordering
starting at source.
dfs_labeled_edges(G[, source, depth_limit]) Iterate over edges in a depth-first-search (DFS) labeled
by type.

networkx.algorithms.traversal.depth_first_search.dfs_edges

dfs_edges(G, source=None, depth_limit=None)


Iterate over edges in a depth-first-search (DFS).
Perform a depth-first-search over the nodes of G and yield the edges in order. This may not generate all edges
in G (see edge_dfs).
Parameters
• G (NetworkX graph)
• source (node, optional) – Specify starting node for depth-first search and return edges in the
component reachable from source.
• depth_limit (int, optional (default=len(G))) – Specify the maximum search depth.
Returns edges – A generator of edges in the depth-first-search.
Return type generator

Examples

>>> G = nx.path_graph(5)
>>> list(nx.dfs_edges(G, source=0))
[(0, 1), (1, 2), (2, 3), (3, 4)]
>>> list(nx.dfs_edges(G, source=0, depth_limit=2))
[(0, 1), (1, 2)]

Notes

If a source is not specified then a source is chosen arbitrarily and repeatedly until all components in the graph
are searched.

460 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

The implementation of this function is adapted from David Eppstein’s depth-first search function in PADS, with
modifications to allow depth limits based on the Wikipedia article “Depth-limited search”.
See also:
dfs_preorder_nodes(), dfs_postorder_nodes(), dfs_labeled_edges(), edge_dfs()

networkx.algorithms.traversal.depth_first_search.dfs_tree

dfs_tree(G, source=None, depth_limit=None)


Returns oriented tree constructed from a depth-first-search from source.
Parameters
• G (NetworkX graph)
• source (node, optional) – Specify starting node for depth-first search.
• depth_limit (int, optional (default=len(G))) – Specify the maximum search depth.
Returns T – An oriented tree
Return type NetworkX DiGraph

Examples

>>> G = nx.path_graph(5)
>>> T = nx.dfs_tree(G, source=0, depth_limit=2)
>>> list(T.edges())
[(0, 1), (1, 2)]
>>> T = nx.dfs_tree(G, source=0)
>>> list(T.edges())
[(0, 1), (1, 2), (2, 3), (3, 4)]

networkx.algorithms.traversal.depth_first_search.dfs_predecessors

dfs_predecessors(G, source=None, depth_limit=None)


Returns dictionary of predecessors in depth-first-search from source.
Parameters
• G (NetworkX graph)
• source (node, optional) – Specify starting node for depth-first search and return edges in the
component reachable from source.
• depth_limit (int, optional (default=len(G))) – Specify the maximum search depth.
Returns pred – A dictionary with nodes as keys and predecessor nodes as values.
Return type dict

Examples

3.54. Traversal 461


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.path_graph(4)
>>> nx.dfs_predecessors(G, source=0)
{1: 0, 2: 1, 3: 2}
>>> nx.dfs_predecessors(G, source=0, depth_limit=2)
{1: 0, 2: 1}

Notes

If a source is not specified then a source is chosen arbitrarily and repeatedly until all components in the graph
are searched.
The implementation of this function is adapted from David Eppstein’s depth-first search function in PADS, with
modifications to allow depth limits based on the Wikipedia article “Depth-limited search”.

networkx.algorithms.traversal.depth_first_search.dfs_successors

dfs_successors(G, source=None, depth_limit=None)


Returns dictionary of successors in depth-first-search from source.
Parameters
• G (NetworkX graph)
• source (node, optional) – Specify starting node for depth-first search and return edges in the
component reachable from source.
• depth_limit (int, optional (default=len(G))) – Specify the maximum search depth.
Returns succ – A dictionary with nodes as keys and list of successor nodes as values.
Return type dict

Examples

>>> G = nx.path_graph(5)
>>> nx.dfs_successors(G, source=0)
{0: [1], 1: [2], 2: [3], 3: [4]}
>>> nx.dfs_successors(G, source=0, depth_limit=2)
{0: [1], 1: [2]}

Notes

If a source is not specified then a source is chosen arbitrarily and repeatedly until all components in the graph
are searched.
The implementation of this function is adapted from David Eppstein’s depth-first search function in PADS, with
modifications to allow depth limits based on the Wikipedia article “Depth-limited search”.

networkx.algorithms.traversal.depth_first_search.dfs_preorder_nodes

dfs_preorder_nodes(G, source=None, depth_limit=None)


Generate nodes in a depth-first-search pre-ordering starting at source.
Parameters

462 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• G (NetworkX graph)
• source (node, optional) – Specify starting node for depth-first search and return edges in the
component reachable from source.
• depth_limit (int, optional (default=len(G))) – Specify the maximum search depth.
Returns nodes – A generator of nodes in a depth-first-search pre-ordering.
Return type generator

Examples

>>> G = nx.path_graph(5)
>>> list(nx.dfs_preorder_nodes(G, source=0))
[0, 1, 2, 3, 4]
>>> list(nx.dfs_preorder_nodes(G, source=0, depth_limit=2))
[0, 1, 2]

Notes

If a source is not specified then a source is chosen arbitrarily and repeatedly until all components in the graph
are searched.
The implementation of this function is adapted from David Eppstein’s depth-first search function in PADS, with
modifications to allow depth limits based on the Wikipedia article “Depth-limited search”.
See also:
dfs_edges(), dfs_postorder_nodes(), dfs_labeled_edges()

networkx.algorithms.traversal.depth_first_search.dfs_postorder_nodes

dfs_postorder_nodes(G, source=None, depth_limit=None)


Generate nodes in a depth-first-search post-ordering starting at source.
Parameters
• G (NetworkX graph)
• source (node, optional) – Specify starting node for depth-first search and return edges in the
component reachable from source.
• depth_limit (int, optional (default=len(G))) – Specify the maximum search depth.
Returns nodes – A generator of nodes in a depth-first-search post-ordering.
Return type generator

Examples

>>> G = nx.path_graph(5)
>>> list(nx.dfs_postorder_nodes(G, source=0))
[4, 3, 2, 1, 0]
>>> list(nx.dfs_postorder_nodes(G, source=0, depth_limit=2))
[1, 0]

3.54. Traversal 463


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

If a source is not specified then a source is chosen arbitrarily and repeatedly until all components in the graph
are searched.
The implementation of this function is adapted from David Eppstein’s depth-first search function in PADS, with
modifications to allow depth limits based on the Wikipedia article “Depth-limited search”.
See also:
dfs_edges(), dfs_preorder_nodes(), dfs_labeled_edges()

networkx.algorithms.traversal.depth_first_search.dfs_labeled_edges

dfs_labeled_edges(G, source=None, depth_limit=None)


Iterate over edges in a depth-first-search (DFS) labeled by type.
Parameters
• G (NetworkX graph)
• source (node, optional) – Specify starting node for depth-first search and return edges in the
component reachable from source.
• depth_limit (int, optional (default=len(G))) – Specify the maximum search depth.
Returns edges – A generator of triples of the form (u, v, d), where (u, v) is the edge being explored
in the depth-first search and d is one of the strings ‘forward’, ‘nontree’, or ‘reverse’. A ‘forward’
edge is one in which u has been visited but v has not. A ‘nontree’ edge is one in which both u
and v have been visited but the edge is not in the DFS tree. A ‘reverse’ edge is on in which both
u and v have been visited and the edge is in the DFS tree.
Return type generator

Examples

The labels reveal the complete transcript of the depth-first search algorithm in more detail than, for example,
dfs_edges():

>>> from pprint import pprint


>>>
>>> G = nx.DiGraph([(0, 1), (1, 2), (2, 1)])
>>> pprint(list(nx.dfs_labeled_edges(G, source=0)))
[(0, 0, 'forward'),
(0, 1, 'forward'),
(1, 2, 'forward'),
(2, 1, 'nontree'),
(1, 2, 'reverse'),
(0, 1, 'reverse'),
(0, 0, 'reverse')]

Notes

If a source is not specified then a source is chosen arbitrarily and repeatedly until all components in the graph
are searched.

464 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

The implementation of this function is adapted from David Eppstein’s depth-first search function in PADS, with
modifications to allow depth limits based on the Wikipedia article “Depth-limited search”.
See also:
dfs_edges(), dfs_preorder_nodes(), dfs_postorder_nodes()

3.54.2 Breadth First Search

Basic algorithms for breadth-first searching the nodes of a graph.

bfs_edges(G, source[, reverse, depth_limit]) Iterate over edges in a breadth-first-search starting at


source.
bfs_tree(G, source[, reverse, depth_limit]) Returns an oriented tree constructed from of a breadth-
first-search starting at source.
bfs_predecessors(G, source[, depth_limit]) Returns an iterator of predecessors in breadth-first-
search from source.
bfs_successors(G, source[, depth_limit]) Returns an iterator of successors in breadth-first-search
from source.

networkx.algorithms.traversal.breadth_first_search.bfs_edges

bfs_edges(G, source, reverse=False, depth_limit=None)


Iterate over edges in a breadth-first-search starting at source.
Parameters
• G (NetworkX graph)
• source (node) – Specify starting node for breadth-first search and return edges in the com-
ponent reachable from source.
• reverse (bool, optional) – If True traverse a directed graph in the reverse direction
• depth_limit (int, optional(default=len(G))) – Specify the maximum search depth
Returns edges – A generator of edges in the breadth-first-search.
Return type generator

Examples

To get the edges in a breadth-first search:


>>> G = nx.path_graph(3)
>>> list(nx.bfs_edges(G, 0))
[(0, 1), (1, 2)]
>>> list(nx.bfs_edges(G, source=0, depth_limit=1))
[(0, 1)]

To get the nodes in a breadth-first search order:


>>> G = nx.path_graph(3)
>>> root = 2
>>> edges = nx.bfs_edges(G, root)
>>> nodes = [root] + [v for u, v in edges]
(continues on next page)

3.54. Traversal 465


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> nodes
[2, 1, 0]

Notes

Based on http://www.ics.uci.edu/~eppstein/PADS/BFS.py. by D. Eppstein, July 2004. The modifications to


allow depth limits based on the Wikipedia article “Depth-limited-search”.

networkx.algorithms.traversal.breadth_first_search.bfs_tree

bfs_tree(G, source, reverse=False, depth_limit=None)


Returns an oriented tree constructed from of a breadth-first-search starting at source.
Parameters
• G (NetworkX graph)
• source (node) – Specify starting node for breadth-first search and return edges in the com-
ponent reachable from source.
• reverse (bool, optional) – If True traverse a directed graph in the reverse direction
• depth_limit (int, optional(default=len(G))) – Specify the maximum search depth
Returns T – An oriented tree
Return type NetworkX DiGraph

Examples

>>> G = nx.path_graph(3)
>>> print(list(nx.bfs_tree(G,1).edges()))
[(1, 0), (1, 2)]
>>> H = nx.Graph()
>>> nx.add_path(H, [0, 1, 2, 3, 4, 5, 6])
>>> nx.add_path(H, [2, 7, 8, 9, 10])
>>> print(sorted(list(nx.bfs_tree(H, source=3, depth_limit=3).edges())))
[(1, 0), (2, 1), (2, 7), (3, 2), (3, 4), (4, 5), (5, 6), (7, 8)]

Notes

Based on http://www.ics.uci.edu/~eppstein/PADS/BFS.py by D. Eppstein, July 2004. The modifications to allow


depth limits based on the Wikipedia article “Depth-limited-search”.

networkx.algorithms.traversal.breadth_first_search.bfs_predecessors

bfs_predecessors(G, source, depth_limit=None)


Returns an iterator of predecessors in breadth-first-search from source.
Parameters
• G (NetworkX graph)

466 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• source (node) – Specify starting node for breadth-first search and return edges in the com-
ponent reachable from source.
• depth_limit (int, optional(default=len(G))) – Specify the maximum search depth
Returns pred – (node, predecessors) iterator where predecessors is the list of predecessors of the
node.
Return type iterator

Examples

>>> G = nx.path_graph(3)
>>> print(dict(nx.bfs_predecessors(G, 0)))
{1: 0, 2: 1}
>>> H = nx.Graph()
>>> H.add_edges_from([(0, 1), (0, 2), (1, 3), (1, 4), (2, 5), (2, 6)])
>>> print(dict(nx.bfs_predecessors(H, 0)))
{1: 0, 2: 0, 3: 1, 4: 1, 5: 2, 6: 2}
>>> M = nx.Graph()
>>> nx.add_path(M, [0, 1, 2, 3, 4, 5, 6])
>>> nx.add_path(M, [2, 7, 8, 9, 10])
>>> print(sorted(nx.bfs_predecessors(M, source=1, depth_limit=3)))
[(0, 1), (2, 1), (3, 2), (4, 3), (7, 2), (8, 7)]

Notes

Based on http://www.ics.uci.edu/~eppstein/PADS/BFS.py by D. Eppstein, July 2004. The modifications to allow


depth limits based on the Wikipedia article “Depth-limited-search”.

networkx.algorithms.traversal.breadth_first_search.bfs_successors

bfs_successors(G, source, depth_limit=None)


Returns an iterator of successors in breadth-first-search from source.
Parameters
• G (NetworkX graph)
• source (node) – Specify starting node for breadth-first search and return edges in the com-
ponent reachable from source.
• depth_limit (int, optional(default=len(G))) – Specify the maximum search depth
Returns succ – (node, successors) iterator where successors is the list of successors of the node.
Return type iterator

Examples

>>> G = nx.path_graph(3)
>>> print(dict(nx.bfs_successors(G,0)))
{0: [1], 1: [2]}
>>> H = nx.Graph()
>>> H.add_edges_from([(0, 1), (0, 2), (1, 3), (1, 4), (2, 5), (2, 6)])
(continues on next page)

3.54. Traversal 467


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> print(dict(nx.bfs_successors(H, 0)))
{0: [1, 2], 1: [3, 4], 2: [5, 6]}
>>> G = nx.Graph()
>>> nx.add_path(G, [0, 1, 2, 3, 4, 5, 6])
>>> nx.add_path(G, [2, 7, 8, 9, 10])
>>> print(dict(nx.bfs_successors(G, source=1, depth_limit=3)))
{1: [0, 2], 2: [3, 7], 3: [4], 7: [8]}

Notes

Based on http://www.ics.uci.edu/~eppstein/PADS/BFS.py by D. Eppstein, July 2004.The modifications to allow


depth limits based on the Wikipedia article “Depth-limited-search”.

3.54.3 Beam search

Basic algorithms for breadth-first searching the nodes of a graph.

bfs_beam_edges(G, source, value[, width]) Iterates over edges in a beam search.

networkx.algorithms.traversal.beamsearch.bfs_beam_edges

bfs_beam_edges(G, source, value, width=None)


Iterates over edges in a beam search.
The beam search is a generalized breadth-first search in which only the “best” w neighbors of the current node
are enqueued, where w is the beam width and “best” is an application-specific heuristic. In general, a beam
search with a small beam width might not visit each node in the graph.
Parameters
• G (NetworkX graph)
• source (node) – Starting node for the breadth-first search; this function iterates over only
those edges in the component reachable from this node.
• value (function) – A function that takes a node of the graph as input and returns a real
number indicating how “good” it is. A higher value means it is more likely to be visited
sooner during the search. When visiting a new node, only the width neighbors with the
highest value are enqueued (in decreasing order of value).
• width (int (default = None)) – The beam width for the search. This is the number of neigh-
bors (ordered by value) to enqueue when visiting each new node.
Yields edge – Edges in the beam search starting from source, given as a pair of nodes.

Examples

To give nodes with, for example, a higher centrality precedence during the search, set the value function to
return the centrality value of the node:

>>> G = nx.karate_club_graph()
>>> centrality = nx.eigenvector_centrality(G)
(continues on next page)

468 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> source = 0
>>> width = 5
>>> for u, v in nx.bfs_beam_edges(G, source, centrality.get, width):
... print((u, v))

3.54.4 Depth First Search on Edges

Depth First Search on Edges

Algorithms for a depth-first traversal of edges in a graph.

edge_dfs(G[, source, orientation]) A directed, depth-first-search of edges in G, beginning


at source.

networkx.algorithms.traversal.edgedfs.edge_dfs

edge_dfs(G, source=None, orientation=None)


A directed, depth-first-search of edges in G, beginning at source.
Yield the edges of G in a depth-first-search order continuing until all edges are generated.
Parameters
• G (graph) – A directed/undirected graph/multigraph.
• source (node, list of nodes) – The node from which the traversal begins. If None, then a
source is chosen arbitrarily and repeatedly until all edges from each node in the graph are
searched.
• orientation (None | ‘original’ | ‘reverse’ | ‘ignore’ (default: None)) – For directed graphs
and directed multigraphs, edge traversals need not respect the original orientation of the
edges. When set to ‘reverse’ every edge is traversed in the reverse direction. When set to
‘ignore’, every edge is treated as undirected. When set to ‘original’, every edge is treated as
directed. In all three cases, the yielded edge tuples add a last entry to indicate the direction
in which that edge was traversed. If orientation is None, the yielded edge has no direction
indicated. The direction is respected, but not reported.
Yields edge (directed edge) – A directed edge indicating the path taken by the depth-first traversal.
For graphs, edge is of the form (u, v) where u and v are the tail and head of the edge as
determined by the traversal. For multigraphs, edge is of the form (u, v, key), where key
is the key of the edge. When the graph is directed, then u and v are always in the order of the
actual directed edge. If orientation is not None then the edge tuple is extended to include the
direction of traversal (‘forward’ or ‘reverse’) on that edge.

Examples

>>> import networkx as nx


>>> nodes = [0, 1, 2, 3]
>>> edges = [(0, 1), (1, 0), (1, 0), (2, 1), (3, 1)]

>>> list(nx.edge_dfs(nx.Graph(edges), nodes))


[(0, 1), (1, 2), (1, 3)]

3.54. Traversal 469


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> list(nx.edge_dfs(nx.DiGraph(edges), nodes))


[(0, 1), (1, 0), (2, 1), (3, 1)]

>>> list(nx.edge_dfs(nx.MultiGraph(edges), nodes))


[(0, 1, 0), (1, 0, 1), (0, 1, 2), (1, 2, 0), (1, 3, 0)]

>>> list(nx.edge_dfs(nx.MultiDiGraph(edges), nodes))


[(0, 1, 0), (1, 0, 0), (1, 0, 1), (2, 1, 0), (3, 1, 0)]

>>> list(nx.edge_dfs(nx.DiGraph(edges), nodes, orientation='ignore'))


[(0, 1, 'forward'), (1, 0, 'forward'), (2, 1, 'reverse'), (3, 1, 'reverse')]

>>> list(nx.edge_dfs(nx.MultiDiGraph(edges), nodes, orientation='ignore'))


[(0, 1, 0, 'forward'), (1, 0, 0, 'forward'), (1, 0, 1, 'reverse'), (2, 1, 0,
˓→'reverse'), (3, 1, 0, 'reverse')]

Notes

The goal of this function is to visit edges. It differs from the more familiar depth-first traversal of nodes, as
provided by networkx.algorithms.traversal.depth_first_search.dfs_edges(), in that
it does not stop once every node has been visited. In a directed graph with edges [(0, 1), (1, 2), (2, 1)], the edge
(2, 1) would not be visited if not for the functionality provided by this function.
See also:
dfs_edges()

3.54.5 Depth First Search on Edges

Breadth First Search on Edges

Algorithms for a depth-first traversal of edges in a graph.

edge_bfs(G[, source, orientation]) A directed, breadth-first-search of edges in G, beginning


at source.

networkx.algorithms.traversal.edgebfs.edge_bfs

edge_bfs(G, source=None, orientation=None)


A directed, breadth-first-search of edges in G, beginning at source.
Yield the edges of G in a breadth-first-search order continuing until all edges are generated.
Parameters
• G (graph) – A directed/undirected graph/multigraph.
• source (node, list of nodes) – The node from which the traversal begins. If None, then a
source is chosen arbitrarily and repeatedly until all edges from each node in the graph are
searched.
• orientation (None | ‘original’ | ‘reverse’ | ‘ignore’ (default: None)) – For directed graphs
and directed multigraphs, edge traversals need not respect the original orientation of the

470 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

edges. When set to ‘reverse’ every edge is traversed in the reverse direction. When set to
‘ignore’, every edge is treated as undirected. When set to ‘original’, every edge is treated as
directed. In all three cases, the yielded edge tuples add a last entry to indicate the direction
in which that edge was traversed. If orientation is None, the yielded edge has no direction
indicated. The direction is respected, but not reported.
Yields edge (directed edge) – A directed edge indicating the path taken by the breadth-first-search.
For graphs, edge is of the form (u, v) where u and v are the tail and head of the edge as
determined by the traversal. For multigraphs, edge is of the form (u, v, key), where key
is the key of the edge. When the graph is directed, then u and v are always in the order of the
actual directed edge. If orientation is not None then the edge tuple is extended to include the
direction of traversal (‘forward’ or ‘reverse’) on that edge.

Examples

>>> import networkx as nx


>>> nodes = [0, 1, 2, 3]
>>> edges = [(0, 1), (1, 0), (1, 0), (2, 0), (2, 1), (3, 1)]

>>> list(nx.edge_bfs(nx.Graph(edges), nodes))


[(0, 1), (0, 2), (1, 2), (1, 3)]

>>> list(nx.edge_bfs(nx.DiGraph(edges), nodes))


[(0, 1), (1, 0), (2, 0), (2, 1), (3, 1)]

>>> list(nx.edge_bfs(nx.MultiGraph(edges), nodes))


[(0, 1, 0), (0, 1, 1), (0, 1, 2), (0, 2, 0), (1, 2, 0), (1, 3, 0)]

>>> list(nx.edge_bfs(nx.MultiDiGraph(edges), nodes))


[(0, 1, 0), (1, 0, 0), (1, 0, 1), (2, 0, 0), (2, 1, 0), (3, 1, 0)]

>>> list(nx.edge_bfs(nx.DiGraph(edges), nodes, orientation='ignore'))


[(0, 1, 'forward'), (1, 0, 'reverse'), (2, 0, 'reverse'), (2, 1, 'reverse'), (3,
˓→1, 'reverse')]

>>> list(nx.edge_bfs(nx.MultiDiGraph(edges), nodes, orientation='ignore'))


[(0, 1, 0, 'forward'), (1, 0, 0, 'reverse'), (1, 0, 1, 'reverse'), (2, 0, 0,
˓→'reverse'), (2, 1, 0, 'reverse'), (3, 1, 0, 'reverse')]

Notes

The goal of this function is to visit edges. It differs from the more familiar breadth-first-search of nodes,
as provided by networkx.algorithms.traversal.breadth_first_search.bfs_edges(), in
that it does not stop once every node has been visited. In a directed graph with edges [(0, 1), (1, 2), (2, 1)], the
edge (2, 1) would not be visited if not for the functionality provided by this function.
See also:
bfs_edges(), bfs_tree(), edge_dfs()

3.54. Traversal 471


NetworkX Reference, Release 2.3rc1.dev20190222214247

3.55 Tree

3.55.1 Recognition

Recognition Tests

A forest is an acyclic, undirected graph, and a tree is a connected forest. Depending on the subfield, there are various
conventions for generalizing these definitions to directed graphs.
In one convention, directed variants of forest and tree are defined in an identical manner, except that the direction of
the edges is ignored. In effect, each directed edge is treated as a single undirected edge. Then, additional restrictions
are imposed to define branchings and arborescences.
In another convention, directed variants of forest and tree correspond to the previous convention’s branchings and
arborescences, respectively. Then two new terms, polyforest and polytree, are defined to correspond to the other
convention’s forest and tree.
Summarizing:

+-----------------------------+
| Convention A | Convention B |
+=============================+
| forest | polyforest |
| tree | polytree |
| branching | forest |
| arborescence | tree |
+-----------------------------+

Each convention has its reasons. The first convention emphasizes definitional similarity in that directed forests and
trees are only concerned with acyclicity and do not have an in-degree constraint, just as their undirected counterparts do
not. The second convention emphasizes functional similarity in the sense that the directed analog of a spanning tree is
a spanning arborescence. That is, take any spanning tree and choose one node as the root. Then every edge is assigned
a direction such there is a directed path from the root to every other node. The result is a spanning arborescence.
NetworkX follows convention “A”. Explicitly, these are:
undirected forest An undirected graph with no undirected cycles.
undirected tree A connected, undirected forest.
directed forest A directed graph with no undirected cycles. Equivalently, the underlying graph structure (which
ignores edge orientations) is an undirected forest. In convention B, this is known as a polyforest.
directed tree A weakly connected, directed forest. Equivalently, the underlying graph structure (which ignores edge
orientations) is an undirected tree. In convention B, this is known as a polytree.
branching A directed forest with each node having, at most, one parent. So the maximum in-degree is equal to 1. In
convention B, this is known as a forest.
arborescence A directed tree with each node having, at most, one parent. So the maximum in-degree is equal to 1.
In convention B, this is known as a tree.
For trees and arborescences, the adjective “spanning” may be added to designate that the graph, when considered as
a forest/branching, consists of a single tree/arborescence that includes all nodes in the graph. It is true, by definition,
that every tree/arborescence is spanning with respect to the nodes that define the tree/arborescence and so, it might
seem redundant to introduce the notion of “spanning”. However, the nodes may represent a subset of nodes from a
larger graph, and it is in this context that the term “spanning” becomes a useful notion.

472 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

is_tree(G) Returns True if G is a tree.


is_forest(G) Returns True if G is a forest.
is_arborescence(G) Returns True if G is an arborescence.
is_branching(G) Returns True if G is a branching.

networkx.algorithms.tree.recognition.is_tree

is_tree(G)
Returns True if G is a tree.
A tree is a connected graph with no undirected cycles.
For directed graphs, G is a tree if the underlying graph is a tree. The underlying graph is obtained by treating
each directed edge as a single undirected edge in a multigraph.
Parameters G (graph) – The graph to test.
Returns b – A boolean that is True if G is a tree.
Return type bool

Notes

In another convention, a directed tree is known as a polytree and then tree corresponds to an arborescence.
See also:
is_arborescence()

networkx.algorithms.tree.recognition.is_forest

is_forest(G)
Returns True if G is a forest.
A forest is a graph with no undirected cycles.
For directed graphs, G is a forest if the underlying graph is a forest. The underlying graph is obtained by treating
each directed edge as a single undirected edge in a multigraph.
Parameters G (graph) – The graph to test.
Returns b – A boolean that is True if G is a forest.
Return type bool

Notes

In another convention, a directed forest is known as a polyforest and then forest corresponds to a branching.
See also:
is_branching()

3.55. Tree 473


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.tree.recognition.is_arborescence

is_arborescence(G)
Returns True if G is an arborescence.
An arborescence is a directed tree with maximum in-degree equal to 1.
Parameters G (graph) – The graph to test.
Returns b – A boolean that is True if G is an arborescence.
Return type bool

Notes

In another convention, an arborescence is known as a tree.


See also:
is_tree()

networkx.algorithms.tree.recognition.is_branching

is_branching(G)
Returns True if G is a branching.
A branching is a directed forest with maximum in-degree equal to 1.
Parameters G (directed graph) – The directed graph to test.
Returns b – A boolean that is True if G is a branching.
Return type bool

Notes

In another convention, a branching is also known as a forest.


See also:
is_forest()

3.55.2 Branchings and Spanning Arborescences

Algorithms for finding optimum branchings and spanning arborescences.


This implementation is based on:
J. Edmonds, Optimum branchings, J. Res. Natl. Bur. Standards 71B (1967), 233–240. URL: http:
//archive.org/details/jresv71Bn4p233

branching_weight(G[, attr, default]) Returns the total weight of a branching.


greedy_branching(G[, attr, default, kind, seed]) Returns a branching obtained through a greedy algo-
rithm.
maximum_branching(G[, attr, default, . . . ]) Returns a maximum branching from G.
minimum_branching(G[, attr, default, . . . ]) Returns a minimum branching from G.
Continued on next page

474 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 139 – continued from previous page


maximum_spanning_arborescence(G[, attr, Returns a maximum spanning arborescence from G.
. . . ])
minimum_spanning_arborescence(G[, attr, Returns a minimum spanning arborescence from G.
. . . ])
Edmonds(G[, seed]) Edmonds algorithm for finding optimal branchings and
spanning arborescences.

networkx.algorithms.tree.branchings.branching_weight

branching_weight(G, attr=’weight’, default=1)


Returns the total weight of a branching.

networkx.algorithms.tree.branchings.greedy_branching

greedy_branching(G, attr=’weight’, default=1, kind=’max’, seed=None)


Returns a branching obtained through a greedy algorithm.
This algorithm is wrong, and cannot give a proper optimal branching. However, we include it for pedagogical
reasons, as it can be helpful to see what its outputs are.
The output is a branching, and possibly, a spanning arborescence. However, it is not guaranteed to be optimal
in either case.
Parameters
• G (DiGraph) – The directed graph to scan.
• attr (str) – The attribute to use as weights. If None, then each edge will be treated equally
with a weight of 1.
• default (float) – When attr is not None, then if an edge does not have that attribute,
default specifies what value it should take.
• kind (str) – The type of optimum to search for: ‘min’ or ‘max’ greedy branching.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns B – The greedily obtained branching.
Return type directed graph

networkx.algorithms.tree.branchings.maximum_branching

maximum_branching(G, attr=’weight’, default=1, preserve_attrs=False)


Returns a maximum branching from G.
Parameters
• G ((multi)digraph-like) – The graph to be searched.
• attr (str) – The edge attribute used to in determining optimality.
• default (float) – The value of the edge attribute used if an edge does not have the attribute
attr.
• preserve_attrs (bool) – If True, preserve the other attributes of the original graph (that are
not passed to attr)

3.55. Tree 475


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns B – A maximum branching.


Return type (multi)digraph-like

networkx.algorithms.tree.branchings.minimum_branching

minimum_branching(G, attr=’weight’, default=1, preserve_attrs=False)


Returns a minimum branching from G.
Parameters
• G ((multi)digraph-like) – The graph to be searched.
• attr (str) – The edge attribute used to in determining optimality.
• default (float) – The value of the edge attribute used if an edge does not have the attribute
attr.
• preserve_attrs (bool) – If True, preserve the other attributes of the original graph (that are
not passed to attr)
Returns B – A minimum branching.
Return type (multi)digraph-like

networkx.algorithms.tree.branchings.maximum_spanning_arborescence

maximum_spanning_arborescence(G, attr=’weight’, default=1, preserve_attrs=False)


Returns a maximum spanning arborescence from G.
Parameters
• G ((multi)digraph-like) – The graph to be searched.
• attr (str) – The edge attribute used to in determining optimality.
• default (float) – The value of the edge attribute used if an edge does not have the attribute
attr.
• preserve_attrs (bool) – If True, preserve the other attributes of the original graph (that are
not passed to attr)
Returns B – A maximum spanning arborescence.
Return type (multi)digraph-like
Raises NetworkXException – If the graph does not contain a maximum spanning arborescence.

networkx.algorithms.tree.branchings.minimum_spanning_arborescence

minimum_spanning_arborescence(G, attr=’weight’, default=1, preserve_attrs=False)


Returns a minimum spanning arborescence from G.
Parameters
• G ((multi)digraph-like) – The graph to be searched.
• attr (str) – The edge attribute used to in determining optimality.
• default (float) – The value of the edge attribute used if an edge does not have the attribute
attr.

476 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

• preserve_attrs (bool) – If True, preserve the other attributes of the original graph (that are
not passed to attr)
Returns B – A minimum spanning arborescence.
Return type (multi)digraph-like
Raises NetworkXException – If the graph does not contain a minimum spanning arborescence.

networkx.algorithms.tree.branchings.Edmonds

class Edmonds(G, seed=None)


Edmonds algorithm for finding optimal branchings and spanning arborescences.
__init__(G, seed=None)
Initialize self. See help(type(self)) for accurate signature.

Methods

__init__(G[, seed]) Initialize self.


find_optimum([attr, default, kind, style, . . . ]) Returns a branching from G.

3.55.3 Encoding and decoding

Functions for encoding and decoding trees.


Since a tree is a highly restricted form of graph, it can be represented concisely in several ways. This module includes
functions for encoding and decoding trees in the form of nested tuples and Prüfer sequences. The former requires a
rooted tree, whereas the latter can be applied to unrooted trees. Furthermore, there is a bijection from Prüfer sequences
to labeled trees.

from_nested_tuple(sequence[, . . . ]) Returns the rooted tree corresponding to the given


nested tuple.
to_nested_tuple(T, root[, canonical_form]) Returns a nested tuple representation of the given tree.
from_prufer_sequence(sequence) Returns the tree corresponding to the given Prüfer se-
quence.
to_prufer_sequence(T) Returns the Prüfer sequence of the given tree.

networkx.algorithms.tree.coding.from_nested_tuple

from_nested_tuple(sequence, sensible_relabeling=False)
Returns the rooted tree corresponding to the given nested tuple.
The nested tuple representation of a tree is defined recursively. The tree with one node and no edges is rep-
resented by the empty tuple, (). A tree with k subtrees is represented by a tuple of length k in which each
element is the nested tuple representation of a subtree.
Parameters
• sequence (tuple) – A nested tuple representing a rooted tree.
• sensible_relabeling (bool) – Whether to relabel the nodes of the tree so that nodes are
labeled in increasing order according to their breadth-first search order from the root node.

3.55. Tree 477


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns The tree corresponding to the given nested tuple, whose root node is node 0. If
sensible_labeling is True, nodes will be labeled in breadth-first search order starting
from the root node.
Return type NetworkX graph

Notes

This function is not the inverse of to_nested_tuple(); the only guarantee is that the rooted trees are
isomorphic.
See also:
to_nested_tuple(), from_prufer_sequence()

Examples

Sensible relabeling ensures that the nodes are labeled from the root starting at 0:

>>> balanced = (((), ()), ((), ()))


>>> T = nx.from_nested_tuple(balanced, sensible_relabeling=True)
>>> edges = [(0, 1), (0, 2), (1, 3), (1, 4), (2, 5), (2, 6)]
>>> all((u, v) in T.edges() or (v, u) in T.edges() for (u, v) in edges)
True

networkx.algorithms.tree.coding.to_nested_tuple

to_nested_tuple(T, root, canonical_form=False)


Returns a nested tuple representation of the given tree.
The nested tuple representation of a tree is defined recursively. The tree with one node and no edges is rep-
resented by the empty tuple, (). A tree with k subtrees is represented by a tuple of length k in which each
element is the nested tuple representation of a subtree.
Parameters
• T (NetworkX graph) – An undirected graph object representing a tree.
• root (node) – The node in T to interpret as the root of the tree.
• canonical_form (bool) – If True, each tuple is sorted so that the function returns a canon-
ical form for rooted trees. This means “lighter” subtrees will appear as nested tuples before
“heavier” subtrees. In this way, each isomorphic rooted tree has the same nested tuple rep-
resentation.
Returns A nested tuple representation of the tree.
Return type tuple

Notes

This function is not the inverse of from_nested_tuple(); the only guarantee is that the rooted trees are
isomorphic.
See also:
from_nested_tuple(), to_prufer_sequence()

478 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

The tree need not be a balanced binary tree:

>>> T = nx.Graph()
>>> T.add_edges_from([(0, 1), (0, 2), (0, 3)])
>>> T.add_edges_from([(1, 4), (1, 5)])
>>> T.add_edges_from([(3, 6), (3, 7)])
>>> root = 0
>>> nx.to_nested_tuple(T, root)
(((), ()), (), ((), ()))

Continuing the above example, if canonical_form is True, the nested tuples will be sorted:

>>> nx.to_nested_tuple(T, root, canonical_form=True)


((), ((), ()), ((), ()))

Even the path graph can be interpreted as a tree:

>>> T = nx.path_graph(4)
>>> root = 0
>>> nx.to_nested_tuple(T, root)
((((),),),)

networkx.algorithms.tree.coding.from_prufer_sequence

from_prufer_sequence(sequence)
Returns the tree corresponding to the given Prüfer sequence.
A Prüfer sequence is a list of n - 2 numbers between 0 and n - 1, inclusive. The tree corresponding to a given
Prüfer sequence can be recovered by repeatedly joining a node in the sequence with a node with the smallest
potential degree according to the sequence.
Parameters sequence (list) – A Prüfer sequence, which is a list of n - 2 integers between zero and
n - 1, inclusive.
Returns The tree corresponding to the given Prüfer sequence.
Return type NetworkX graph

Notes

There is a bijection from labeled trees to Prüfer sequences. This function is the inverse of the
from_prufer_sequence() function.
Sometimes Prüfer sequences use nodes labeled from 1 to n instead of from 0 to n - 1. This function requires
nodes to be labeled in the latter form. You can use networkx.relabel_nodes() to relabel the nodes of
your tree to the appropriate format.
This implementation is from1 and has a running time of 𝑂(𝑛 log 𝑛).
1 Wang, Xiaodong, Lei Wang, and Yingjie Wu. “An optimal algorithm for Prufer codes.” Journal of Software Engineering and Applications

2.02 (2009): 111. <https://doi.org/10.4236/jsea.2009.22016>

3.55. Tree 479


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

See also:
from_nested_tuple(), to_prufer_sequence()

Examples

There is a bijection between Prüfer sequences and labeled trees, so this function is the inverse of the
to_prufer_sequence() function:
>>> edges = [(0, 3), (1, 3), (2, 3), (3, 4), (4, 5)]
>>> tree = nx.Graph(edges)
>>> sequence = nx.to_prufer_sequence(tree)
>>> sequence
[3, 3, 3, 4]
>>> tree2 = nx.from_prufer_sequence(sequence)
>>> list(tree2.edges()) == edges
True

networkx.algorithms.tree.coding.to_prufer_sequence

to_prufer_sequence(T)
Returns the Prüfer sequence of the given tree.
A Prüfer sequence is a list of n - 2 numbers between 0 and n - 1, inclusive. The tree corresponding to a given
Prüfer sequence can be recovered by repeatedly joining a node in the sequence with a node with the smallest
potential degree according to the sequence.
Parameters T (NetworkX graph) – An undirected graph object representing a tree.
Returns The Prüfer sequence of the given tree.
Return type list
Raises
• NetworkXPointlessConcept – If the number of nodes in T is less than two.
• NotATree – If T is not a tree.
• KeyError – If the set of nodes in T is not {0, . . . , n - 1}.

Notes

There is a bijection from labeled trees to Prüfer sequences. This function is the inverse of the
from_prufer_sequence() function.
Sometimes Prüfer sequences use nodes labeled from 1 to n instead of from 0 to n - 1. This function requires
nodes to be labeled in the latter form. You can use relabel_nodes() to relabel the nodes of your tree to the
appropriate format.
This implementation is from1 and has a running time of 𝑂(𝑛 log 𝑛).
See also:
to_nested_tuple(), from_prufer_sequence()
1 Wang, Xiaodong, Lei Wang, and Yingjie Wu. “An optimal algorithm for Prufer codes.” Journal of Software Engineering and Applications
2.02 (2009): 111. <https://doi.org/10.4236/jsea.2009.22016>

480 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

Examples

There is a bijection between Prüfer sequences and labeled trees, so this function is the inverse of the
from_prufer_sequence() function:

>>> edges = [(0, 3), (1, 3), (2, 3), (3, 4), (4, 5)]
>>> tree = nx.Graph(edges)
>>> sequence = nx.to_prufer_sequence(tree)
>>> sequence
[3, 3, 3, 4]
>>> tree2 = nx.from_prufer_sequence(sequence)
>>> list(tree2.edges()) == edges
True

3.55.4 Operations

Operations on trees.

join(rooted_trees[, label_attribute]) Returns a new rooted tree with a root node joined with
the roots of each of the given rooted trees.

networkx.algorithms.tree.operations.join

join(rooted_trees, label_attribute=None)
Returns a new rooted tree with a root node joined with the roots of each of the given rooted trees.
Parameters
• rooted_trees (list) – A list of pairs in which each left element is a NetworkX graph object
representing a tree and each right element is the root node of that tree. The nodes of these
trees will be relabeled to integers.
• label_attribute (str) – If provided, the old node labels will be stored in the new tree under
this node attribute. If not provided, the node attribute '_old' will store the original label
of the node in the rooted trees given in the input.
Returns The rooted tree whose subtrees are the given rooted trees. The new root node is la-
beled 0. Each non-root node has an attribute, as described under the keyword argument
label_attribute, that indicates the label of the original node in the input tree.
Return type NetworkX graph

Notes

Graph, edge, and node attributes are propagated from the given rooted trees to the created tree. If there are
any overlapping graph attributes, those from later trees will overwrite those from earlier trees in the tuple of
positional arguments.

Examples

Join two full balanced binary trees of height h to get a full balanced binary tree of depth h + 1:

3.55. Tree 481


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> h = 4
>>> left = nx.balanced_tree(2, h)
>>> right = nx.balanced_tree(2, h)
>>> joined_tree = nx.join([(left, 0), (right, 0)])
>>> nx.is_isomorphic(joined_tree, nx.balanced_tree(2, h + 1))
True

3.55.5 Spanning Trees

Algorithms for calculating min/max spanning trees/forests.

minimum_spanning_tree(G[, weight, . . . ]) Returns a minimum spanning tree or forest on an undi-


rected graph G.
maximum_spanning_tree(G[, weight, . . . ]) Returns a maximum spanning tree or forest on an undi-
rected graph G.
minimum_spanning_edges(G[, algorithm, . . . ]) Generate edges in a minimum spanning forest of an
undirected weighted graph.
maximum_spanning_edges(G[, algorithm, . . . ]) Generate edges in a maximum spanning forest of an
undirected weighted graph.

networkx.algorithms.tree.mst.minimum_spanning_tree

minimum_spanning_tree(G, weight=’weight’, algorithm=’kruskal’, ignore_nan=False)


Returns a minimum spanning tree or forest on an undirected graph G.
Parameters
• G (undirected graph) – An undirected graph. If G is connected, then the algorithm finds a
spanning tree. Otherwise, a spanning forest is found.
• weight (str) – Data key to use for edge weights.
• algorithm (string) – The algorithm to use when finding a minimum spanning tree. Valid
choices are ‘kruskal’, ‘prim’, or ‘boruvka’. The default is ‘kruskal’.
• ignore_nan (bool (default: False)) – If a NaN is found as an edge weight normally an
exception is raised. If ignore_nan is True then that edge is ignored instead.
Returns G – A minimum spanning tree or forest.
Return type NetworkX Graph

Examples

>>> G = nx.cycle_graph(4)
>>> G.add_edge(0, 3, weight=2)
>>> T = nx.minimum_spanning_tree(G)
>>> sorted(T.edges(data=True))
[(0, 1, {}), (1, 2, {}), (2, 3, {})]

Notes

For Borůvka’s algorithm, each edge must have a weight attribute, and each edge weight must be distinct.

482 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

For the other algorithms, if the graph edges do not have a weight attribute a default weight of 1 will be used.
There may be more than one tree with the same minimum or maximum weight. See networkx.tree.
recognition for more detailed definitions.
Isolated nodes with self-loops are in the tree as edgeless isolated nodes.

networkx.algorithms.tree.mst.maximum_spanning_tree

maximum_spanning_tree(G, weight=’weight’, algorithm=’kruskal’, ignore_nan=False)


Returns a maximum spanning tree or forest on an undirected graph G.
Parameters
• G (undirected graph) – An undirected graph. If G is connected, then the algorithm finds a
spanning tree. Otherwise, a spanning forest is found.
• weight (str) – Data key to use for edge weights.
• algorithm (string) – The algorithm to use when finding a maximum spanning tree. Valid
choices are ‘kruskal’, ‘prim’, or ‘boruvka’. The default is ‘kruskal’.
• ignore_nan (bool (default: False)) – If a NaN is found as an edge weight normally an
exception is raised. If ignore_nan is True then that edge is ignored instead.
Returns G – A maximum spanning tree or forest.
Return type NetworkX Graph

Examples

>>> G = nx.cycle_graph(4)
>>> G.add_edge(0, 3, weight=2)
>>> T = nx.maximum_spanning_tree(G)
>>> sorted(T.edges(data=True))
[(0, 1, {}), (0, 3, {'weight': 2}), (1, 2, {})]

Notes

For Borůvka’s algorithm, each edge must have a weight attribute, and each edge weight must be distinct.
For the other algorithms, if the graph edges do not have a weight attribute a default weight of 1 will be used.
There may be more than one tree with the same minimum or maximum weight. See networkx.tree.
recognition for more detailed definitions.
Isolated nodes with self-loops are in the tree as edgeless isolated nodes.

networkx.algorithms.tree.mst.minimum_spanning_edges

minimum_spanning_edges(G, algorithm=’kruskal’, weight=’weight’, keys=True, data=True, ig-


nore_nan=False)
Generate edges in a minimum spanning forest of an undirected weighted graph.
A minimum spanning tree is a subgraph of the graph (a tree) with the minimum sum of edge weights. A spanning
forest is a union of the spanning trees for each connected component of the graph.
Parameters

3.55. Tree 483


NetworkX Reference, Release 2.3rc1.dev20190222214247

• G (undirected Graph) – An undirected graph. If G is connected, then the algorithm finds a


spanning tree. Otherwise, a spanning forest is found.
• algorithm (string) – The algorithm to use when finding a minimum spanning tree. Valid
choices are ‘kruskal’, ‘prim’, or ‘boruvka’. The default is ‘kruskal’.
• weight (string) – Edge data key to use for weight (default ‘weight’).
• keys (bool) – Whether to yield edge key in multigraphs in addition to the edge. If G is not a
multigraph, this is ignored.
• data (bool, optional) – If True yield the edge data along with the edge.
• ignore_nan (bool (default: False)) – If a NaN is found as an edge weight normally an
exception is raised. If ignore_nan is True then that edge is ignored instead.
Returns
edges – An iterator over edges in a maximum spanning tree of G. Edges connecting nodes u and
v are represented as tuples: (u, v, k, d) or (u, v, k) or (u, v, d) or (u, v)
If G is a multigraph, keys indicates whether the edge key k will be reported in the third position
in the edge tuple. data indicates whether the edge datadict d will appear at the end of the edge
tuple.
If G is not a multigraph, the tuples are (u, v, d) if data is True or (u, v) if data is
False.
Return type iterator

Examples

>>> from networkx.algorithms import tree

Find minimum spanning edges by Kruskal’s algorithm

>>> G = nx.cycle_graph(4)
>>> G.add_edge(0, 3, weight=2)
>>> mst = tree.minimum_spanning_edges(G, algorithm='kruskal', data=False)
>>> edgelist = list(mst)
>>> sorted(edgelist)
[(0, 1), (1, 2), (2, 3)]

Find minimum spanning edges by Prim’s algorithm

>>> G = nx.cycle_graph(4)
>>> G.add_edge(0, 3, weight=2)
>>> mst = tree.minimum_spanning_edges(G, algorithm='prim', data=False)
>>> edgelist = list(mst)
>>> sorted(edgelist)
[(0, 1), (1, 2), (2, 3)]

Notes

For Borůvka’s algorithm, each edge must have a weight attribute, and each edge weight must be distinct.
For the other algorithms, if the graph edges do not have a weight attribute a default weight of 1 will be used.
Modified code from David Eppstein, April 2006 http://www.ics.uci.edu/~eppstein/PADS/

484 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.tree.mst.maximum_spanning_edges

maximum_spanning_edges(G, algorithm=’kruskal’, weight=’weight’, keys=True, data=True, ig-


nore_nan=False)
Generate edges in a maximum spanning forest of an undirected weighted graph.
A maximum spanning tree is a subgraph of the graph (a tree) with the maximum possible sum of edge weights.
A spanning forest is a union of the spanning trees for each connected component of the graph.
Parameters
• G (undirected Graph) – An undirected graph. If G is connected, then the algorithm finds a
spanning tree. Otherwise, a spanning forest is found.
• algorithm (string) – The algorithm to use when finding a maximum spanning tree. Valid
choices are ‘kruskal’, ‘prim’, or ‘boruvka’. The default is ‘kruskal’.
• weight (string) – Edge data key to use for weight (default ‘weight’).
• keys (bool) – Whether to yield edge key in multigraphs in addition to the edge. If G is not a
multigraph, this is ignored.
• data (bool, optional) – If True yield the edge data along with the edge.
• ignore_nan (bool (default: False)) – If a NaN is found as an edge weight normally an
exception is raised. If ignore_nan is True then that edge is ignored instead.
Returns
edges – An iterator over edges in a maximum spanning tree of G. Edges connecting nodes u and
v are represented as tuples: (u, v, k, d) or (u, v, k) or (u, v, d) or (u, v)
If G is a multigraph, keys indicates whether the edge key k will be reported in the third position
in the edge tuple. data indicates whether the edge datadict d will appear at the end of the edge
tuple.
If G is not a multigraph, the tuples are (u, v, d) if data is True or (u, v) if data is
False.
Return type iterator

Examples

>>> from networkx.algorithms import tree

Find maximum spanning edges by Kruskal’s algorithm

>>> G = nx.cycle_graph(4)
>>> G.add_edge(0, 3, weight=2)
>>> mst = tree.maximum_spanning_edges(G, algorithm='kruskal', data=False)
>>> edgelist = list(mst)
>>> sorted(edgelist)
[(0, 1), (0, 3), (1, 2)]

Find maximum spanning edges by Prim’s algorithm

>>> G = nx.cycle_graph(4)
>>> G.add_edge(0, 3, weight=2) # assign weight 2 to edge 0-3
>>> mst = tree.maximum_spanning_edges(G, algorithm='prim', data=False)
>>> edgelist = list(mst)
(continues on next page)

3.55. Tree 485


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> sorted(edgelist)
[(0, 1), (0, 3), (3, 2)]

Notes

For Borůvka’s algorithm, each edge must have a weight attribute, and each edge weight must be distinct.
For the other algorithms, if the graph edges do not have a weight attribute a default weight of 1 will be used.
Modified code from David Eppstein, April 2006 http://www.ics.uci.edu/~eppstein/PADS/

3.55.6 Exceptions

Functions for encoding and decoding trees.


Since a tree is a highly restricted form of graph, it can be represented concisely in several ways. This module includes
functions for encoding and decoding trees in the form of nested tuples and Prüfer sequences. The former requires a
rooted tree, whereas the latter can be applied to unrooted trees. Furthermore, there is a bijection from Prüfer sequences
to labeled trees.

NotATree Raised when a function expects a tree (that is, a con-


nected undirected graph with no cycles) but gets a non-
tree graph as input instead.

networkx.algorithms.tree.coding.NotATree

exception NotATree
Raised when a function expects a tree (that is, a connected undirected graph with no cycles) but gets a non-tree
graph as input instead.

3.56 Triads

Functions for analyzing triads of a graph.

triadic_census(G) Determines the triadic census of a directed graph.

3.56.1 networkx.algorithms.triads.triadic_census

triadic_census(G)
Determines the triadic census of a directed graph.
The triadic census is a count of how many of the 16 possible types of triads are present in a directed graph.
Parameters G (digraph) – A NetworkX DiGraph
Returns census – Dictionary with triad names as keys and number of occurrences as values.
Return type dict

486 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This algorithm has complexity 𝑂(𝑚) where 𝑚 is the number of edges in the graph.
See also:
triad_graph()

References

3.57 Vitality

Vitality measures.

closeness_vitality(G[, node, weight, . . . ]) Returns the closeness vitality for nodes in the graph.

3.57.1 networkx.algorithms.vitality.closeness_vitality

closeness_vitality(G, node=None, weight=None, wiener_index=None)


Returns the closeness vitality for nodes in the graph.
The closeness vitality of a node, defined in Section 3.6.2 of [1], is the change in the sum of distances between
all node pairs when excluding that node.
Parameters
• G (NetworkX graph) – A strongly-connected graph.
• weight (string) – The name of the edge attribute used as weight. This is passed directly to
the wiener_index() function.
• node (object) – If specified, only the closeness vitality for this node will be returned. Oth-
erwise, a dictionary mapping each node to its closeness vitality will be returned.
Other Parameters wiener_index (number) – If you have already computed the Wiener index of the
graph G, you can provide that value here. Otherwise, it will be computed for you.
Returns
If node is None, this function returns a dictionary with nodes as keys and closeness vitality as
the value. Otherwise, it returns only the closeness vitality for the specified node.
The closeness vitality of a node may be negative infinity if removing that node would disconnect
the graph.
Return type dictionary or float

Examples

>>> G = nx.cycle_graph(3)
>>> nx.closeness_vitality(G)
{0: 2.0, 1: 2.0, 2: 2.0}

See also:
closeness_centrality()

3.57. Vitality 487


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.58 Voronoi cells

Functions for computing the Voronoi cells of a graph.

voronoi_cells(G, center_nodes[, weight]) Returns the Voronoi cells centered at center_nodes


with respect to the shortest-path distance metric.

3.58.1 networkx.algorithms.voronoi.voronoi_cells

voronoi_cells(G, center_nodes, weight=’weight’)


Returns the Voronoi cells centered at center_nodes with respect to the shortest-path distance metric.
If C is a set of nodes in the graph and c is an element of C, the Voronoi cell centered at a node c is the set of all
nodes v that are closer to c than to any other center node in C with respect to the shortest-path distance metric.1
For directed graphs, this will compute the “outward” Voronoi cells, as defined in1 , in which distance is measured
from the center nodes to the target node. For the “inward” Voronoi cells, use the DiGraph.reverse()
method to reverse the orientation of the edges before invoking this function on the directed graph.
Parameters
• G (NetworkX graph)
• center_nodes (set) – A nonempty set of nodes in the graph G that represent the center of the
Voronoi cells.
• weight (string or function) – The edge attribute (or an arbitrary function) representing
the weight of an edge. This keyword argument is as described in the documentation for
multi_source_dijkstra_path(), for example.
Returns A mapping from center node to set of all nodes in the graph closer to that center node than
to any other center node. The keys of the dictionary are the element of center_nodes, and
the values of the dictionary form a partition of the nodes of G.
Return type dictionary

Examples

To get only the partition of the graph induced by the Voronoi cells, take the collection of all values in the returned
dictionary:

>>> G = nx.path_graph(6)
>>> center_nodes = {0, 3}
>>> cells = nx.voronoi_cells(G, center_nodes)
>>> partition = set(map(frozenset, cells.values()))
>>> sorted(map(sorted, partition))
[[0, 1], [2, 3, 4, 5]]

Raises ValueError – If center_nodes is empty.


1 Erwig, Martin. (2000), “The graph Voronoi diagram with applications.” Networks, 36: 156–163. <dx.doi.org/10.1002/1097-
0037(200010)36:3<156::AID-NET2>3.0.CO;2-L>

488 Chapter 3. Algorithms


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

3.59 Wiener index

Functions related to the Wiener index of a graph.

wiener_index(G[, weight]) Returns the Wiener index of the given graph.

3.59.1 networkx.algorithms.wiener.wiener_index

wiener_index(G, weight=None)
Returns the Wiener index of the given graph.
The Wiener index of a graph is the sum of the shortest-path distances between each pair of reachable nodes. For
pairs of nodes in undirected graphs, only one orientation of the pair is counted.
Parameters
• G (NetworkX graph)
• weight (object) – The edge attribute to use as distance when computing shortest-path dis-
tances. This is passed directly to the networkx.shortest_path_length() func-
tion.
Returns The Wiener index of the graph G.
Return type float
Raises NetworkXError – If the graph G is not connected.

Notes

If a pair of nodes is not reachable, the distance is assumed to be infinity. This means that for graphs that are not
strongly-connected, this function returns inf.
The Wiener index is not usually defined for directed graphs, however this function uses the natural generalization
of the Wiener index to directed graphs.

Examples

The Wiener index of the (unweighted) complete graph on n nodes equals the number of pairs of the n nodes,
since each pair of nodes is at distance one:

>>> import networkx as nx


>>> n = 10
>>> G = nx.complete_graph(n)
>>> nx.wiener_index(G) == n * (n - 1) / 2
True

Graphs that are not strongly-connected have infinite Wiener index:

>>> G = nx.empty_graph(2)
>>> nx.wiener_index(G)
inf

3.59. Wiener index 489


NetworkX Reference, Release 2.3rc1.dev20190222214247

490 Chapter 3. Algorithms


CHAPTER

FOUR

FUNCTIONS

Functional interface to graph methods and assorted utilities.

4.1 Graph

degree(G[, nbunch, weight]) Returns a degree view of single node or of nbunch of


nodes.
degree_histogram(G) Returns a list of the frequency of each degree value.
density(G) Returns the density of a graph.
info(G[, n]) Print short summary of information for the graph G or
the node n.
create_empty_copy(G[, with_data]) Returns a copy of the graph G with all of the edges re-
moved.
is_directed(G) Return True if graph is directed.
to_directed(graph) Returns a directed view of the graph graph.
to_undirected(graph) Returns an undirected view of the graph graph.
is_empty(G) Returns True if G has no edges.
add_star(G_to_add_to, nodes_for_star, **attr) Add a star to Graph G_to_add_to.
add_path(G_to_add_to, nodes_for_path, **attr) Add a path to the Graph G_to_add_to.
add_cycle(G_to_add_to, nodes_for_cycle, **attr) Add a cycle to the Graph G_to_add_to.
subgraph(G, nbunch) Returns the subgraph induced on nodes in nbunch.
induced_subgraph(G, nbunch) Returns a SubGraph view of G showing only nodes in
nbunch.
restricted_view(G, nodes, edges) Returns a view of G with hidden nodes and edges.
reverse_view(digraph) Provide a reverse view of the digraph with edges re-
versed.
edge_subgraph(G, edges) Returns a view of the subgraph induced by the specified
edges.

4.1.1 networkx.classes.function.degree

degree(G, nbunch=None, weight=None)


Returns a degree view of single node or of nbunch of nodes. If nbunch is omitted, then return degrees of all
nodes.

491
NetworkX Reference, Release 2.3rc1.dev20190222214247

4.1.2 networkx.classes.function.degree_histogram

degree_histogram(G)
Returns a list of the frequency of each degree value.
Parameters G (Networkx graph) – A graph
Returns hist – A list of frequencies of degrees. The degree values are the index in the list.
Return type list

Notes

Note: the bins are width one, hence len(list) can be large (Order(number_of_edges))

4.1.3 networkx.classes.function.density

density(G)
Returns the density of a graph.
The density for undirected graphs is
2𝑚
𝑑= ,
𝑛(𝑛 − 1)

and for directed graphs is


𝑚
𝑑= ,
𝑛(𝑛 − 1)

where n is the number of nodes and m is the number of edges in G.

Notes

The density is 0 for a graph without edges and 1 for a complete graph. The density of multigraphs can be higher
than 1.
Self loops are counted in the total number of edges so graphs with self loops can have density higher than 1.

4.1.4 networkx.classes.function.info

info(G, n=None)
Print short summary of information for the graph G or the node n.
Parameters
• G (Networkx graph) – A graph
• n (node (any hashable)) – A node in the graph G

4.1.5 networkx.classes.function.create_empty_copy

create_empty_copy(G, with_data=True)
Returns a copy of the graph G with all of the edges removed.
Parameters

492 Chapter 4. Functions


NetworkX Reference, Release 2.3rc1.dev20190222214247

• G (graph) – A NetworkX graph


• with_data (bool (default=True)) – Propagate Graph and Nodes data to the new graph.
See also:
empty_graph()

4.1.6 networkx.classes.function.is_directed

is_directed(G)
Return True if graph is directed.

4.1.7 networkx.classes.function.to_directed

to_directed(graph)
Returns a directed view of the graph graph.
Identical to graph.to_directed(as_view=True) Note that graph.to_directed defaults to as_view=False while
this function always provides a view.

4.1.8 networkx.classes.function.to_undirected

to_undirected(graph)
Returns an undirected view of the graph graph.
Identical to graph.to_undirected(as_view=True) Note that graph.to_undirected defaults to as_view=False
while this function always provides a view.

4.1.9 networkx.classes.function.is_empty

is_empty(G)
Returns True if G has no edges.
Parameters G (graph) – A NetworkX graph.
Returns True if G has no edges, and False otherwise.
Return type bool

Notes

An empty graph can have nodes but not edges. The empty graph with zero nodes is known as the null graph.
This is an 𝑂(𝑛) operation where n is the number of nodes in the graph.

4.1.10 networkx.classes.function.add_star

add_star(G_to_add_to, nodes_for_star, **attr)


Add a star to Graph G_to_add_to.
The first node in nodes_for_star is the middle of the star. It is connected to all other nodes.
Parameters

4.1. Graph 493


NetworkX Reference, Release 2.3rc1.dev20190222214247

• G_to_add_to (graph) – A NetworkX graph


• nodes_for_star (iterable container) – A container of nodes.
• attr (keyword arguments, optional (default= no attributes)) – Attributes to add to every edge
in star.
See also:
add_path(), add_cycle()

Examples

>>> G = nx.Graph()
>>> nx.add_star(G, [0, 1, 2, 3])
>>> nx.add_star(G, [10, 11, 12], weight=2)

4.1.11 networkx.classes.function.add_path

add_path(G_to_add_to, nodes_for_path, **attr)


Add a path to the Graph G_to_add_to.
Parameters
• G_to_add_to (graph) – A NetworkX graph
• nodes_for_path (iterable container) – A container of nodes. A path will be constructed
from the nodes (in order) and added to the graph.
• attr (keyword arguments, optional (default= no attributes)) – Attributes to add to every edge
in path.
See also:
add_star(), add_cycle()

Examples

>>> G = nx.Graph()
>>> nx.add_path(G, [0, 1, 2, 3])
>>> nx.add_path(G, [10, 11, 12], weight=7)

4.1.12 networkx.classes.function.add_cycle

add_cycle(G_to_add_to, nodes_for_cycle, **attr)


Add a cycle to the Graph G_to_add_to.
Parameters
• G_to_add_to (graph) – A NetworkX graph
• nodes_for_cycle (iterable container) – A container of nodes. A cycle will be constructed
from the nodes (in order) and added to the graph.
• attr (keyword arguments, optional (default= no attributes)) – Attributes to add to every edge
in cycle.

494 Chapter 4. Functions


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
add_path(), add_star()

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> nx.add_cycle(G, [0, 1, 2, 3])
>>> nx.add_cycle(G, [10, 11, 12], weight=7)

4.1.13 networkx.classes.function.subgraph

subgraph(G, nbunch)
Returns the subgraph induced on nodes in nbunch.
Parameters
• G (graph) – A NetworkX graph
• nbunch (list, iterable) – A container of nodes that will be iterated through once (thus it
should be an iterator or be iterable). Each element of the container should be a valid node
type: any hashable type except None. If nbunch is None, return all edges data in the graph.
Nodes in nbunch that are not in the graph will be (quietly) ignored.

Notes

subgraph(G) calls G.subgraph()

4.1.14 networkx.classes.function.induced_subgraph

induced_subgraph(G, nbunch)
Returns a SubGraph view of G showing only nodes in nbunch.
The induced subgraph of a graph on a set of nodes N is the graph with nodes N and edges from G which have
both ends in N.
Parameters
• G (NetworkX Graph)
• nbunch (node, container of nodes or None (for all nodes))
Returns subgraph – A read-only view of the subgraph in G induced by the nodes. Changes to the
graph G will be reflected in the view.
Return type SubGraph View

Notes

To create a mutable subgraph with its own copies of nodes edges and attributes use subgraph.copy() or
Graph(subgraph)
For an inplace reduction of a graph to a subgraph you can remove nodes: G.remove_nodes_from(n in
G if n not in set(nbunch))

4.1. Graph 495


NetworkX Reference, Release 2.3rc1.dev20190222214247

If you are going to compute subgraphs of your subgraphs you could end up with a chain of views that can be
very slow once the chain has about 15 views in it. If they are all induced subgraphs, you can short-cut the chain
by making them all subgraphs of the original graph. The graph class method G.subgraph does this when G
is a subgraph. In contrast, this function allows you to choose to build chains or not, as you wish. The returned
subgraph is a view on G.

Examples

>>> import networkx as nx


>>> G = nx.path_graph(4) # or DiGraph, MultiGraph, MultiDiGraph, etc
>>> H = G.subgraph([0, 1, 2])
>>> list(H.edges)
[(0, 1), (1, 2)]

4.1.15 networkx.classes.function.restricted_view

restricted_view(G, nodes, edges)


Returns a view of G with hidden nodes and edges.
The resulting subgraph filters out node nodes and edges edges. Filtered out nodes also filter out any of their
edges.
Parameters
• G (NetworkX Graph)
• nodes (iterable) – An iterable of nodes. Nodes not present in G are ignored.
• edges (iterable) – An iterable of edges. Edges not present in G are ignored.
Returns subgraph – A read-only restricted view of G filtering out nodes and edges. Changes to G
are reflected in the view.
Return type SubGraph View

Notes

To create a mutable subgraph with its own copies of nodes edges and attributes use subgraph.copy() or
Graph(subgraph)
If you create a subgraph of a subgraph recursively you may end up with a chain of subgraph views. Such
chains can get quite slow for lengths near 15. To avoid long chains, try to make your subgraph based on the
original graph. We do not rule out chains programmatically so that odd cases like an edge_subgraph of a
restricted_view can be created.

Examples

>>> import networkx as nx


>>> G = nx.path_graph(5)
>>> H = nx.restricted_view(G, [0], [(1, 2), (3, 4)])
>>> list(H.nodes)
[1, 2, 3, 4]
>>> list(H.edges)
[(2, 3)]

496 Chapter 4. Functions


NetworkX Reference, Release 2.3rc1.dev20190222214247

4.1.16 networkx.classes.function.reverse_view

reverse_view(digraph)
Provide a reverse view of the digraph with edges reversed.
Identical to digraph.reverse(copy=False)

4.1.17 networkx.classes.function.edge_subgraph

edge_subgraph(G, edges)
Returns a view of the subgraph induced by the specified edges.
The induced subgraph contains each edge in edges and each node incident to any of those edges.
Parameters
• G (NetworkX Graph)
• edges (iterable) – An iterable of edges. Edges not present in G are ignored.
Returns subgraph – A read-only edge-induced subgraph of G. Changes to G are reflected in the
view.
Return type SubGraph View

Notes

To create a mutable subgraph with its own copies of nodes edges and attributes use subgraph.copy() or
Graph(subgraph)
If you create a subgraph of a subgraph recursively you can end up with a chain of subgraphs that becomes very
slow with about 15 nested subgraph views. Luckily the edge_subgraph filter nests nicely so you can use the
original graph as G in this function to avoid chains. We do not rule out chains programmatically so that odd
cases like an edge_subgraph of a restricted_view can be created.

Examples

>>> import networkx as nx


>>> G = nx.path_graph(5)
>>> H = G.edge_subgraph([(0, 1), (3, 4)])
>>> list(H.nodes)
[0, 1, 3, 4]
>>> list(H.edges)
[(0, 1), (3, 4)]

4.2 Nodes

nodes(G) Returns an iterator over the graph nodes.


number_of_nodes(G) Returns the number of nodes in the graph.
neighbors(G, n) Returns a list of nodes connected to node n.
all_neighbors(graph, node) Returns all of the neighbors of a node in the graph.
non_neighbors(graph, node) Returns the non-neighbors of the node in the graph.
Continued on next page

4.2. Nodes 497


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 2 – continued from previous page


common_neighbors(G, u, v) Returns the common neighbors of two nodes in a graph.

4.2.1 networkx.classes.function.nodes

nodes(G)
Returns an iterator over the graph nodes.

4.2.2 networkx.classes.function.number_of_nodes

number_of_nodes(G)
Returns the number of nodes in the graph.

4.2.3 networkx.classes.function.neighbors

neighbors(G, n)
Returns a list of nodes connected to node n.

4.2.4 networkx.classes.function.all_neighbors

all_neighbors(graph, node)
Returns all of the neighbors of a node in the graph.
If the graph is directed returns predecessors as well as successors.
Parameters
• graph (NetworkX graph) – Graph to find neighbors.
• node (node) – The node whose neighbors will be returned.
Returns neighbors – Iterator of neighbors
Return type iterator

4.2.5 networkx.classes.function.non_neighbors

non_neighbors(graph, node)
Returns the non-neighbors of the node in the graph.
Parameters
• graph (NetworkX graph) – Graph to find neighbors.
• node (node) – The node whose neighbors will be returned.
Returns non_neighbors – Iterator of nodes in the graph that are not neighbors of the node.
Return type iterator

498 Chapter 4. Functions


NetworkX Reference, Release 2.3rc1.dev20190222214247

4.2.6 networkx.classes.function.common_neighbors

common_neighbors(G, u, v)
Returns the common neighbors of two nodes in a graph.
Parameters
• G (graph) – A NetworkX undirected graph.
• u, v (nodes) – Nodes in the graph.
Returns cnbors – Iterator of common neighbors of u and v in the graph.
Return type iterator
Raises NetworkXError – If u or v is not a node in the graph.

Examples

>>> G = nx.complete_graph(5)
>>> sorted(nx.common_neighbors(G, 0, 1))
[2, 3, 4]

4.3 Edges

edges(G[, nbunch]) Returns an edge view of edges incident to nodes in


nbunch.
number_of_edges(G) Returns the number of edges in the graph.
density(G) Returns the density of a graph.
non_edges(graph) Returns the non-existent edges in the graph.

4.3.1 networkx.classes.function.edges

edges(G, nbunch=None)
Returns an edge view of edges incident to nodes in nbunch.
Return all edges if nbunch is unspecified or nbunch=None.
For digraphs, edges=out_edges

4.3.2 networkx.classes.function.number_of_edges

number_of_edges(G)
Returns the number of edges in the graph.

4.3.3 networkx.classes.function.non_edges

non_edges(graph)
Returns the non-existent edges in the graph.
Parameters graph (NetworkX graph.) – Graph to find non-existent edges.
Returns non_edges – Iterator of edges that are not in the graph.

4.3. Edges 499


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type iterator

4.4 Self loops

selfloop_edges(G[, data, keys, default]) Returns an iterator over selfloop edges.


number_of_selfloops(G) Returns the number of selfloop edges.
nodes_with_selfloops(G) Returns an iterator over nodes with self loops.

4.4.1 networkx.classes.function.selfloop_edges

selfloop_edges(G, data=False, keys=False, default=None)


Returns an iterator over selfloop edges.
A selfloop edge has the same node at both ends.
Parameters
• data (string or bool, optional (default=False)) – Return selfloop edges as two tuples (u,
v) (data=False) or three-tuples (u, v, datadict) (data=True) or three-tuples (u, v, datavalue)
(data=’attrname’)
• keys (bool, optional (default=False)) – If True, return edge keys with each edge.
• default (value, optional (default=None)) – Value used for edges that don’t have the re-
quested attribute. Only relevant if data is not True or False.
Returns edgeiter – An iterator over all selfloop edges.
Return type iterator over edge tuples
See also:
nodes_with_selfloops(), number_of_selfloops()

Examples

>>> G = nx.MultiGraph() # or Graph, DiGraph, MultiDiGraph, etc


>>> ekey = G.add_edge(1, 1)
>>> ekey = G.add_edge(1, 2)
>>> list(nx.selfloop_edges(G))
[(1, 1)]
>>> list(nx.selfloop_edges(G, data=True))
[(1, 1, {})]
>>> list(nx.selfloop_edges(G, keys=True))
[(1, 1, 0)]
>>> list(nx.selfloop_edges(G, keys=True, data=True))
[(1, 1, 0, {})]

4.4.2 networkx.classes.function.number_of_selfloops

number_of_selfloops(G)
Returns the number of selfloop edges.
A selfloop edge has the same node at both ends.

500 Chapter 4. Functions


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns nloops – The number of selfloops.


Return type int
See also:
nodes_with_selfloops(), selfloop_edges()

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_edge(1, 1)
>>> G.add_edge(1, 2)
>>> nx.number_of_selfloops(G)
1

4.4.3 networkx.classes.function.nodes_with_selfloops

nodes_with_selfloops(G)
Returns an iterator over nodes with self loops.
A node with a self loop has an edge with both ends adjacent to that node.
Returns nodelist – A iterator over nodes with self loops.
Return type iterator
See also:
selfloop_edges(), number_of_selfloops()

Examples

>>> G = nx.Graph() # or DiGraph, MultiGraph, MultiDiGraph, etc


>>> G.add_edge(1, 1)
>>> G.add_edge(1, 2)
>>> list(nx.nodes_with_selfloops(G))
[1]

4.5 Attributes

is_weighted(G[, edge, weight]) Returns True if G has weighted edges.


is_negatively_weighted(G[, edge, weight]) Returns True if G has negatively weighted edges.
set_node_attributes(G, values[, name]) Sets node attributes from a given value or dictionary of
values.
get_node_attributes(G, name) Get node attributes from graph
set_edge_attributes(G, values[, name]) Sets edge attributes from a given value or dictionary of
values.
get_edge_attributes(G, name) Get edge attributes from graph

4.5. Attributes 501


NetworkX Reference, Release 2.3rc1.dev20190222214247

4.5.1 networkx.classes.function.is_weighted

is_weighted(G, edge=None, weight=’weight’)


Returns True if G has weighted edges.
Parameters
• G (graph) – A NetworkX graph.
• edge (tuple, optional) – A 2-tuple specifying the only edge in G that will be tested. If None,
then every edge in G is tested.
• weight (string, optional) – The attribute name used to query for edge weights.
Returns A boolean signifying if G, or the specified edge, is weighted.
Return type bool
Raises NetworkXError – If the specified edge does not exist.

Examples

>>> G = nx.path_graph(4)
>>> nx.is_weighted(G)
False
>>> nx.is_weighted(G, (2, 3))
False

>>> G = nx.DiGraph()
>>> G.add_edge(1, 2, weight=1)
>>> nx.is_weighted(G)
True

4.5.2 networkx.classes.function.is_negatively_weighted

is_negatively_weighted(G, edge=None, weight=’weight’)


Returns True if G has negatively weighted edges.
Parameters
• G (graph) – A NetworkX graph.
• edge (tuple, optional) – A 2-tuple specifying the only edge in G that will be tested. If None,
then every edge in G is tested.
• weight (string, optional) – The attribute name used to query for edge weights.
Returns A boolean signifying if G, or the specified edge, is negatively weighted.
Return type bool
Raises NetworkXError – If the specified edge does not exist.

Examples

502 Chapter 4. Functions


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.Graph()
>>> G.add_edges_from([(1, 3), (2, 4), (2, 6)])
>>> G.add_edge(1, 2, weight=4)
>>> nx.is_negatively_weighted(G, (1, 2))
False
>>> G[2][4]['weight'] = -2
>>> nx.is_negatively_weighted(G)
True
>>> G = nx.DiGraph()
>>> edges = [('0', '3', 3), ('0', '1', -5), ('1', '0', -2)]
>>> G.add_weighted_edges_from(edges)
>>> nx.is_negatively_weighted(G)
True

4.5.3 networkx.classes.function.set_node_attributes

set_node_attributes(G, values, name=None)


Sets node attributes from a given value or dictionary of values.

Warning: The call order of arguments values and name switched between v1.x & v2.x.

Parameters
• G (NetworkX Graph)
• values (scalar value, dict-like) – What the node attribute should be set to. If values is not
a dictionary, then it is treated as a single attribute value that is then applied to every node in
G. This means that if you provide a mutable object, like a list, updates to that object will be
reflected in the node attribute for every node. The attribute name will be name.
If values is a dict or a dict of dict, it should be keyed by node to either an attribute value
or a dict of attribute key/value pairs used to update the node’s attributes.
• name (string (optional, default=None)) – Name of the node attribute to set if values is a
scalar.

Examples

After computing some property of the nodes of a graph, you may want to assign a node attribute to store the
value of that property for each node:

>>> G = nx.path_graph(3)
>>> bb = nx.betweenness_centrality(G)
>>> isinstance(bb, dict)
True
>>> nx.set_node_attributes(G, bb, 'betweenness')
>>> G.nodes[1]['betweenness']
1.0

If you provide a list as the second argument, updates to the list will be reflected in the node attribute for each
node:

4.5. Attributes 503


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.path_graph(3)
>>> labels = []
>>> nx.set_node_attributes(G, labels, 'labels')
>>> labels.append('foo')
>>> G.nodes[0]['labels']
['foo']
>>> G.nodes[1]['labels']
['foo']
>>> G.nodes[2]['labels']
['foo']

If you provide a dictionary of dictionaries as the second argument, the outer dictionary is assumed to be keyed
by node to an inner dictionary of node attributes for that node:

>>> G = nx.path_graph(3)
>>> attrs = {0: {'attr1': 20, 'attr2': 'nothing'}, 1: {'attr2': 3}}
>>> nx.set_node_attributes(G, attrs)
>>> G.nodes[0]['attr1']
20
>>> G.nodes[0]['attr2']
'nothing'
>>> G.nodes[1]['attr2']
3
>>> G.nodes[2]
{}

4.5.4 networkx.classes.function.get_node_attributes

get_node_attributes(G, name)
Get node attributes from graph
Parameters
• G (NetworkX Graph)
• name (string) – Attribute name
Returns
Return type Dictionary of attributes keyed by node.

Examples

>>> G = nx.Graph()
>>> G.add_nodes_from([1, 2, 3], color='red')
>>> color = nx.get_node_attributes(G, 'color')
>>> color[1]
'red'

4.5.5 networkx.classes.function.set_edge_attributes

set_edge_attributes(G, values, name=None)


Sets edge attributes from a given value or dictionary of values.

504 Chapter 4. Functions


NetworkX Reference, Release 2.3rc1.dev20190222214247

Warning: The call order of arguments values and name switched between v1.x & v2.x.

Parameters
• G (NetworkX Graph)
• values (scalar value, dict-like) – What the edge attribute should be set to. If values is not
a dictionary, then it is treated as a single attribute value that is then applied to every edge in
G. This means that if you provide a mutable object, like a list, updates to that object will be
reflected in the edge attribute for each edge. The attribute name will be name.
If values is a dict or a dict of dict, it should be keyed by edge tuple to either an attribute
value or a dict of attribute key/value pairs used to update the edge’s attributes. For multi-
graphs, the edge tuples must be of the form (u, v, key), where u and v are nodes and
key is the edge key. For non-multigraphs, the keys must be tuples of the form (u, v).
• name (string (optional, default=None)) – Name of the edge attribute to set if values is a
scalar.

Examples

After computing some property of the edges of a graph, you may want to assign a edge attribute to store the
value of that property for each edge:

>>> G = nx.path_graph(3)
>>> bb = nx.edge_betweenness_centrality(G, normalized=False)
>>> nx.set_edge_attributes(G, bb, 'betweenness')
>>> G.edges[1, 2]['betweenness']
2.0

If you provide a list as the second argument, updates to the list will be reflected in the edge attribute for each
edge:

>>> labels = []
>>> nx.set_edge_attributes(G, labels, 'labels')
>>> labels.append('foo')
>>> G.edges[0, 1]['labels']
['foo']
>>> G.edges[1, 2]['labels']
['foo']

If you provide a dictionary of dictionaries as the second argument, the entire dictionary will be used to update
edge attributes:

>>> G = nx.path_graph(3)
>>> attrs = {(0, 1): {'attr1': 20, 'attr2': 'nothing'},
... (1, 2): {'attr2': 3}}
>>> nx.set_edge_attributes(G, attrs)
>>> G[0][1]['attr1']
20
>>> G[0][1]['attr2']
'nothing'
>>> G[1][2]['attr2']
3

4.5. Attributes 505


NetworkX Reference, Release 2.3rc1.dev20190222214247

4.5.6 networkx.classes.function.get_edge_attributes

get_edge_attributes(G, name)
Get edge attributes from graph
Parameters
• G (NetworkX Graph)
• name (string) – Attribute name
Returns
• Dictionary of attributes keyed by edge. For (di)graphs, the keys are
• 2-tuples of the form ((u, v). For multi(di)graphs, the keys are 3-tuples of )
• the form ((u, v, key).)

Examples

>>> G = nx.Graph()
>>> nx.add_path(G, [1, 2, 3], color='red')
>>> color = nx.get_edge_attributes(G, 'color')
>>> color[(1, 2)]
'red'

4.6 Freezing graph structure

freeze(G) Modify graph to prevent further change by adding or


removing nodes or edges.
is_frozen(G) Returns True if graph is frozen.

4.6.1 networkx.classes.function.freeze

freeze(G)
Modify graph to prevent further change by adding or removing nodes or edges.
Node and edge data can still be modified.
Parameters G (graph) – A NetworkX graph

Examples

>>> G = nx.path_graph(4)
>>> G = nx.freeze(G)
>>> try:
... G.add_edge(4, 5)
... except nx.NetworkXError as e:
... print(str(e))
Frozen graph can't be modified

506 Chapter 4. Functions


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

To “unfreeze” a graph you must make a copy by creating a new graph object:

>>> graph = nx.path_graph(4)


>>> frozen_graph = nx.freeze(graph)
>>> unfrozen_graph = nx.Graph(frozen_graph)
>>> nx.is_frozen(unfrozen_graph)
False

See also:
is_frozen()

4.6.2 networkx.classes.function.is_frozen

is_frozen(G)
Returns True if graph is frozen.
Parameters G (graph) – A NetworkX graph
See also:
freeze()

4.6. Freezing graph structure 507


NetworkX Reference, Release 2.3rc1.dev20190222214247

508 Chapter 4. Functions


CHAPTER

FIVE

GRAPH GENERATORS

5.1 Atlas

Generators for the small graph atlas.

graph_atlas(i) Returns graph number i from the Graph Atlas.


graph_atlas_g() Returns the list of all graphs with up to seven nodes
named in the Graph Atlas.

5.1.1 networkx.generators.atlas.graph_atlas

graph_atlas(i)
Returns graph number i from the Graph Atlas.
For more information, see graph_atlas_g().
Parameters i (int) – The index of the graph from the atlas to get. The graph at index 0 is assumed
to be the null graph.
Returns A list of Graph objects, the one at index i corresponding to the graph i in the Graph Atlas.
Return type list
See also:
graph_atlas_g()

Notes

The time required by this function increases linearly with the argument i, since it reads a large file sequentially
in order to generate the graph1 .

References

5.1.2 networkx.generators.atlas.graph_atlas_g

graph_atlas_g()
Returns the list of all graphs with up to seven nodes named in the Graph Atlas.
The graphs are listed in increasing order by
1 Ronald C. Read and Robin J. Wilson, An Atlas of Graphs. Oxford University Press, 1998.

509
NetworkX Reference, Release 2.3rc1.dev20190222214247

1. number of nodes,
2. number of edges,
3. degree sequence (for example 111223 < 112222),
4. number of automorphisms,
in that order, with three exceptions as described in the Notes section below. This causes the list to correspond
with the index of the graphs in the Graph Atlas [atlas], with the first graph, G[0], being the null graph.
Returns A list of Graph objects, the one at index i corresponding to the graph i in the Graph Atlas.
Return type list
See also:
graph_atlas()

Notes

This function may be expensive in both time and space, since it reads a large file sequentially in order to populate
the list.
Although the NetworkX atlas functions match the order of graphs given in the “Atlas of Graphs” book, there
are (at least) three errors in the ordering described in the book. The following three pairs of nodes violate the
lexicographically nondecreasing sorted degree sequence rule:
• graphs 55 and 56 with degree sequences 001111 and 000112,
• graphs 1007 and 1008 with degree sequences 3333444 and 3333336,
• graphs 1012 and 1213 with degree sequences 1244555 and 1244456.

References

5.2 Classic

Generators for some classic graphs.


The typical graph generator is called as follows:

>>> G = nx.complete_graph(100)

returning the complete graph on n nodes labeled 0, .., 99 as a simple graph. Except for empty_graph, all the generators
in this module return a Graph class (i.e. a simple, undirected graph).

balanced_tree(r, h[, create_using]) Returns the perfectly balanced r-ary tree of height h.
barbell_graph(m1, m2[, create_using]) Returns the Barbell Graph: two complete graphs con-
nected by a path.
complete_graph(n[, create_using]) Return the complete graph K_n with n nodes.
complete_multipartite_graph(*subset_sizes) Returns the complete multipartite graph with the speci-
fied subset sizes.
circular_ladder_graph(n[, create_using]) Returns the circular ladder graph 𝐶𝐿𝑛 of length n.
circulant_graph(n, offsets[, create_using]) Generates the circulant graph 𝐶𝑖𝑛 (𝑥1 , 𝑥2 , ..., 𝑥𝑚 ) with
𝑛 vertices.
Continued on next page

510 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 2 – continued from previous page


cycle_graph(n[, create_using]) Returns the cycle graph 𝐶𝑛 of cyclically connected
nodes.
dorogovtsev_goltsev_mendes_graph(n[, Returns the hierarchically constructed Dorogovtsev-
. . . ]) Goltsev-Mendes graph.
empty_graph([n, create_using, default]) Returns the empty graph with n nodes and zero edges.
full_rary_tree(r, n[, create_using]) Creates a full r-ary tree of n vertices.
ladder_graph(n[, create_using]) Returns the Ladder graph of length n.
lollipop_graph(m, n[, create_using]) Returns the Lollipop Graph; K_m connected to P_n.
null_graph([create_using]) Returns the Null graph with no nodes or edges.
path_graph(n[, create_using]) Returns the Path graph P_n of linearly connected nodes.
star_graph(n[, create_using]) Return the star graph
trivial_graph([create_using]) Return the Trivial graph with one node (with label 0)
and no edges.
turan_graph(n, r) Return the Turan Graph
wheel_graph(n[, create_using]) Return the wheel graph

5.2.1 networkx.generators.classic.balanced_tree

balanced_tree(r, h, create_using=None)
Returns the perfectly balanced r-ary tree of height h.
Parameters
• r (int) – Branching factor of the tree; each node will have r children.
• h (int) – Height of the tree.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
Returns G – A balanced r-ary tree of height h.
Return type NetworkX graph

Notes

This is the rooted tree where all leaves are at distance h from the root. The root has degree r and all other
internal nodes have degree r + 1.
Node labels are integers, starting from zero.
A balanced tree is also known as a complete r-ary tree.

5.2.2 networkx.generators.classic.barbell_graph

barbell_graph(m1, m2, create_using=None)


Returns the Barbell Graph: two complete graphs connected by a path.
For 𝑚1 > 1 and 𝑚2 >= 0.
Two identical complete graphs 𝐾𝑚1 form the left and right bells, and are connected by a path 𝑃𝑚2 .
The 2*m1+m2 nodes are numbered 0, ..., m1-1 for the left barbell, m1, ..., m1+m2-1 for the
path, and m1+m2, ..., 2*m1+m2-1 for the right barbell.

5.2. Classic 511


NetworkX Reference, Release 2.3rc1.dev20190222214247

The 3 subgraphs are joined via the edges (m1-1, m1) and (m1+m2-1, m1+m2). If m2=0, this is merely
two complete graphs joined together.
This graph is an extremal example in David Aldous and Jim Fill’s e-text on Random Walks on Graphs.

5.2.3 networkx.generators.classic.complete_graph

complete_graph(n, create_using=None)
Return the complete graph K_n with n nodes.
Parameters
• n (int or iterable container of nodes) – If n is an integer, nodes are from range(n). If n is a
container of nodes, those nodes appear in the graph.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.

Examples

>>> G = nx.complete_graph(9)
>>> len(G)
9
>>> G.size()
36
>>> G = nx.complete_graph(range(11, 14))
>>> list(G.nodes())
[11, 12, 13]
>>> G = nx.complete_graph(4, nx.DiGraph())
>>> G.is_directed()
True

5.2.4 networkx.generators.classic.complete_multipartite_graph

complete_multipartite_graph(*subset_sizes)
Returns the complete multipartite graph with the specified subset sizes.
Parameters subset_sizes (tuple of integers or tuple of node iterables) – The arguments can either
all be integer number of nodes or they can all be iterables of nodes. If integers, they represent
the number of vertices in each subset of the multipartite graph. If iterables, each is used to create
the nodes for that subset. The length of subset_sizes is the number of subsets.
Returns
G – Returns the complete multipartite graph with the specified subsets.
For each node, the node attribute ‘subset’ is an integer indicating which subset contains the node.
Return type NetworkX Graph

Examples

Creating a complete tripartite graph, with subsets of one, two, and three vertices, respectively.

512 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> import networkx as nx


>>> G = nx.complete_multipartite_graph(1, 2, 3)
>>> [G.nodes[u]['subset'] for u in G]
[0, 1, 1, 2, 2, 2]
>>> list(G.edges(0))
[(0, 1), (0, 2), (0, 3), (0, 4), (0, 5)]
>>> list(G.edges(2))
[(2, 0), (2, 3), (2, 4), (2, 5)]
>>> list(G.edges(4))
[(4, 0), (4, 1), (4, 2)]

>>> G = nx.complete_multipartite_graph('a', 'bc', 'def')


>>> [G.nodes[u]['subset'] for u in sorted(G)]
[0, 1, 1, 2, 2, 2]

Notes

This function generalizes several other graph generator functions.


• If no subset sizes are given, this returns the null graph.
• If a single subset size n is given, this returns the empty graph on n nodes.
• If two subset sizes m and n are given, this returns the complete bipartite graph on m + n nodes.
• If subset sizes 1 and n are given, this returns the star graph on n + 1 nodes.
See also:
complete_bipartite_graph()

5.2.5 networkx.generators.classic.circular_ladder_graph

circular_ladder_graph(n, create_using=None)
Returns the circular ladder graph 𝐶𝐿𝑛 of length n.
𝐶𝐿𝑛 consists of two concentric n-cycles in which each of the n pairs of concentric nodes are joined by an edge.
Node labels are the integers 0 to n-1

5.2.6 networkx.generators.classic.circulant_graph

circulant_graph(n, offsets, create_using=None)


Generates the circulant graph 𝐶𝑖𝑛 (𝑥1 , 𝑥2 , ..., 𝑥𝑚 ) with 𝑛 vertices.
Returns
• The graph :math:‘Ci_n(x_1, . . . , x_m)‘ consisting of :math:‘n‘ vertices :math:‘0, . . . , n-1‘
such
• that the vertex with label :math:‘i‘ is connected to the vertices labelled :math:‘(i + x)‘
• and :math:‘(i - x)‘, for all :math:‘x‘ in :math:‘x_1‘ up to :math:‘x_m‘, with the indices taken
modulo :math:‘n‘.
Parameters
• n (integer) – The number of vertices the generated graph is to contain.

5.2. Classic 513


NetworkX Reference, Release 2.3rc1.dev20190222214247

• offsets (list of integers) – A list of vertex offsets, 𝑥1 up to 𝑥𝑚 , as described above.


• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.

Examples

Many well-known graph families are subfamilies of the circulant graphs; for example, to generate the cycle
graph on n points, we connect every vertex to every other at offset plus or minus one. For n = 10,

>>> import networkx


>>> G = networkx.generators.classic.circulant_graph(10, [1])
>>> edges = [
... (0, 9), (0, 1), (1, 2), (2, 3), (3, 4),
... (4, 5), (5, 6), (6, 7), (7, 8), (8, 9)]
...
>>> sorted(edges) == sorted(G.edges())
True

Similarly, we can generate the complete graph on 5 points with the set of offsets [1, 2]:

>>> G = networkx.generators.classic.circulant_graph(5, [1, 2])


>>> edges = [
... (0, 1), (0, 2), (0, 3), (0, 4), (1, 2),
... (1, 3), (1, 4), (2, 3), (2, 4), (3, 4)]
...
>>> sorted(edges) == sorted(G.edges())
True

5.2.7 networkx.generators.classic.cycle_graph

cycle_graph(n, create_using=None)
Returns the cycle graph 𝐶𝑛 of cyclically connected nodes.
𝐶𝑛 is a path with its two end-nodes connected.
Parameters
• n (int or iterable container of nodes) – If n is an integer, nodes are from range(n). If n is
a container of nodes, those nodes appear in the graph.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.

Notes

If create_using is directed, the direction is in increasing order.

5.2.8 networkx.generators.classic.dorogovtsev_goltsev_mendes_graph

dorogovtsev_goltsev_mendes_graph(n, create_using=None)
Returns the hierarchically constructed Dorogovtsev-Goltsev-Mendes graph.
n is the generation. See: arXiv:/cond-mat/0112143 by Dorogovtsev, Goltsev and Mendes.

514 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

5.2.9 networkx.generators.classic.empty_graph

empty_graph(n=0, create_using=None, default=<class ’networkx.classes.graph.Graph’>)


Returns the empty graph with n nodes and zero edges.
Parameters
• n (int or iterable container of nodes (default = 0)) – If n is an integer, nodes are from
range(n). If n is a container of nodes, those nodes appear in the graph.
• create_using (Graph Instance, Constructor or None) – Indicator of type of graph to return.
If a Graph-type instance, then clear and use it. If None, use the default constructor. If a
constructor, call it to create an empty graph.
• default (Graph constructor (optional, default = nx.Graph)) – The constructor to use if cre-
ate_using is None. If None, then nx.Graph is used. This is used when passing an unknown
create_using value through your home-grown function to empty_graph and you
want a default constructor other than nx.Graph.

Examples

>>> G = nx.empty_graph(10)
>>> G.number_of_nodes()
10
>>> G.number_of_edges()
0
>>> G = nx.empty_graph("ABC")
>>> G.number_of_nodes()
3
>>> sorted(G)
['A', 'B', 'C']

Notes

The variable create_using should be a Graph Constructor or a “graph”-like object. Constructors, e.g. nx.
Graph or nx.MultiGraph will be used to create the returned graph. “graph”-like objects will be cleared
(nodes and edges will be removed) and refitted as an empty “graph” with nodes specified in n. This capability is
useful for specifying the class-nature of the resulting empty “graph” (i.e. Graph, DiGraph, MyWeirdGraphClass,
etc.).
The variable create_using has three main uses: Firstly, the variable create_using can be used to create an empty
digraph, multigraph, etc. For example,

>>> n = 10
>>> G = nx.empty_graph(n, create_using=nx.DiGraph)

will create an empty digraph on n nodes.


Secondly, one can pass an existing graph (digraph, multigraph, etc.) via create_using. For example, if G is an
existing graph (resp. digraph, multigraph, etc.), then empty_graph(n, create_using=G) will empty G (i.e. delete
all nodes and edges using G.clear()) and then add n nodes and zero edges, and return the modified graph.
Thirdly, when constructing your home-grown graph creation function you can use empty_graph to construct the
graph by passing a user defined create_using to empty_graph. In this case, if you want the default constructor
to be other than nx.Graph, specify default.

5.2. Classic 515


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> def mygraph(n, create_using=None):


... G = nx.empty_graph(n, create_using, nx.MultiGraph)
... G.add_edges_from([(0, 1), (0, 1)])
... return G
>>> G = mygraph(3)
>>> G.is_multigraph()
True
>>> G = mygraph(3, nx.Graph)
>>> G.is_multigraph()
False

See also create_empty_copy(G).

5.2.10 networkx.generators.classic.full_rary_tree

full_rary_tree(r, n, create_using=None)
Creates a full r-ary tree of n vertices.
Sometimes called a k-ary, n-ary, or m-ary tree. “. . . all non-leaf vertices have exactly r children and all levels
are full except for some rightmost position of the bottom level (if a leaf at the bottom level is missing, then so
are all of the leaves to its right.”1
Parameters
• r (int) – branching factor of the tree
• n (int) – Number of nodes in the tree
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
Returns G – An r-ary tree with n nodes
Return type networkx Graph

References

5.2.11 networkx.generators.classic.ladder_graph

ladder_graph(n, create_using=None)
Returns the Ladder graph of length n.
This is two paths of n nodes, with each pair connected by a single edge.
Node labels are the integers 0 to 2*n - 1.

5.2.12 networkx.generators.classic.lollipop_graph

lollipop_graph(m, n, create_using=None)
Returns the Lollipop Graph; K_m connected to P_n.
This is the Barbell Graph without the right barbell.
Parameters
1 An introduction to data structures and algorithms, James Andrew Storer, Birkhauser Boston 2001, (page 225).

516 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

• m, n (int or iterable container of nodes (default = 0)) – If an integer, nodes are from
range(m) and range(m,m+n). If a container, the entries are the coordinate of the node.
The nodes for m appear in the complete graph 𝐾𝑚 and the nodes for n appear in the path
𝑃𝑛
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.

Notes

The 2 subgraphs are joined via an edge (m-1, m). If n=0, this is merely a complete graph.
(This graph is an extremal example in David Aldous and Jim Fill’s etext on Random Walks on Graphs.)

5.2.13 networkx.generators.classic.null_graph

null_graph(create_using=None)
Returns the Null graph with no nodes or edges.
See empty_graph for the use of create_using.

5.2.14 networkx.generators.classic.path_graph

path_graph(n, create_using=None)
Returns the Path graph P_n of linearly connected nodes.
Parameters
• n (int or iterable) – If an integer, node labels are 0 to n with center 0. If an iterable of nodes,
the center is the first.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.

5.2.15 networkx.generators.classic.star_graph

star_graph(n, create_using=None)
Return the star graph
The star graph consists of one center node connected to n outer nodes.
Parameters
• n (int or iterable) – If an integer, node labels are 0 to n with center 0. If an iterable of nodes,
the center is the first.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.

Notes

The graph has n+1 nodes for integer n. So star_graph(3) is the same as star_graph(range(4)).

5.2. Classic 517


NetworkX Reference, Release 2.3rc1.dev20190222214247

5.2.16 networkx.generators.classic.trivial_graph

trivial_graph(create_using=None)
Return the Trivial graph with one node (with label 0) and no edges.

5.2.17 networkx.generators.classic.turan_graph

turan_graph(n, r)
Return the Turan Graph
The Turan Graph is a complete multipartite graph on 𝑛 vertices with 𝑟 disjoint subsets. It is the graph with the
edges for any graph with 𝑛 vertices and 𝑟 disjoint subsets.
Given 𝑛 and 𝑟, we generate a complete multipartite graph with 𝑟 − (𝑛 mod 𝑟) partitions of size 𝑛/𝑟, rounded
down, and 𝑛 mod 𝑟 partitions of size 𝑛/𝑟 + 1, rounded down.
Parameters
• n (int) – The number of vertices.
• r (int) – The number of partitions. Must be less than or equal to n.

Notes

Must satisfy 1 <= 𝑟 <= 𝑛. The graph has (𝑟 − 1)(𝑛2 )/(2𝑟) edges, rounded down.

5.2.18 networkx.generators.classic.wheel_graph

wheel_graph(n, create_using=None)
Return the wheel graph
The wheel graph consists of a hub node connected to a cycle of (n-1) nodes.
Parameters
• n (int or iterable) – If an integer, node labels are 0 to n with center 0. If an iterable of nodes,
the center is the first.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
• Node labels are the integers 0 to n - 1.

5.3 Expanders

Provides explicit constructions of expander graphs.

margulis_gabber_galil_graph(n[, cre- Returns the Margulis-Gabber-Galil undirected Multi-


ate_using]) Graph on n^2 nodes.
chordal_cycle_graph(p[, create_using]) Returns the chordal cycle graph on p nodes.

518 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

5.3.1 networkx.generators.expanders.margulis_gabber_galil_graph

margulis_gabber_galil_graph(n, create_using=None)
Returns the Margulis-Gabber-Galil undirected MultiGraph on n^2 nodes.
The undirected MultiGraph is regular with degree 8. Nodes are integer pairs. The second-largest eigenvalue of
the adjacency matrix of the graph is at most 5 sqrt{2}, regardless of n.
Parameters
• n (int) – Determines the number of nodes in the graph: n^2.
• create_using (NetworkX graph constructor, optional (default MultiGraph)) – Graph type to
create. If graph instance, then cleared before populated.
Returns G – The constructed undirected multigraph.
Return type graph
Raises NetworkXError – If the graph is directed or not a multigraph.

5.3.2 networkx.generators.expanders.chordal_cycle_graph

chordal_cycle_graph(p, create_using=None)
Returns the chordal cycle graph on p nodes.
The returned graph is a cycle graph on p nodes with chords joining each vertex x to its inverse modulo p. This
graph is a (mildly explicit) 3-regular expander1 .
p must be a prime number.
Parameters
• p (a prime number) – The number of vertices in the graph. This also indicates where the
chordal edges in the cycle will be created.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
Returns G – The constructed undirected multigraph.
Return type graph
Raises NetworkXError – If create_using indicates directed or not a multigraph.

References

5.4 Lattice

Functions for generating grid graphs and lattices


The grid_2d_graph(), triangular_lattice_graph(), and hexagonal_lattice_graph() func-
tions correspond to the three regular tilings of the plane, the square, triangular, and hexagonal tilings, respectively.
grid_graph() and hypercube_graph() are similar for arbitrary dimensions. Useful relevant discussion can
be found about Triangular Tiling, and Square, Hex and Triangle Grids

1 Theorem 4.4.2 in A. Lubotzky. “Discrete groups, expanding graphs and invariant measures”, volume 125 of Progress in Mathematics.

Birkhäuser Verlag, Basel, 1994.

5.4. Lattice 519


NetworkX Reference, Release 2.3rc1.dev20190222214247

grid_2d_graph(m, n[, periodic, create_using]) Returns the two-dimensional grid graph.


grid_graph(dim[, periodic]) Returns the n-dimensional grid graph.
hexagonal_lattice_graph(m, n[, periodic, Returns an m by n hexagonal lattice graph.
. . . ])
hypercube_graph(n) Returns the n-dimensional hypercube graph.
triangular_lattice_graph(m, n[, periodic, Returns the 𝑚 by 𝑛 triangular lattice graph.
. . . ])

5.4.1 networkx.generators.lattice.grid_2d_graph

grid_2d_graph(m, n, periodic=False, create_using=None)


Returns the two-dimensional grid graph.
The grid graph has each node connected to its four nearest neighbors.
Parameters
• m, n (int or iterable container of nodes) – If an integer, nodes are from range(n). If a
container, elements become the coordinate of the nodes.
• periodic (bool (default: False)) – If this is True the nodes on the grid boundaries are joined
to the corresponding nodes on the opposite grid boundaries.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
Returns The (possibly periodic) grid graph of the specified dimensions.
Return type NetworkX graph

5.4.2 networkx.generators.lattice.grid_graph

grid_graph(dim, periodic=False)
Returns the n-dimensional grid graph.
The dimension n is the length of the list dim and the size in each dimension is the value of the corresponding
list element.
Parameters
• dim (list or tuple of numbers or iterables of nodes) – ‘dim’ is a tuple or list with, for each
dimension, either a number that is the size of that dimension or an iterable of nodes for that
dimension. The dimension of the grid_graph is the length of dim.
• periodic (bool) – If periodic is True the nodes on the grid boundaries are joined to
the corresponding nodes on the opposite grid boundaries.
Returns The (possibly periodic) grid graph of the specified dimensions.
Return type NetworkX graph

Examples

To produce a 2 by 3 by 4 grid graph, a graph on 24 nodes:

520 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> from networkx import grid_graph


>>> G = grid_graph(dim=[2, 3, 4])
>>> len(G)
24
>>> G = grid_graph(dim=[range(7, 9), range(3, 6)])
>>> len(G)
6

5.4.3 networkx.generators.lattice.hexagonal_lattice_graph

hexagonal_lattice_graph(m, n, periodic=False, with_positions=True, create_using=None)


Returns an m by n hexagonal lattice graph.
The hexagonal lattice graph is a graph whose nodes and edges are the hexagonal tiling of the plane.
The returned graph will have m rows and n columns of hexagons. Odd numbered columns are shifted up relative
to even numbered columns.
Positions of nodes are computed by default or with_positions is True. Node positions creating the
standard embedding in the plane with sidelength 1 and are stored in the node attribute ‘pos’. pos = nx.
get_node_attributes(G, 'pos') creates a dict ready for drawing.
Parameters
• m (int) – The number of rows of hexagons in the lattice.
• n (int) – The number of columns of hexagons in the lattice.
• periodic (bool) – Whether to make a periodic grid by joining the boundary vertices. For
this to work n must be odd and both n > 1 and m > 1. The periodic connections create
another row and column of hexagons so these graphs have fewer nodes as boundary nodes
are identified.
• with_positions (bool (default: True)) – Store the coordinates of each node in the graph node
attribute ‘pos’. The coordinates provide a lattice with vertical columns of hexagons offset
to interleave and cover the plane. Periodic positions shift the nodes vertically in a nonlinear
way so the edges don’t overlap so much.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated. If graph is directed, edges will
point up or right.
Returns The m by n hexagonal lattice graph.
Return type NetworkX graph

5.4.4 networkx.generators.lattice.hypercube_graph

hypercube_graph(n)
Returns the n-dimensional hypercube graph.
The nodes are the integers between 0 and 2 ** n - 1, inclusive.
For more information on the hypercube graph, see the Wikipedia article Hypercube graph.
Parameters n (int) – The dimension of the hypercube. The number of nodes in the graph will be 2
** n.
Returns The hypercube graph of dimension n.

5.4. Lattice 521


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type NetworkX graph

5.4.5 networkx.generators.lattice.triangular_lattice_graph

triangular_lattice_graph(m, n, periodic=False, with_positions=True, create_using=None)


Returns the 𝑚 by 𝑛 triangular lattice graph.
The triangular lattice graph is a two-dimensional grid graph in which each square unit has a diagonal edge (each
grid unit has a chord).
The returned graph has 𝑚 rows and 𝑛 columns of triangles. Rows and columns include both triangles pointing
up and down. Rows form a strip of constant height. Columns form a series of diamond shapes, staggered with
the columns on either side. Another way to state the size is that the nodes form a grid of m+1 rows and (n +
1) // 2 columns. The odd row nodes are shifted horizontally relative to the even rows.
Directed graph types have edges pointed up or right.
Positions of nodes are computed by default or with_positions is True. The position of each node
(embedded in a euclidean plane)
√︀ is stored in the graph using equilateral triangles with sidelength 1. The height
between rows of nodes is thus (3)/2. Nodes lie in the first quadrant with the node (0, 0) at the origin.
Parameters
• m (int) – The number of rows in the lattice.
• n (int) – The number of columns in the lattice.
• periodic (bool (default: False)) – If True, join the boundary vertices of the grid using pe-
riodic boundary conditions. The join between boundaries is the final row and column of
triangles. This means there is one row and one column fewer nodes for the periodic lattice.
Periodic lattices require m >= 3, n >= 5 and are allowed but misaligned if m or n are odd
• with_positions (bool (default: True)) – Store the coordinates of each node in the graph
node attribute ‘pos’. The coordinates provide a lattice with equilateral triangles. Periodic
positions shift the nodes vertically in a nonlinear way so the edges don’t overlap so much.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
Returns The m by n triangular lattice graph.
Return type NetworkX graph

5.5 Small

Various small and named graphs, together with some compact generators.

make_small_graph(graph_description[, . . . ]) Return the small graph described by graph_description.


LCF_graph(n, shift_list, repeats[, create_using]) Return the cubic graph specified in LCF notation.
bull_graph([create_using]) Returns the Bull graph.
chvatal_graph([create_using]) Returns the Chvátal graph.
cubical_graph([create_using]) Returns the 3-regular Platonic Cubical graph.
desargues_graph([create_using]) Return the Desargues graph.
diamond_graph([create_using]) Returns the Diamond graph.
dodecahedral_graph([create_using]) Return the Platonic Dodecahedral graph.
frucht_graph([create_using]) Returns the Frucht Graph.
Continued on next page

522 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 5 – continued from previous page


heawood_graph([create_using]) Return the Heawood graph, a (3,6) cage.
hoffman_singleton_graph() Return the Hoffman-Singleton Graph.
house_graph([create_using]) Returns the House graph (square with triangle on top).
house_x_graph([create_using]) Returns the House graph with a cross inside the house
square.
icosahedral_graph([create_using]) Returns the Platonic Icosahedral graph.
krackhardt_kite_graph([create_using]) Return the Krackhardt Kite Social Network.
moebius_kantor_graph([create_using]) Returns the Moebius-Kantor graph.
octahedral_graph([create_using]) Returns the Platonic Octahedral graph.
pappus_graph() Return the Pappus graph.
petersen_graph([create_using]) Returns the Petersen graph.
sedgewick_maze_graph([create_using]) Return a small maze with a cycle.
tetrahedral_graph([create_using]) Return the 3-regular Platonic Tetrahedral graph.
truncated_cube_graph([create_using]) Returns the skeleton of the truncated cube.
truncated_tetrahedron_graph([create_using]) Returns the skeleton of the truncated Platonic tetrahe-
dron.
tutte_graph([create_using]) Returns the Tutte graph.

5.5.1 networkx.generators.small.make_small_graph

make_small_graph(graph_description, create_using=None)
Return the small graph described by graph_description.
graph_description is a list of the form [ltype,name,n,xlist]
Here ltype is one of “adjacencylist” or “edgelist”, name is the name of the graph and n the number of nodes.
This constructs a graph of n nodes with integer labels 0,..,n-1.
If ltype=”adjacencylist” then xlist is an adjacency list with exactly n entries, in with the j’th entry (which can be
empty) specifies the nodes connected to vertex j. e.g. the “square” graph C_4 can be obtained by

>>> G=nx.make_small_graph(["adjacencylist","C_4",4,[[2,4],[1,3],[2,4],[1,3]]])

or, since we do not need to add edges twice,

>>> G=nx.make_small_graph(["adjacencylist","C_4",4,[[2,4],[3],[4],[]]])

If ltype=”edgelist” then xlist is an edge list written as [[v1,w2],[v2,w2],. . . ,[vk,wk]], where vj and wj integers
in the range 1,..,n e.g. the “square” graph C_4 can be obtained by

>>> G=nx.make_small_graph(["edgelist","C_4",4,[[1,2],[3,4],[2,3],[4,1]]])

Use the create_using argument to choose the graph class/type.

5.5.2 networkx.generators.small.LCF_graph

LCF_graph(n, shift_list, repeats, create_using=None)


Return the cubic graph specified in LCF notation.
LCF notation (LCF=Lederberg-Coxeter-Fruchte) is a compressed notation used in the generation of various
cubic Hamiltonian graphs of high symmetry. See, for example, dodecahedral_graph, desargues_graph, hea-
wood_graph and pappus_graph below.

5.5. Small 523


NetworkX Reference, Release 2.3rc1.dev20190222214247

n (number of nodes) The starting graph is the n-cycle with nodes 0,. . . ,n-1. (The null graph is returned if n <
0.)
shift_list = [s1,s2,..,sk], a list of integer shifts mod n,
repeats integer specifying the number of times that shifts in shift_list are successively applied to each v_current
in the n-cycle to generate an edge between v_current and v_current+shift mod n.
For v1 cycling through the n-cycle a total of k*repeats with shift cycling through shiftlist repeats times connect
v1 with v1+shift mod n
The utility graph 𝐾3,3

>>> G = nx.LCF_graph(6, [3, -3], 3)

The Heawood graph

>>> G = nx.LCF_graph(14, [5, -5], 7)

See http://mathworld.wolfram.com/LCFNotation.html for a description and references.

5.5.3 networkx.generators.small.bull_graph

bull_graph(create_using=None)
Returns the Bull graph.

5.5.4 networkx.generators.small.chvatal_graph

chvatal_graph(create_using=None)
Returns the Chvátal graph.

5.5.5 networkx.generators.small.cubical_graph

cubical_graph(create_using=None)
Returns the 3-regular Platonic Cubical graph.

5.5.6 networkx.generators.small.desargues_graph

desargues_graph(create_using=None)
Return the Desargues graph.

5.5.7 networkx.generators.small.diamond_graph

diamond_graph(create_using=None)
Returns the Diamond graph.

5.5.8 networkx.generators.small.dodecahedral_graph

dodecahedral_graph(create_using=None)
Return the Platonic Dodecahedral graph.

524 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

5.5.9 networkx.generators.small.frucht_graph

frucht_graph(create_using=None)
Returns the Frucht Graph.
The Frucht Graph is the smallest cubical graph whose automorphism group consists only of the identity element.

5.5.10 networkx.generators.small.heawood_graph

heawood_graph(create_using=None)
Return the Heawood graph, a (3,6) cage.

5.5.11 networkx.generators.small.hoffman_singleton_graph

hoffman_singleton_graph()
Return the Hoffman-Singleton Graph.

5.5.12 networkx.generators.small.house_graph

house_graph(create_using=None)
Returns the House graph (square with triangle on top).

5.5.13 networkx.generators.small.house_x_graph

house_x_graph(create_using=None)
Returns the House graph with a cross inside the house square.

5.5.14 networkx.generators.small.icosahedral_graph

icosahedral_graph(create_using=None)
Returns the Platonic Icosahedral graph.

5.5.15 networkx.generators.small.krackhardt_kite_graph

krackhardt_kite_graph(create_using=None)
Return the Krackhardt Kite Social Network.
A 10 actor social network introduced by David Krackhardt to illustrate: degree, betweenness, centrality, close-
ness, etc. The traditional labeling is: Andre=1, Beverley=2, Carol=3, Diane=4, Ed=5, Fernando=6, Garth=7,
Heather=8, Ike=9, Jane=10.

5.5.16 networkx.generators.small.moebius_kantor_graph

moebius_kantor_graph(create_using=None)
Returns the Moebius-Kantor graph.

5.5. Small 525


NetworkX Reference, Release 2.3rc1.dev20190222214247

5.5.17 networkx.generators.small.octahedral_graph

octahedral_graph(create_using=None)
Returns the Platonic Octahedral graph.

5.5.18 networkx.generators.small.pappus_graph

pappus_graph()
Return the Pappus graph.

5.5.19 networkx.generators.small.petersen_graph

petersen_graph(create_using=None)
Returns the Petersen graph.

5.5.20 networkx.generators.small.sedgewick_maze_graph

sedgewick_maze_graph(create_using=None)
Return a small maze with a cycle.
This is the maze used in Sedgewick,3rd Edition, Part 5, Graph Algorithms, Chapter 18, e.g. Figure 18.2 and
following. Nodes are numbered 0,..,7

5.5.21 networkx.generators.small.tetrahedral_graph

tetrahedral_graph(create_using=None)
Return the 3-regular Platonic Tetrahedral graph.

5.5.22 networkx.generators.small.truncated_cube_graph

truncated_cube_graph(create_using=None)
Returns the skeleton of the truncated cube.

5.5.23 networkx.generators.small.truncated_tetrahedron_graph

truncated_tetrahedron_graph(create_using=None)
Returns the skeleton of the truncated Platonic tetrahedron.

5.5.24 networkx.generators.small.tutte_graph

tutte_graph(create_using=None)
Returns the Tutte graph.

5.6 Random Graphs

Generators for random graphs.

526 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

fast_gnp_random_graph(n, p[, seed, directed]) Returns a 𝐺𝑛,𝑝 random graph, also known as an Erdős-
Rényi graph or a binomial graph.
gnp_random_graph(n, p[, seed, directed]) Returns a 𝐺𝑛,𝑝 random graph, also known as an Erdős-
Rényi graph or a binomial graph.
dense_gnm_random_graph(n, m[, seed]) Returns a 𝐺𝑛,𝑚 random graph.
gnm_random_graph(n, m[, seed, directed]) Returns a 𝐺𝑛,𝑚 random graph.
erdos_renyi_graph(n, p[, seed, directed]) Returns a 𝐺𝑛,𝑝 random graph, also known as an Erdős-
Rényi graph or a binomial graph.
binomial_graph(n, p[, seed, directed]) Returns a 𝐺𝑛,𝑝 random graph, also known as an Erdős-
Rényi graph or a binomial graph.
newman_watts_strogatz_graph(n, k, p[, seed]) Returns a Newman–Watts–Strogatz small-world graph.
watts_strogatz_graph(n, k, p[, seed]) Returns a Watts–Strogatz small-world graph.
connected_watts_strogatz_graph(n, k, p[, Returns a connected Watts–Strogatz small-world graph.
. . . ])
random_regular_graph(d, n[, seed]) Returns a random 𝑑-regular graph on 𝑛 nodes.
barabasi_albert_graph(n, m[, seed]) Returns a random graph according to the
Barabási–Albert preferential attachment model.
dual_barabasi_albert_graph(n, m1, m2, p[, Returns a random graph according to the dual
seed]) Barabási–Albert preferential attachment model.
extended_barabasi_albert_graph(n, m, p, Returns an extended Barabási–Albert model graph.
q[, . . . ])
powerlaw_cluster_graph(n, m, p[, seed]) Holme and Kim algorithm for growing graphs with
powerlaw degree distribution and approximate average
clustering.
random_kernel_graph(n, kernel_integral[, . . . ]) Returns an random graph based on the specified kernel.
random_lobster(n, p1, p2[, seed]) Returns a random lobster graph.
random_shell_graph(constructor[, seed]) Returns a random shell graph for the constructor given.
random_powerlaw_tree(n[, gamma, seed, tries]) Returns a tree with a power law degree distribution.
random_powerlaw_tree_sequence(n[, gamma, Returns a degree sequence for a tree with a power law
. . . ]) distribution.
random_kernel_graph(n, kernel_integral[, . . . ]) Returns an random graph based on the specified kernel.

5.6.1 networkx.generators.random_graphs.fast_gnp_random_graph

fast_gnp_random_graph(n, p, seed=None, directed=False)


Returns a 𝐺𝑛,𝑝 random graph, also known as an Erdős-Rényi graph or a binomial graph.
Parameters
• n (int) – The number of nodes.
• p (float) – Probability for edge creation.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• directed (bool, optional (default=False)) – If True, this function returns a directed graph.

Notes

The 𝐺𝑛,𝑝 graph algorithm chooses each of the [𝑛(𝑛 − 1)]/2 (undirected) or 𝑛(𝑛 − 1) (directed) possible edges
with probability 𝑝.

5.6. Random Graphs 527


NetworkX Reference, Release 2.3rc1.dev20190222214247

This algorithm1 runs in 𝑂(𝑛 + 𝑚) time, where m is the expected number of edges, which equals 𝑝𝑛(𝑛 − 1)/2.
This should be faster than gnp_random_graph() when 𝑝 is small and the expected number of edges is small
(that is, the graph is sparse).
See also:
gnp_random_graph()

References

5.6.2 networkx.generators.random_graphs.gnp_random_graph

gnp_random_graph(n, p, seed=None, directed=False)


Returns a 𝐺𝑛,𝑝 random graph, also known as an Erdős-Rényi graph or a binomial graph.
The 𝐺𝑛,𝑝 model chooses each of the possible edges with probability 𝑝.
The functions binomial_graph() and erdos_renyi_graph() are aliases of this function.
Parameters
• n (int) – The number of nodes.
• p (float) – Probability for edge creation.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• directed (bool, optional (default=False)) – If True, this function returns a directed graph.
See also:
fast_gnp_random_graph()

Notes

This algorithm2 runs in 𝑂(𝑛2 ) time. For sparse graphs (that is, for small values of 𝑝),
fast_gnp_random_graph() is a faster algorithm.

References

5.6.3 networkx.generators.random_graphs.dense_gnm_random_graph

dense_gnm_random_graph(n, m, seed=None)
Returns a 𝐺𝑛,𝑚 random graph.
In the 𝐺𝑛,𝑚 model, a graph is chosen uniformly at random from the set of all graphs with 𝑛 nodes and 𝑚 edges.
This algorithm should be faster than gnm_random_graph() for dense graphs.
Parameters
• n (int) – The number of nodes.
• m (int) – The number of edges.
1 Vladimir Batagelj and Ulrik Brandes, “Efficient generation of large random networks”, Phys. Rev. E, 71, 036113, 2005.
2

E. N. Gilbert, Random Graphs, Ann. Math. Stat., 30, 1141 (1959).

528 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

• seed (integer, random_state, or None (default)) – Indicator of random number generation


state. See Randomness.
See also:
gnm_random_graph()

Notes

Algorithm by Keith M. Briggs Mar 31, 2006. Inspired by Knuth’s Algorithm S (Selection sampling technique),
in section 3.4.2 of1 .

References

5.6.4 networkx.generators.random_graphs.gnm_random_graph

gnm_random_graph(n, m, seed=None, directed=False)


Returns a 𝐺𝑛,𝑚 random graph.
In the 𝐺𝑛,𝑚 model, a graph is chosen uniformly at random from the set of all graphs with 𝑛 nodes and 𝑚 edges.
This algorithm should be faster than dense_gnm_random_graph() for sparse graphs.
Parameters
• n (int) – The number of nodes.
• m (int) – The number of edges.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• directed (bool, optional (default=False)) – If True return a directed graph
See also:
dense_gnm_random_graph()

5.6.5 networkx.generators.random_graphs.erdos_renyi_graph

erdos_renyi_graph(n, p, seed=None, directed=False)


Returns a 𝐺𝑛,𝑝 random graph, also known as an Erdős-Rényi graph or a binomial graph.
The 𝐺𝑛,𝑝 model chooses each of the possible edges with probability 𝑝.
The functions binomial_graph() and erdos_renyi_graph() are aliases of this function.
Parameters
• n (int) – The number of nodes.
• p (float) – Probability for edge creation.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• directed (bool, optional (default=False)) – If True, this function returns a directed graph.
1 Donald E. Knuth, The Art of Computer Programming, Volume 2/Seminumerical algorithms, Third Edition, Addison-Wesley, 1997.

5.6. Random Graphs 529


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
fast_gnp_random_graph()

Notes

This algorithm2 runs in 𝑂(𝑛2 ) time. For sparse graphs (that is, for small values of 𝑝),
fast_gnp_random_graph() is a faster algorithm.

References

5.6.6 networkx.generators.random_graphs.binomial_graph

binomial_graph(n, p, seed=None, directed=False)


Returns a 𝐺𝑛,𝑝 random graph, also known as an Erdős-Rényi graph or a binomial graph.
The 𝐺𝑛,𝑝 model chooses each of the possible edges with probability 𝑝.
The functions binomial_graph() and erdos_renyi_graph() are aliases of this function.
Parameters
• n (int) – The number of nodes.
• p (float) – Probability for edge creation.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• directed (bool, optional (default=False)) – If True, this function returns a directed graph.
See also:
fast_gnp_random_graph()

Notes

This algorithm2 runs in 𝑂(𝑛2 ) time. For sparse graphs (that is, for small values of 𝑝),
fast_gnp_random_graph() is a faster algorithm.

References

5.6.7 networkx.generators.random_graphs.newman_watts_strogatz_graph

newman_watts_strogatz_graph(n, k, p, seed=None)
Returns a Newman–Watts–Strogatz small-world graph.
Parameters
• n (int) – The number of nodes.
• k (int) – Each node is joined with its k nearest neighbors in a ring topology.
2

E. N. Gilbert, Random Graphs, Ann. Math. Stat., 30, 1141 (1959).

E. N. Gilbert, Random Graphs, Ann. Math. Stat., 30, 1141 (1959).

530 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

• p (float) – The probability of adding a new edge for each edge.


• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

Notes

First create a ring over 𝑛 nodes1 . Then each node in the ring is connected with its 𝑘 nearest neighbors (or 𝑘 − 1
neighbors if 𝑘 is odd). Then shortcuts are created by adding new edges as follows: for each edge (𝑢, 𝑣) in the
underlying “𝑛-ring with 𝑘 nearest neighbors” with probability 𝑝 add a new edge (𝑢, 𝑤) with randomly-chosen
existing node 𝑤. In contrast with watts_strogatz_graph(), no edges are removed.
See also:
watts_strogatz_graph()

References

5.6.8 networkx.generators.random_graphs.watts_strogatz_graph

watts_strogatz_graph(n, k, p, seed=None)
Returns a Watts–Strogatz small-world graph.
Parameters
• n (int) – The number of nodes
• k (int) – Each node is joined with its k nearest neighbors in a ring topology.
• p (float) – The probability of rewiring each edge
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
See also:
newman_watts_strogatz_graph(), connected_watts_strogatz_graph()

Notes

First create a ring over 𝑛 nodes1 . Then each node in the ring is joined to its 𝑘 nearest neighbors (or 𝑘 − 1
neighbors if 𝑘 is odd). Then shortcuts are created by replacing some edges as follows: for each edge (𝑢, 𝑣)
in the underlying “𝑛-ring with 𝑘 nearest neighbors” with probability 𝑝 replace it with a new edge (𝑢, 𝑤) with
uniformly random choice of existing node 𝑤.
In contrast with newman_watts_strogatz_graph(), the random rewiring does not in-
crease the number of edges. The rewired graph is not guaranteed to be connected as in
connected_watts_strogatz_graph().
1 M. E. J. Newman and D. J. Watts, Renormalization group analysis of the small-world network model, Physics Letters A, 263, 341, 1999.

https://doi.org/10.1016/S0375-9601(99)00757-4
1 Duncan J. Watts and Steven H. Strogatz, Collective dynamics of small-world networks, Nature, 393, pp. 440–442, 1998.

5.6. Random Graphs 531


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

5.6.9 networkx.generators.random_graphs.connected_watts_strogatz_graph

connected_watts_strogatz_graph(n, k, p, tries=100, seed=None)


Returns a connected Watts–Strogatz small-world graph.
Attempts to generate a connected graph by repeated generation of Watts–Strogatz small-world graphs. An
exception is raised if the maximum number of tries is exceeded.
Parameters
• n (int) – The number of nodes
• k (int) – Each node is joined with its k nearest neighbors in a ring topology.
• p (float) – The probability of rewiring each edge
• tries (int) – Number of attempts to generate a connected graph.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

Notes

First create a ring over 𝑛 nodes1 . Then each node in the ring is joined to its 𝑘 nearest neighbors (or 𝑘 − 1
neighbors if 𝑘 is odd). Then shortcuts are created by replacing some edges as follows: for each edge (𝑢, 𝑣)
in the underlying “𝑛-ring with 𝑘 nearest neighbors” with probability 𝑝 replace it with a new edge (𝑢, 𝑤) with
uniformly random choice of existing node 𝑤. The entire process is repeated until a connected graph results.
See also:
newman_watts_strogatz_graph(), watts_strogatz_graph()

References

5.6.10 networkx.generators.random_graphs.random_regular_graph

random_regular_graph(d, n, seed=None)
Returns a random 𝑑-regular graph on 𝑛 nodes.
The resulting graph has no self-loops or parallel edges.
Parameters
• d (int) – The degree of each node.
• n (integer) – The number of nodes. The value of 𝑛 × 𝑑 must be even.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
1 Duncan J. Watts and Steven H. Strogatz, Collective dynamics of small-world networks, Nature, 393, pp. 440–442, 1998.

532 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The nodes are numbered from 0 to 𝑛 − 1.


Kim and Vu’s paper2 shows that this algorithm samples in an asymptotically uniform way from the space of
random graphs when 𝑑 = 𝑂(𝑛1/3−𝜖 ).
Raises NetworkXError – If 𝑛 × 𝑑 is odd or 𝑑 is greater than or equal to 𝑛.

References

5.6.11 networkx.generators.random_graphs.barabasi_albert_graph

barabasi_albert_graph(n, m, seed=None)
Returns a random graph according to the Barabási–Albert preferential attachment model.
A graph of 𝑛 nodes is grown by attaching new nodes each with 𝑚 edges that are preferentially attached to
existing nodes with high degree.
Parameters
• n (int) – Number of nodes
• m (int) – Number of edges to attach from a new node to existing nodes
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G
Return type Graph
Raises NetworkXError – If m does not satisfy 1 <= m < n.

References

5.6.12 networkx.generators.random_graphs.dual_barabasi_albert_graph

dual_barabasi_albert_graph(n, m1, m2, p, seed=None)


Returns a random graph according to the dual Barabási–Albert preferential attachment model.
A graph of 𝑛 nodes is grown by attaching new nodes each with either 𝑚1 edges (with probability 𝑝) or 𝑚2 edges
(with probability 1 − 𝑝) that are preferentially attached to existing nodes with high degree.
Parameters
• n (int) – Number of nodes
• m1 (int) – Number of edges to attach from a new node to existing nodes with probability 𝑝
• m2 (int) – Number of edges to attach from a new node to existing nodes with probability
1−𝑝
• p (float) – The probability of attaching 𝑚1 edges (as opposed to 𝑚2 edges)
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G
2Jeong Han Kim and Van H. Vu, Generating random regular graphs, Proceedings of the thirty-fifth ACM symposium on Theory of computing,
San Diego, CA, USA, pp 213–222, 2003. http://portal.acm.org/citation.cfm?id=780542.780576

5.6. Random Graphs 533


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type Graph


Raises NetworkXError – If m1 and m2 do not satisfy 1 <= m1,m2 < n or p does not satisfy
0 <= p <= 1.

References

5.6.13 networkx.generators.random_graphs.extended_barabasi_albert_graph

extended_barabasi_albert_graph(n, m, p, q, seed=None)
Returns an extended Barabási–Albert model graph.
An extended Barabási–Albert model graph is a random graph constructed using preferential attachment. The
extended model allows new edges, rewired edges or new nodes. Based on the probabilities 𝑝 and 𝑞 with 𝑝+𝑞 < 1,
the growing behavior of the graph is determined as:
1) With 𝑝 probability, 𝑚 new edges are added to the graph, starting from randomly chosen existing nodes and
attached preferentially at the other end.
2) With 𝑞 probability, 𝑚 existing edges are rewired by randomly choosing an edge and rewiring one end to a
preferentially chosen node.
3) With (1 − 𝑝 − 𝑞) probability, 𝑚 new nodes are added to the graph with edges attached preferentially.
When 𝑝 = 𝑞 = 0, the model behaves just like the Barabási–Alber mo
Parameters
• n (int) – Number of nodes
• m (int) – Number of edges with which a new node attaches to existing nodes
• p (float) – Probability value for adding an edge between existing nodes. p + q < 1
• q (float) – Probability value of rewiring of existing edges. p + q < 1
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G
Return type Graph
Raises NetworkXError – If m does not satisfy 1 <= m < n or 1 >= p + q

References

5.6.14 networkx.generators.random_graphs.powerlaw_cluster_graph

powerlaw_cluster_graph(n, m, p, seed=None)
Holme and Kim algorithm for growing graphs with powerlaw degree distribution and approximate average
clustering.
Parameters
• n (int) – the number of nodes
• m (int) – the number of random edges to add for each new node
• p (float,) – Probability of adding a triangle after adding a random edge

534 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

• seed (integer, random_state, or None (default)) – Indicator of random number generation


state. See Randomness.

Notes

The average clustering has a hard time getting above a certain cutoff that depends on m. This cutoff is often
quite low. The transitivity (fraction of triangles to possible triangles) seems to decrease with network size.
It is essentially the Barabási–Albert (BA) growth model with an extra step that each random edge is followed
by a chance of making an edge to one of its neighbors too (and thus a triangle).
This algorithm improves on BA in the sense that it enables a higher average clustering to be attained if desired.
It seems possible to have a disconnected graph with this algorithm since the initial m nodes may not be all linked
to a new node on the first iteration like the BA model.
Raises NetworkXError – If m does not satisfy 1 <= m <= n or p does not satisfy 0 <= p
<= 1.

References

5.6.15 networkx.generators.random_graphs.random_kernel_graph

random_kernel_graph(n, kernel_integral, kernel_root=None, seed=None)


Returns an random graph based on the specified kernel.
The algorithm chooses each of the [𝑛(𝑛 − 1)]/2 possible edges with probability specified by a kernel 𝜅(𝑥, 𝑦)1 .
The kernel 𝜅(𝑥, 𝑦) must be a symmetric (in 𝑥, 𝑦), non-negative, bounded function.
Parameters
• n (int) – The number of nodes
• kernal_integral (function) – Function that returns the definite integral of the kernel 𝜅(𝑥, 𝑦),
∫︀ 𝑏
𝐹 (𝑦, 𝑎, 𝑏) := 𝑎 𝜅(𝑥, 𝑦)𝑑𝑥
• kernel_root (function (optional)) – Function that returns the root 𝑏 of the equation
𝐹 (𝑦, 𝑎, 𝑏) = 𝑟. If None, the root is found using scipy.optimize.brentq() (this
requires SciPy).
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

Notes

The kernel is specified through its definite integral which must be provided as one of the arguments. If the
integral and root of the kernel integral can be found in 𝑂(1) time then this algorithm runs in time 𝑂(𝑛 + 𝑚)
where m is the expected number of edges2 .
The nodes are set to integers from 0 to 𝑛 − 1.
1 Bollobás, Béla, Janson, S. and Riordan, O. “The phase transition in inhomogeneous random graphs”, Random Structures Algorithms, 31,

3–122, 2007.
2 Hagberg A, Lemons N (2015), “Fast Generation of Sparse Random Kernel Graphs”. PLoS ONE 10(9): e0135177, 2015.
doi:10.1371/journal.pone.0135177

5.6. Random Graphs 535


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

Generate an Erdős–Rényi random graph 𝐺(𝑛, 𝑐/𝑛), with kernel 𝜅(𝑥, 𝑦) = 𝑐 where 𝑐 is the mean expected
degree.
>>> def integral(u, w, z):
... return c * (z - w)
>>> def root(u, w, r):
... return r / c + w
>>> c = 1
>>> graph = nx.random_kernel_graph(1000, integral, root)

See also:
gnp_random_graph(), expected_degree_graph()

References

5.6.16 networkx.generators.random_graphs.random_lobster

random_lobster(n, p1, p2, seed=None)


Returns a random lobster graph.
A lobster is a tree that reduces to a caterpillar when pruning all leaf nodes. A caterpillar is a tree that reduces to
a path graph when pruning all leaf nodes; setting p2 to zero produces a caterpillar.
Parameters
• n (int) – The expected number of nodes in the backbone
• p1 (float) – Probability of adding an edge to the backbone
• p2 (float) – Probability of adding an edge one level beyond backbone
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

5.6.17 networkx.generators.random_graphs.random_shell_graph

random_shell_graph(constructor, seed=None)
Returns a random shell graph for the constructor given.
Parameters
• constructor (list of three-tuples) – Represents the parameters for a shell, starting at the
center shell. Each element of the list must be of the form (n, m, d), where n is the
number of nodes in the shell, m is the number of edges in the shell, and d is the ratio of
inter-shell (next) edges to intra-shell edges. If d is zero, there will be no intra-shell edges,
and if d is one there will be all possible intra-shell edges.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

Examples

>>> constructor = [(10, 20, 0.8), (20, 40, 0.8)]


>>> G = nx.random_shell_graph(constructor)

536 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

5.6.18 networkx.generators.random_graphs.random_powerlaw_tree

random_powerlaw_tree(n, gamma=3, seed=None, tries=100)


Returns a tree with a power law degree distribution.
Parameters
• n (int) – The number of nodes.
• gamma (float) – Exponent of the power law.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• tries (int) – Number of attempts to adjust the sequence to make it a tree.
Raises NetworkXError – If no valid sequence is found within the maximum number of attempts.

Notes

A trial power law degree sequence is chosen and then elements are swapped with new elements from a powerlaw
distribution until the sequence makes a tree (by checking, for example, that the number of edges is one smaller
than the number of nodes).

5.6.19 networkx.generators.random_graphs.random_powerlaw_tree_sequence

random_powerlaw_tree_sequence(n, gamma=3, seed=None, tries=100)


Returns a degree sequence for a tree with a power law distribution.
Parameters
• n (int,) – The number of nodes.
• gamma (float) – Exponent of the power law.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• tries (int) – Number of attempts to adjust the sequence to make it a tree.
Raises NetworkXError – If no valid sequence is found within the maximum number of attempts.

Notes

A trial power law degree sequence is chosen and then elements are swapped with new elements from a power
law distribution until the sequence makes a tree (by checking, for example, that the number of edges is one
smaller than the number of nodes).

5.7 Duplication Divergence

Functions for generating graphs based on the “duplication” method.


These graph generators start with a small initial graph then duplicate nodes and (partially) duplicate their edges. These
functions are generally inspired by biological networks.

5.7. Duplication Divergence 537


NetworkX Reference, Release 2.3rc1.dev20190222214247

duplication_divergence_graph(n, p[, seed]) Returns an undirected graph using the duplication-


divergence model.
partial_duplication_graph(N, n, p, q[, seed]) Returns a random graph using the partial duplication
model.

5.7.1 networkx.generators.duplication.duplication_divergence_graph

duplication_divergence_graph(n, p, seed=None)
Returns an undirected graph using the duplication-divergence model.
A graph of n nodes is created by duplicating the initial nodes and retaining edges incident to the original nodes
with a retention probability p.
Parameters
• n (int) – The desired number of nodes in the graph.
• p (float) – The probability for retaining the edge of the replicated node.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G
Return type Graph
Raises NetworkXError – If p is not a valid probability. If n is less than 2.

Notes

This algorithm appears in [1].


This implementation disallows the possibility of generating disconnected graphs.

References

5.7.2 networkx.generators.duplication.partial_duplication_graph

partial_duplication_graph(N, n, p, q, seed=None)
Returns a random graph using the partial duplication model.
Parameters
• N (int) – The total number of nodes in the final graph.
• n (int) – The number of nodes in the initial clique.
• p (float) – The probability of joining each neighbor of a node to the duplicate node. Must
be a number in the between zero and one, inclusive.
• q (float) – The probability of joining the source node to the duplicate node. Must be a
number in the between zero and one, inclusive.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

538 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

A graph of nodes is grown by creating a fully connected graph of size n. The following procedure is then
repeated until a total of N nodes have been reached.
1. A random node, u, is picked and a new node, v, is created.
2. For each neighbor of u an edge from the neighbor to v is created with probability p.
3. An edge from u to v is created with probability q.
This algorithm appears in [1].
This implementation allows the possibility of generating disconnected graphs.

References

5.8 Degree Sequence

Generate graphs with a given degree sequence or expected degree sequence.

configuration_model(deg_sequence[, . . . ]) Returns a random graph with the given degree sequence.


directed_configuration_model(. . . [, . . . ]) Returns a directed_random graph with the given degree
sequences.
expected_degree_graph(w[, seed, selfloops]) Returns a random graph with given expected degrees.
havel_hakimi_graph(deg_sequence[, cre- Returns a simple graph with given degree sequence con-
ate_using]) structed using the Havel-Hakimi algorithm.
directed_havel_hakimi_graph(in_deg_sequence,Returns a directed graph with the given degree se-
...) quences.
degree_sequence_tree(deg_sequence[, . . . ]) Make a tree for the given degree sequence.
random_degree_sequence_graph(sequence[, Returns a simple random graph with the given degree
. . . ]) sequence.

5.8.1 networkx.generators.degree_seq.configuration_model

configuration_model(deg_sequence, create_using=None, seed=None)


Returns a random graph with the given degree sequence.
The configuration model generates a random pseudograph (graph with parallel edges and self loops) by ran-
domly assigning edges to match the given degree sequence.
Parameters
• deg_sequence (list of nonnegative integers) – Each list entry corresponds to the degree of a
node.
• create_using (NetworkX graph constructor, optional (default MultiGraph)) – Graph type to
create. If graph instance, then cleared before populated.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G – A graph with the specified degree sequence. Nodes are labeled starting at 0 with an
index corresponding to the position in deg_sequence.
Return type MultiGraph

5.8. Degree Sequence 539


NetworkX Reference, Release 2.3rc1.dev20190222214247

Raises NetworkXError – If the degree sequence does not have an even sum.
See also:
is_graphical()

Notes

As described by Newman1 .
A non-graphical degree sequence (not realizable by some simple graph) is allowed since this function returns
graphs with self loops and parallel edges. An exception is raised if the degree sequence does not have an even
sum.
This configuration model construction process can lead to duplicate edges and loops. You can remove the
self-loops and parallel edges (see below) which will likely result in a graph that doesn’t have the exact degree
sequence specified.
The density of self-loops and parallel edges tends to decrease as the number of nodes increases. However,
typically the number of self-loops will approach a Poisson distribution with a nonzero mean, and similarly for
the number of parallel edges. Consider a node with k stubs. The probability of being joined to another stub of
the same node is basically (k - 1) / N, where k is the degree and N is the number of nodes. So the probability of
a self-loop scales like c / N for some constant c. As N grows, this means we expect c self-loops. Similarly for
parallel edges.

References

Examples

You can create a degree sequence following a particular distribution by using the one of the distribution functions
in random_sequence (or one of your own). For example, to create an undirected multigraph on one hundred
nodes with degree sequence chosen from the power law distribution:

>>> sequence = nx.random_powerlaw_tree_sequence(100, tries=5000)


>>> G = nx.configuration_model(sequence)
>>> len(G)
100
>>> actual_degrees = [d for v, d in G.degree()]
>>> actual_degrees == sequence
True

The returned graph is a multigraph, which may have parallel edges. To remove any parallel edges from the
returned graph:

>>> G = nx.Graph(G)

Similarly, to remove self-loops:

>>> G.remove_edges_from(nx.selfloop_edges(G))

1 M.E.J. Newman, “The structure and function of complex networks”, SIAM REVIEW 45-2, pp 167-256, 2003.

540 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

5.8.2 networkx.generators.degree_seq.directed_configuration_model

directed_configuration_model(in_degree_sequence, out_degree_sequence, create_using=None,


seed=None)
Returns a directed_random graph with the given degree sequences.
The configuration model generates a random directed pseudograph (graph with parallel edges and self loops) by
randomly assigning edges to match the given degree sequences.
Parameters
• in_degree_sequence (list of nonnegative integers) – Each list entry corresponds to the in-
degree of a node.
• out_degree_sequence (list of nonnegative integers) – Each list entry corresponds to the
out-degree of a node.
• create_using (NetworkX graph constructor, optional (default MultiDiGraph)) – Graph type
to create. If graph instance, then cleared before populated.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G – A graph with the specified degree sequences. Nodes are labeled starting at 0 with an
index corresponding to the position in deg_sequence.
Return type MultiDiGraph
Raises NetworkXError – If the degree sequences do not have the same sum.
See also:
configuration_model()

Notes

Algorithm as described by Newman1 .


A non-graphical degree sequence (not realizable by some simple graph) is allowed since this function returns
graphs with self loops and parallel edges. An exception is raised if the degree sequences does not have the same
sum.
This configuration model construction process can lead to duplicate edges and loops. You can remove the
self-loops and parallel edges (see below) which will likely result in a graph that doesn’t have the exact degree
sequence specified. This “finite-size effect” decreases as the size of the graph increases.

References

Examples

One can modify the in- and out-degree sequences from an existing directed graph in order to create a new
directed graph. For example, here we modify the directed path graph:
>>> D = nx.DiGraph([(0, 1), (1, 2), (2, 3)])
>>> din = list(d for n, d in D.in_degree())
>>> dout = list(d for n, d in D.out_degree())
>>> din.append(1)
(continues on next page)
1 Newman, M. E. J. and Strogatz, S. H. and Watts, D. J. Random graphs with arbitrary degree distributions and their applications Phys. Rev. E,
64, 026118 (2001)

5.8. Degree Sequence 541


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> dout[0] = 2
>>> # We now expect an edge from node 0 to a new node, node 3.
... D = nx.directed_configuration_model(din, dout)

The returned graph is a directed multigraph, which may have parallel edges. To remove any parallel edges from
the returned graph:

>>> D = nx.DiGraph(D)

Similarly, to remove self-loops:

>>> D.remove_edges_from(nx.selfloop_edges(D))

5.8.3 networkx.generators.degree_seq.expected_degree_graph

expected_degree_graph(w, seed=None, selfloops=True)


Returns a random graph with given expected degrees.
Given a sequence of expected degrees 𝑊 = (𝑤0 , 𝑤1 , . . . , 𝑤𝑛−1 ) of length 𝑛 this algorithm assigns an edge
between node 𝑢 and node 𝑣 with probability
𝑤𝑢 𝑤𝑣
𝑝𝑢𝑣 = ∑︀ .
𝑘 𝑤𝑘

Parameters
• w (list) – The list of expected degrees.
• selfloops (bool (default=True)) – Set to False to remove the possibility of self-loop edges.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns
Return type Graph

Examples

>>> z=[10 for i in range(100)]


>>> G=nx.expected_degree_graph(z)

Notes

The nodes have integer labels corresponding to index of expected degrees input sequence.
The complexity of this algorithm is 𝒪(𝑛 + 𝑚) where 𝑛 is the number of nodes and 𝑚 is the expected number
of edges.
The model in1 includes the possibility of self-loop edges. Set selfloops=False to produce a graph without self
loops.
1 Fan Chung and L. Lu, Connected components in random graphs with given expected degree sequences, Ann. Combinatorics, 6, pp. 125-145,

2002.

542 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

For finite graphs this model doesn’t produce exactly the given expected degree sequence. Instead the expected
degrees are as follows.
For the case without self loops (selfloops=False),
(︂ )︂
∑︁ 𝑤𝑢
𝐸[𝑑𝑒𝑔(𝑢)] = 𝑝𝑢𝑣 = 𝑤𝑢 1 − ∑︀ .
𝑣̸=𝑢 𝑘 𝑤𝑘

NetworkX uses the standard convention that a self-loop edge counts 2 in the degree of a node, so with self loops
(selfloops=True),
(︂ )︂
∑︁ 𝑤𝑢
𝐸[𝑑𝑒𝑔(𝑢)] = 𝑝𝑢𝑣 + 2𝑝𝑢𝑢 = 𝑤𝑢 1 + ∑︀ .
𝑣̸=𝑢 𝑘 𝑤𝑘

References

5.8.4 networkx.generators.degree_seq.havel_hakimi_graph

havel_hakimi_graph(deg_sequence, create_using=None)
Returns a simple graph with given degree sequence constructed using the Havel-Hakimi algorithm.
Parameters
• deg_sequence (list of integers) – Each integer corresponds to the degree of a node (need not
be sorted).
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated. Directed graphs are not allowed.
Raises NetworkXException – For a non-graphical degree sequence (i.e. one not realizable by
some simple graph).

Notes

The Havel-Hakimi algorithm constructs a simple graph by successively connecting the node of highest degree
to other nodes of highest degree, resorting remaining nodes by degree, and repeating the process. The resulting
graph has a high degree-associativity. Nodes are labeled 1,.., len(deg_sequence), corresponding to their position
in deg_sequence.
The basic algorithm is from Hakimi1 and was generalized by Kleitman and Wang2 .

References

5.8.5 networkx.generators.degree_seq.directed_havel_hakimi_graph

directed_havel_hakimi_graph(in_deg_sequence, out_deg_sequence, create_using=None)


Returns a directed graph with the given degree sequences.
Parameters
• in_deg_sequence (list of integers) – Each list entry corresponds to the in-degree of a node.
1 Hakimi S., On Realizability of a Set of Integers as Degrees of the Vertices of a Linear Graph. I, Journal of SIAM, 10(3), pp. 496-506 (1962)
2 Kleitman D.J. and Wang D.L. Algorithms for Constructing Graphs and Digraphs with Given Valences and Factors Discrete Mathematics, 6(1),
pp. 79-88 (1973)

5.8. Degree Sequence 543


NetworkX Reference, Release 2.3rc1.dev20190222214247

• out_deg_sequence (list of integers) – Each list entry corresponds to the out-degree of a


node.
• create_using (NetworkX graph constructor, optional (default DiGraph)) – Graph type to
create. If graph instance, then cleared before populated.
Returns G – A graph with the specified degree sequences. Nodes are labeled starting at 0 with an
index corresponding to the position in deg_sequence
Return type DiGraph
Raises NetworkXError – If the degree sequences are not digraphical.
See also:
configuration_model()

Notes

Algorithm as described by Kleitman and Wang1 .

References

5.8.6 networkx.generators.degree_seq.degree_sequence_tree

degree_sequence_tree(deg_sequence, create_using=None)
Make a tree for the given degree sequence.
A tree has #nodes-#edges=1 so the degree sequence must have len(deg_sequence)-sum(deg_sequence)/2=1

5.8.7 networkx.generators.degree_seq.random_degree_sequence_graph

random_degree_sequence_graph(sequence, seed=None, tries=10)


Returns a simple random graph with the given degree sequence.
If the maximum degree 𝑑𝑚 in the sequence is 𝑂(𝑚1/4 ) then the algorithm produces almost uniform random
graphs in 𝑂(𝑚𝑑𝑚 ) time where 𝑚 is the number of edges.
Parameters
• sequence (list of integers) – Sequence of degrees
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• tries (int, optional) – Maximum number of tries to create a graph
Returns G – A graph with the specified degree sequence. Nodes are labeled starting at 0 with an
index corresponding to the position in the sequence.
Return type Graph
Raises
• NetworkXUnfeasible – If the degree sequence is not graphical.
• NetworkXError – If a graph is not produced in specified number of tries
1 D.J. Kleitman and D.L. Wang Algorithms for Constructing Graphs and Digraphs with Given Valences and Factors Discrete Mathematics, 6(1),

pp. 79-88 (1973)

544 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
is_graphical(), configuration_model()

Notes

The generator algorithm1 is not guaranteed to produce a graph.

References

Examples

>>> sequence = [1, 2, 2, 3]


>>> G = nx.random_degree_sequence_graph(sequence)
>>> sorted(d for n, d in G.degree())
[1, 2, 2, 3]

5.9 Random Clustered

Generate graphs with given degree and triangle sequence.

random_clustered_graph(joint_degree_sequence) Generate a random graph with the given joint indepen-


dent edge degree and triangle degree sequence.

5.9.1 networkx.generators.random_clustered.random_clustered_graph

random_clustered_graph(joint_degree_sequence, create_using=None, seed=None)


Generate a random graph with the given joint independent edge degree and triangle degree sequence.
This uses a configuration model-like approach to generate a random graph (with parallel edges and self-loops)
by randomly assigning edges to match the given joint degree sequence.
The joint degree sequence is a list of pairs of integers of the form [(𝑑1,𝑖 , 𝑑1,𝑡 ), . . . , (𝑑𝑛,𝑖 , 𝑑𝑛,𝑡 )]. According to
this list, vertex 𝑢 is a member of 𝑑𝑢,𝑡 triangles and has 𝑑𝑢,𝑖 other edges. The number 𝑑𝑢,𝑡 is the triangle degree
of 𝑢 and the number 𝑑𝑢,𝑖 is the independent edge degree.
Parameters
• joint_degree_sequence (list of integer pairs) – Each list entry corresponds to the indepen-
dent edge degree and triangle degree of a node.
• create_using (NetworkX graph constructor, optional (default MultiGraph)) – Graph type to
create. If graph instance, then cleared before populated.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G – A graph with the specified degree sequence. Nodes are labeled starting at 0 with an
index corresponding to the position in deg_sequence.
Return type MultiGraph
1 Moshen Bayati, Jeong Han Kim, and Amin Saberi, A sequential algorithm for generating random graphs. Algorithmica, Volume 58, Number

4, 860-910, DOI: 10.1007/s00453-009-9340-1

5.9. Random Clustered 545


NetworkX Reference, Release 2.3rc1.dev20190222214247

Raises NetworkXError – If the independent edge degree sequence sum is not even or the triangle
degree sequence sum is not divisible by 3.

Notes

As described by Miller1 (see also Newman2 for an equivalent description).


A non-graphical degree sequence (not realizable by some simple graph) is allowed since this function returns
graphs with self loops and parallel edges. An exception is raised if the independent degree sequence does not
have an even sum or the triangle degree sequence sum is not divisible by 3.
This configuration model-like construction process can lead to duplicate edges and loops. You can remove the
self-loops and parallel edges (see below) which will likely result in a graph that doesn’t have the exact degree
sequence specified. This “finite-size effect” decreases as the size of the graph increases.

References

Examples

>>> deg = [(1, 0), (1, 0), (1, 0), (2, 0), (1, 0), (2, 1), (0, 1), (0, 1)]
>>> G = nx.random_clustered_graph(deg)

To remove parallel edges:

>>> G = nx.Graph(G)

To remove self loops:

>>> G.remove_edges_from(nx.selfloop_edges(G))

5.10 Directed

Generators for some directed graphs, including growing network (GN) graphs and scale-free graphs.

gn_graph(n[, kernel, create_using, seed]) Returns the growing network (GN) digraph with n
nodes.
gnr_graph(n, p[, create_using, seed]) Returns the growing network with redirection (GNR)
digraph with n nodes and redirection probability p.
gnc_graph(n[, create_using, seed]) Returns the growing network with copying (GNC) di-
graph with n nodes.
random_k_out_graph(n, k, alpha[, . . . ]) Returns a random k-out graph with preferential attach-
ment.
scale_free_graph(n[, alpha, beta, gamma, . . . ]) Returns a scale-free directed graph.
1 Joel C. Miller. “Percolation and epidemics in random clustered networks”. In: Physical review. E, Statistical, nonlinear, and soft matter

physics 80 (2 Part 1 August 2009).


2 M. E. J. Newman. “Random Graphs with Clustering”. In: Physical Review Letters 103 (5 July 2009)

546 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

5.10.1 networkx.generators.directed.gn_graph

gn_graph(n, kernel=None, create_using=None, seed=None)


Returns the growing network (GN) digraph with n nodes.
The GN graph is built by adding nodes one at a time with a link to one previously added node. The target node
for the link is chosen with probability based on degree. The default attachment kernel is a linear function of the
degree of a node.
The graph is always a (directed) tree.
Parameters
• n (int) – The number of nodes for the generated graph.
• kernel (function) – The attachment kernel.
• create_using (NetworkX graph constructor, optional (default DiGraph)) – Graph type to
create. If graph instance, then cleared before populated.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

Examples

To create the undirected GN graph, use the to_directed() method:

>>> D = nx.gn_graph(10) # the GN graph


>>> G = D.to_undirected() # the undirected version

To specify an attachment kernel, use the kernel keyword argument:

>>> D = nx.gn_graph(10, kernel=lambda x: x ** 1.5) # A_k = k^1.5

References

5.10.2 networkx.generators.directed.gnr_graph

gnr_graph(n, p, create_using=None, seed=None)


Returns the growing network with redirection (GNR) digraph with n nodes and redirection probability p.
The GNR graph is built by adding nodes one at a time with a link to one previously added node. The previous
target node is chosen uniformly at random. With probabiliy p the link is instead “redirected” to the successor
node of the target.
The graph is always a (directed) tree.
Parameters
• n (int) – The number of nodes for the generated graph.
• p (float) – The redirection probability.
• create_using (NetworkX graph constructor, optional (default DiGraph)) – Graph type to
create. If graph instance, then cleared before populated.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

5.10. Directed 547


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

To create the undirected GNR graph, use the to_directed() method:

>>> D = nx.gnr_graph(10, 0.5) # the GNR graph


>>> G = D.to_undirected() # the undirected version

References

5.10.3 networkx.generators.directed.gnc_graph

gnc_graph(n, create_using=None, seed=None)


Returns the growing network with copying (GNC) digraph with n nodes.
The GNC graph is built by adding nodes one at a time with a link to one previously added node (chosen
uniformly at random) and to all of that node’s successors.
Parameters
• n (int) – The number of nodes for the generated graph.
• create_using (NetworkX graph constructor, optional (default DiGraph)) – Graph type to
create. If graph instance, then cleared before populated.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

References

5.10.4 networkx.generators.directed.random_k_out_graph

random_k_out_graph(n, k, alpha, self_loops=True, seed=None)


Returns a random k-out graph with preferential attachment.
A random k-out graph with preferential attachment is a multidigraph generated by the following algorithm.
1. Begin with an empty digraph, and initially set each node to have weight alpha.
2. Choose a node u with out-degree less than k uniformly at random.
3. Choose a node v from with probability proportional to its weight.
4. Add a directed edge from u to v, and increase the weight of v by one.
5. If each node has out-degree k, halt, otherwise repeat from step 2.
For more information on this model of random graph, see [1].
Parameters
• n (int) – The number of nodes in the returned graph.
• k (int) – The out-degree of each node in the returned graph.
• alpha (float) – A positive float representing the initial weight of each vertex. A higher
number means that in step 3 above, nodes will be chosen more like a true uniformly random
sample, and a lower number means that nodes are more likely to be chosen as their in-degree
increases. If this parameter is not positive, a ValueError is raised.
• self_loops (bool) – If True, self-loops are allowed when generating the graph.

548 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

• seed (integer, random_state, or None (default)) – Indicator of random number generation


state. See Randomness.
Returns A k-out-regular multidigraph generated according to the above algorithm.
Return type MultiDiGraph
Raises ValueError – If alpha is not positive.

Notes

The returned multidigraph may not be strongly connected, or even weakly connected.

References

[1]: Peterson, Nicholas R., and Boris Pittel. “Distance between two random k-out digraphs, with and without
preferential attachment.” arXiv preprint arXiv:1311.5961 (2013). <https://arxiv.org/abs/1311.5961>

5.10.5 networkx.generators.directed.scale_free_graph

scale_free_graph(n, alpha=0.41, beta=0.54, gamma=0.05, delta_in=0.2, delta_out=0, cre-


ate_using=None, seed=None)
Returns a scale-free directed graph.
Parameters
• n (integer) – Number of nodes in graph
• alpha (float) – Probability for adding a new node connected to an existing node chosen
randomly according to the in-degree distribution.
• beta (float) – Probability for adding an edge between two existing nodes. One existing node
is chosen randomly according the in-degree distribution and the other chosen randomly
according to the out-degree distribution.
• gamma (float) – Probability for adding a new node connected to an existing node chosen
randomly according to the out-degree distribution.
• delta_in (float) – Bias for choosing nodes from in-degree distribution.
• delta_out (float) – Bias for choosing nodes from out-degree distribution.
• create_using (NetworkX graph constructor, optional) – The default is a MultiDiGraph 3-
cycle. If a graph instance, use it without clearing first. If a graph constructor, call it to
construct an empty graph.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

Examples

Create a scale-free graph on one hundred nodes:

>>> G = nx.scale_free_graph(100)

5.10. Directed 549


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The sum of alpha, beta, and gamma must be 1.

References

5.11 Geometric

Generators for geometric graphs.

random_geometric_graph(n, radius[, dim, . . . ]) Returns a random geometric graph in the unit cube of
dimensions dim.
soft_random_geometric_graph(n, radius[, Returns a soft random geometric graph in the unit cube.
. . . ])
geographical_threshold_graph(n, theta[, Returns a geographical threshold graph.
. . . ])
waxman_graph(n[, beta, alpha, L, domain, . . . ]) Returns a Waxman random graph.
navigable_small_world_graph(n[, p, q, r, Returns a navigable small-world graph.
. . . ])
thresholded_random_geometric_graph(n, Returns a thresholded random geometric graph in the
. . . [, . . . ]) unit cube.

5.11.1 networkx.generators.geometric.random_geometric_graph

random_geometric_graph(n, radius, dim=2, pos=None, p=2, seed=None)


Returns a random geometric graph in the unit cube of dimensions dim.
The random geometric graph model places n nodes uniformly at random in the unit cube. Two nodes are joined
by an edge if the distance between the nodes is at most radius.
Edges are determined using a KDTree when SciPy is available. This reduces the time complexity from 𝑂(𝑛2 )
to 𝑂(𝑛).
Parameters
• n (int or iterable) – Number of nodes or iterable of nodes
• radius (float) – Distance threshold value
• dim (int, optional) – Dimension of graph
• pos (dict, optional) – A dictionary keyed by node with node positions as values.
• p (float, optional) – Which Minkowski distance metric to use. p has to meet the condition
1 <= p <= infinity.
If this argument is not specified, the 𝐿2 metric (the Euclidean distance metric), p = 2 is used.
This should not be confused with the p of an Erdős-Rényi random graph, which represents
probability.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns A random geometric graph, undirected and without self-loops. Each node has a node at-
tribute 'pos' that stores the position of that node in Euclidean space as provided by the pos
keyword argument or, if pos was not provided, as generated by this function.

550 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type Graph

Examples

Create a random geometric graph on twenty nodes where nodes are joined by an edge if their distance is at most
0.1:

>>> G = nx.random_geometric_graph(20, 0.1)

Notes

This uses a k-d tree to build the graph.


The pos keyword argument can be used to specify node positions so you can create an arbitrary distribution
and domain for positions.
For example, to use a 2D Gaussian distribution of node positions with mean (0, 0) and standard deviation 2:

>>> import random


>>> n = 20
>>> pos = {i: (random.gauss(0, 2), random.gauss(0, 2)) for i in range(n)}
>>> G = nx.random_geometric_graph(n, 0.2, pos=pos)

References

5.11.2 networkx.generators.geometric.soft_random_geometric_graph

soft_random_geometric_graph(n, radius, dim=2, pos=None, p=2, p_dist=None, seed=None)


Returns a soft random geometric graph in the unit cube.
The soft random geometric graph [1] model places n nodes uniformly at random in the unit cube in dimension
dim. Two nodes of distance, dist, computed by the p-Minkowski distance metric are joined by an edge with
probability p_dist if the computed distance metric value of the nodes is at most radius, otherwise they are
not joined.
Edges within radius of each other are determined using a KDTree when SciPy is available. This reduces the
time complexity from 𝑂(𝑛2 ) to 𝑂(𝑛).
Parameters
• n (int or iterable) – Number of nodes or iterable of nodes
• radius (float) – Distance threshold value
• dim (int, optional) – Dimension of graph
• pos (dict, optional) – A dictionary keyed by node with node positions as values.
• p (float, optional) – Which Minkowski distance metric to use. p has to meet the condition
1 <= p <= infinity.
If this argument is not specified, the 𝐿2 metric (the Euclidean distance metric), p = 2 is used.
This should not be confused with the p of an Erdős-Rényi random graph, which represents
probability.

5.11. Geometric 551


NetworkX Reference, Release 2.3rc1.dev20190222214247

• p_dist (function, optional) – A probability density function computing the probability of


connecting two nodes that are of distance, dist, computed by the Minkowski distance met-
ric. The probability density function, p_dist, must be any function that takes the metric
value as input and outputs a single probability value between 0-1. The scipy.stats package
has many probability distribution functions implemented and tools for custom probability
distribution definitions [2], and passing the .pdf method of scipy.stats distributions can be
used here. If the probability function, p_dist, is not supplied, the default function is an
exponential distribution with rate parameter 𝜆 = 1.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns A soft random geometric graph, undirected and without self-loops. Each node has a node
attribute 'pos' that stores the position of that node in Euclidean space as provided by the pos
keyword argument or, if pos was not provided, as generated by this function.
Return type Graph

Examples

Default Graph:
G = nx.soft_random_geometric_graph(50, 0.2)
Custom Graph:
Create a soft random geometric graph on 100 uniformly distributed nodes where nodes are joined by an edge
with probability computed from an exponential distribution with rate parameter 𝜆 = 1 if their Euclidean distance
is at most 0.2.

Notes

This uses a k-d tree to build the graph.


The pos keyword argument can be used to specify node positions so you can create an arbitrary distribution
and domain for positions.
For example, to use a 2D Gaussian distribution of node positions with mean (0, 0) and standard deviation 2
The scipy.stats package can be used to define the probability distribution with the .pdf method used as p_dist.

>>> import random


>>> import math
>>> n = 100
>>> pos = {i: (random.gauss(0, 2), random.gauss(0, 2)) for i in range(n)}
>>> p_dist = lambda dist : math.exp(-dist)
>>> G = nx.soft_random_geometric_graph(n, 0.2, pos=pos, p_dist=p_dist)

References

5.11.3 networkx.generators.geometric.geographical_threshold_graph

geographical_threshold_graph(n, theta, dim=2, pos=None, weight=None, metric=None,


p_dist=None, seed=None)
Returns a geographical threshold graph.

552 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

The geographical threshold graph model places 𝑛 nodes uniformly at random in a rectangular domain. Each
node 𝑢 is assigned a weight 𝑤𝑢 . Two nodes 𝑢 and 𝑣 are joined by an edge if

(𝑤𝑢 + 𝑤𝑣 )ℎ(𝑟) ≥ 𝜃

where r is the distance between u and v, h(r) is a probability of connection as a function of r, and 𝜃 as the
threshold parameter. h(r) corresponds to the p_dist parameter.
Parameters
• n (int or iterable) – Number of nodes or iterable of nodes
• theta (float) – Threshold value
• dim (int, optional) – Dimension of graph
• pos (dict) – Node positions as a dictionary of tuples keyed by node.
• weight (dict) – Node weights as a dictionary of numbers keyed by node.
• metric (function) – A metric on vectors of numbers (represented as lists or tuples). This
must be a function that accepts two lists (or tuples) as input and yields a number as output.
The function must also satisfy the four requirements of a metric. Specifically, if 𝑑 is the
function and 𝑥, 𝑦, and 𝑧 are vectors in the graph, then 𝑑 must satisfy
1. 𝑑(𝑥, 𝑦) ≥ 0,
2. 𝑑(𝑥, 𝑦) = 0 if and only if 𝑥 = 𝑦,
3. 𝑑(𝑥, 𝑦) = 𝑑(𝑦, 𝑥),
4. 𝑑(𝑥, 𝑧) ≤ 𝑑(𝑥, 𝑦) + 𝑑(𝑦, 𝑧).
If this argument is not specified, the Euclidean distance metric is used.
• p_dist (function, optional) – A probability density function computing the probability of
connecting two nodes that are of distance, r, computed by metric. The probability density
function, p_dist, must be any function that takes the metric value as input and outputs a
single probability value between 0-1. The scipy.stats package has many probability distri-
bution functions implemented and tools for custom probability distribution definitions [2],
and passing the .pdf method of scipy.stats distributions can be used here. If the probability
function, p_dist, is not supplied, the default exponential function :math: r^{-2} is used.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns
A random geographic threshold graph, undirected and without self-loops.
Each node has a node attribute pos that stores the position of that node in Euclidean space
as provided by the pos keyword argument or, if pos was not provided, as generated by this
function. Similarly, each node has a node attribute weight that stores the weight of that node
as provided or as generated.
Return type Graph

Examples

Specify an alternate distance metric using the metric keyword argument. For example, to use the taxicab
metric instead of the default Euclidean metric:

5.11. Geometric 553


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> dist = lambda x, y: sum(abs(a - b) for a, b in zip(x, y))


>>> G = nx.geographical_threshold_graph(10, 0.1, metric=dist)

Notes

If weights are not specified they are assigned to nodes by drawing randomly from the exponential distribution
with rate parameter 𝜆 = 1. To specify weights from a different distribution, use the weight keyword argument:

>>> import random


>>> n = 20
>>> w = {i: random.expovariate(5.0) for i in range(n)}
>>> G = nx.geographical_threshold_graph(20, 50, weight=w)

If node positions are not specified they are randomly assigned from the uniform distribution.
Starting in NetworkX 2.1 the parameter alpha is deprecated and replaced with the customizable p_dist
function parameter, which defaults to r^-2 if p_dist is not supplied. To reproduce networks of earlier Net-
workX versions, a custom function needs to be defined and passed as the p_dist parameter. For example,
if the parameter alpha = 2 was used in NetworkX 2.0, the custom function def custom_dist(r): r**-2 can be
passed in versions >=2.1 as the parameter p_dist = custom_dist to produce an equivalent network. Note the
change in sign from +2 to -2 in this parameter change.

References

5.11.4 networkx.generators.geometric.waxman_graph

waxman_graph(n, beta=0.4, alpha=0.1, L=None, domain=(0, 0, 1, 1), metric=None, seed=None)


Returns a Waxman random graph.
The Waxman random graph model places n nodes uniformly at random in a rectangular domain. Each pair of
nodes at distance d is joined by an edge with probability

𝑝 = 𝛽 exp(−𝑑/𝛼𝐿).

This function implements both Waxman models, using the L keyword argument.
• Waxman-1: if L is not specified, it is set to be the maximum distance between any pair of nodes.
• Waxman-2: if L is specified, the distance between a pair of nodes is chosen uniformly at random from the
interval [0, L].

Parameters
• n (int or iterable) – Number of nodes or iterable of nodes
• beta (float) – Model parameter
• alpha (float) – Model parameter
• L (float, optional) – Maximum distance between nodes. If not specified, the actual distance
is calculated.
• domain (four-tuple of numbers, optional) – Domain size, given as a tuple of the form
(x_min, y_min, x_max, y_max).

554 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

• metric (function) – A metric on vectors of numbers (represented as lists or tuples). This


must be a function that accepts two lists (or tuples) as input and yields a number as output.
The function must also satisfy the four requirements of a metric. Specifically, if 𝑑 is the
function and 𝑥, 𝑦, and 𝑧 are vectors in the graph, then 𝑑 must satisfy
1. 𝑑(𝑥, 𝑦) ≥ 0,
2. 𝑑(𝑥, 𝑦) = 0 if and only if 𝑥 = 𝑦,
3. 𝑑(𝑥, 𝑦) = 𝑑(𝑦, 𝑥),
4. 𝑑(𝑥, 𝑧) ≤ 𝑑(𝑥, 𝑦) + 𝑑(𝑦, 𝑧).
If this argument is not specified, the Euclidean distance metric is used.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns A random Waxman graph, undirected and without self-loops. Each node has a node at-
tribute 'pos' that stores the position of that node in Euclidean space as generated by this
function.
Return type Graph

Examples

Specify an alternate distance metric using the metric keyword argument. For example, to use the “taxicab
metric” instead of the default Euclidean metric:

>>> dist = lambda x, y: sum(abs(a - b) for a, b in zip(x, y))


>>> G = nx.waxman_graph(10, 0.5, 0.1, metric=dist)

Notes

Starting in NetworkX 2.0 the parameters alpha and beta align with their usual roles in the probability distribution.
In earlier versions their positions in the expression were reversed. Their position in the calling sequence reversed
as well to minimize backward incompatibility.

References

5.11.5 networkx.generators.geometric.navigable_small_world_graph

navigable_small_world_graph(n, p=1, q=1, r=2, dim=2, seed=None)


Returns a navigable small-world graph.
A navigable small-world graph is a directed grid with additional long-range connections that are chosen ran-
domly.
[. . . ] we begin with a set of nodes [. . . ] that are identified with the set of lattice points in an 𝑛 × 𝑛
square, {(𝑖, 𝑗) : 𝑖 ∈ {1, 2, . . . , 𝑛}, 𝑗 ∈ {1, 2, . . . , 𝑛}}, and we define the lattice distance between
two nodes (𝑖, 𝑗) and (𝑘, 𝑙) to be the number of “lattice steps” separating them: 𝑑((𝑖, 𝑗), (𝑘, 𝑙)) =
|𝑘 − 𝑖| + |𝑙 − 𝑗|.
For a universal constant 𝑝 >= 1, the node 𝑢 has a directed edge to every other node within lattice
distance 𝑝—these are its local contacts. For universal constants 𝑞 >= 0 and 𝑟 >= 0 we also
construct directed edges from 𝑢 to 𝑞 other nodes (the long-range contacts) using independent random
trials; the 𝑖 has endpoint 𝑣 with probability proportional to [𝑑(𝑢, 𝑣)]−𝑟 .

5.11. Geometric 555


NetworkX Reference, Release 2.3rc1.dev20190222214247

—1

Parameters
• n (int) – The length of one side of the lattice; the number of nodes in the graph is therefore
𝑛2 .
• p (int) – The diameter of short range connections. Each node is joined with every other node
within this lattice distance.
• q (int) – The number of long-range connections for each node.
• r (float) – Exponent for decaying probability of connections. The probability of connecting
to a node at lattice distance 𝑑 is 1/𝑑𝑟 .
• dim (int) – Dimension of grid
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

References

5.11.6 networkx.generators.geometric.thresholded_random_geometric_graph

thresholded_random_geometric_graph(n, radius, theta, dim=2, pos=None, weight=None, p=2,


seed=None)
Returns a thresholded random geometric graph in the unit cube.
The thresholded random geometric graph [1] model places n nodes uniformly at random in the unit cube of
dimensions dim. Each node u is assigned a weight 𝑤𝑢 . Two nodes u and v are joined by an edge if they are
within the maximum connection distance, radius computed by the p-Minkowski distance and the summation
of weights 𝑤𝑢 + 𝑤𝑣 is greater than or equal to the threshold parameter theta.
Edges within radius of each other are determined using a KDTree when SciPy is available. This reduces the
time complexity from 𝑂(𝑛2 ) to 𝑂(𝑛).
Parameters
• n (int or iterable) – Number of nodes or iterable of nodes
• radius (float) – Distance threshold value
• theta (float) – Threshold value
• dim (int, optional) – Dimension of graph
• pos (dict, optional) – A dictionary keyed by node with node positions as values.
• weight (dict, optional) – Node weights as a dictionary of numbers keyed by node.
• p (float, optional) – Which Minkowski distance metric to use. p has to meet the condition
1 <= p <= infinity.
If this argument is not specified, the 𝐿2 metric (the Euclidean distance metric), p = 2 is used.
This should not be confused with the p of an Erdős-Rényi random graph, which represents
probability.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
1 J. Kleinberg. The small-world phenomenon: An algorithmic perspective. Proc. 32nd ACM Symposium on Theory of Computing, 2000.

556 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns
A thresholded random geographic graph, undirected and without self-loops.
Each node has a node attribute 'pos' that stores the position of that node in Euclidean space
as provided by the pos keyword argument or, if pos was not provided, as generated by this
function. Similarly, each node has a nodethre attribute 'weight' that stores the weight of that
node as provided or as generated.
Return type Graph

Examples

Default Graph:
G = nx.thresholded_random_geometric_graph(50, 0.2, 0.1)
Custom Graph:
Create a thresholded random geometric graph on 50 uniformly distributed nodes where nodes are joined by
an edge if their sum weights drawn from a exponential distribution with rate = 5 are >= theta = 0.1 and their
Euclidean distance is at most 0.2.

Notes

This uses a k-d tree to build the graph.


The pos keyword argument can be used to specify node positions so you can create an arbitrary distribution
and domain for positions.
For example, to use a 2D Gaussian distribution of node positions with mean (0, 0) and standard deviation 2
If weights are not specified they are assigned to nodes by drawing randomly from the exponential distribution
with rate parameter 𝜆 = 1. To specify weights from a different distribution, use the weight keyword argument:

::

>>> import random


>>> import math
>>> n = 50
>>> pos = {i: (random.gauss(0, 2), random.gauss(0, 2)) for i in range(n)}
>>> w = {i: random.expovariate(5.0) for i in range(n)}
>>> G = nx.thresholded_random_geometric_graph(n, 0.2, 0.1, 2, pos, w)

References

5.12 Line Graph

Functions for generating line graphs.

line_graph(G[, create_using]) Returns the line graph of the graph or digraph G.


inverse_line_graph(G) Returns the inverse line graph of graph G.

5.12. Line Graph 557


NetworkX Reference, Release 2.3rc1.dev20190222214247

5.12.1 networkx.generators.line.line_graph

line_graph(G, create_using=None)
Returns the line graph of the graph or digraph G.
The line graph of a graph G has a node for each edge in G and an edge joining those nodes if the two edges in
G share a common node. For directed graphs, nodes are adjacent exactly when the edges they represent form a
directed path of length two.
The nodes of the line graph are 2-tuples of nodes in the original graph (or 3-tuples for multigraphs, with the key
of the edge as the third element).
For information about self-loops and more discussion, see the Notes section below.
Parameters
• G (graph) – A NetworkX Graph, DiGraph, MultiGraph, or MultiDigraph.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
Returns L – The line graph of G.
Return type graph

Examples

>>> import networkx as nx


>>> G = nx.star_graph(3)
>>> L = nx.line_graph(G)
>>> print(sorted(map(sorted, L.edges()))) # makes a 3-clique, K3
[[(0, 1), (0, 2)], [(0, 1), (0, 3)], [(0, 2), (0, 3)]]

Notes

Graph, node, and edge data are not propagated to the new graph. For undirected graphs, the nodes in G must be
sortable, otherwise the constructed line graph may not be correct.
Self-loops in undirected graphs
For an undirected graph G without multiple edges, each edge can be written as a set {u, v}. Its line graph
L has the edges of G as its nodes. If x and y are two nodes in L, then {x, y} is an edge in L if and only
if the intersection of x and y is nonempty. Thus, the set of all edges is determined by the set of all pairwise
intersections of edges in G.
Trivially, every edge in G would have a nonzero intersection with itself, and so every node in L should have
a self-loop. This is not so interesting, and the original context of line graphs was with simple graphs, which
had no self-loops or multiple edges. The line graph was also meant to be a simple graph and thus, self-loops in
L are not part of the standard definition of a line graph. In a pairwise intersection matrix, this is analogous to
excluding the diagonal entries from the line graph definition.
Self-loops and multiple edges in G add nodes to L in a natural way, and do not require any fundamental changes
to the definition. It might be argued that the self-loops we excluded before should now be included. However,
the self-loops are still “trivial” in some sense and thus, are usually excluded.
Self-loops in directed graphs

558 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

For a directed graph G without multiple edges, each edge can be written as a tuple (u, v). Its line graph L has
the edges of G as its nodes. If x and y are two nodes in L, then (x, y) is an edge in L if and only if the tail of
x matches the head of y, for example, if x = (a, b) and y = (b, c) for some vertices a, b, and c in G.
Due to the directed nature of the edges, it is no longer the case that every edge in G should have a self-loop in
L. Now, the only time self-loops arise is if a node in G itself has a self-loop. So such self-loops are no longer
“trivial” but instead, represent essential features of the topology of G. For this reason, the historical development
of line digraphs is such that self-loops are included. When the graph G has multiple edges, once again only
superficial changes are required to the definition.

References

• Harary, Frank, and Norman, Robert Z., “Some properties of line digraphs”, Rend. Circ. Mat. Palermo, II.
Ser. 9 (1960), 161–168.
• Hemminger, R. L.; Beineke, L. W. (1978), “Line graphs and line digraphs”, in Beineke, L. W.; Wilson, R.
J., Selected Topics in Graph Theory, Academic Press Inc., pp. 271–305.

5.12.2 networkx.generators.line.inverse_line_graph

inverse_line_graph(G)
Returns the inverse line graph of graph G.
If H is a graph, and G is the line graph of H, such that H = L(G). Then H is the inverse line graph of G.
Not all graphs are line graphs and these do not have an inverse line graph. In these cases this generator returns
a NetworkXError.
Parameters G (graph) – A NetworkX Graph
Returns H – The inverse line graph of G.
Return type graph
Raises
• NetworkXNotImplemented – If G is directed or a multigraph
• NetworkXError – If G is not a line graph

Notes

This is an implementation of the Roussopoulos algorithm.

References

• Roussopolous, N, “A max {m, n} algorithm for determining the graph H from its line graph G”, Informa-
tion Processing Letters 2, (1973), 108–112.

5.13 Ego Graph

Ego graph.

5.13. Ego Graph 559


NetworkX Reference, Release 2.3rc1.dev20190222214247

ego_graph(G, n[, radius, center, . . . ]) Returns induced subgraph of neighbors centered at node
n within a given radius.

5.13.1 networkx.generators.ego.ego_graph

ego_graph(G, n, radius=1, center=True, undirected=False, distance=None)


Returns induced subgraph of neighbors centered at node n within a given radius.
Parameters
• G (graph) – A NetworkX Graph or DiGraph
• n (node) – A single node
• radius (number, optional) – Include all neighbors of distance<=radius from n.
• center (bool, optional) – If False, do not include center node in graph
• undirected (bool, optional) – If True use both in- and out-neighbors of directed graphs.
• distance (key, optional) – Use specified edge data key as distance. For example, setting
distance=’weight’ will use the edge weight to measure the distance from the node n.

Notes

For directed graphs D this produces the “out” neighborhood or successors. If you want the neighborhood of
predecessors first reverse the graph with D.reverse(). If you want both directions use the keyword argument
undirected=True.
Node, edge, and graph attributes are copied to the returned subgraph.

5.14 Stochastic

Functions for generating stochastic graphs from a given weighted directed graph.

stochastic_graph(G[, copy, weight]) Returns a right-stochastic representation of directed


graph G.

5.14.1 networkx.generators.stochastic.stochastic_graph

stochastic_graph(G, copy=True, weight=’weight’)


Returns a right-stochastic representation of directed graph G.
A right-stochastic graph is a weighted digraph in which for each node, the sum of the weights of all the out-edges
of that node is 1. If the graph is already weighted (for example, via a ‘weight’ edge attribute), the reweighting
takes that into account.
Parameters
• G (directed graph) – A DiGraph or MultiDiGraph.
• copy (boolean, optional) – If this is True, then this function returns a new graph with the
stochastic reweighting. Otherwise, the original graph is modified in-place (and also re-
turned, for convenience).

560 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

• weight (edge attribute key (optional, default=’weight’)) – Edge attribute key used for read-
ing the existing weight and setting the new weight. If no attribute with this key is found for
an edge, then the edge weight is assumed to be 1. If an edge has a weight, it must be a a
positive number.

5.15 Intersection

Generators for random intersection graphs.

uniform_random_intersection_graph(n, m, Returns a uniform random intersection graph.


p[, . . . ])
k_random_intersection_graph(n, m, k[, Returns a intersection graph with randomly chosen at-
seed]) tribute sets for each node that are of equal size (k).
general_random_intersection_graph(n, m, Returns a random intersection graph with independent
p[, . . . ]) probabilities for connections between node and attribute
sets.

5.15.1 networkx.generators.intersection.uniform_random_intersection_graph

uniform_random_intersection_graph(n, m, p, seed=None)
Returns a uniform random intersection graph.
Parameters
• n (int) – The number of nodes in the first bipartite set (nodes)
• m (int) – The number of nodes in the second bipartite set (attributes)
• p (float) – Probability of connecting nodes between bipartite sets
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
See also:
gnp_random_graph()

References

5.15.2 networkx.generators.intersection.k_random_intersection_graph

k_random_intersection_graph(n, m, k, seed=None)
Returns a intersection graph with randomly chosen attribute sets for each node that are of equal size (k).
Parameters
• n (int) – The number of nodes in the first bipartite set (nodes)
• m (int) – The number of nodes in the second bipartite set (attributes)
• k (float) – Size of attribute set to assign to each node.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.

5.15. Intersection 561


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
gnp_random_graph(), uniform_random_intersection_graph()

References

5.15.3 networkx.generators.intersection.general_random_intersection_graph

general_random_intersection_graph(n, m, p, seed=None)
Returns a random intersection graph with independent probabilities for connections between node and attribute
sets.
Parameters
• n (int) – The number of nodes in the first bipartite set (nodes)
• m (int) – The number of nodes in the second bipartite set (attributes)
• p (list of floats of length m) – Probabilities for connecting nodes to each attribute
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
See also:
gnp_random_graph(), uniform_random_intersection_graph()

References

5.16 Social Networks

Famous social networks.

karate_club_graph() Returns Zachary’s Karate Club graph.


davis_southern_women_graph() Returns Davis Southern women social network.
florentine_families_graph() Returns Florentine families graph.
les_miserables_graph() Returns coappearance network of characters in the novel
Les Miserables.

5.16.1 networkx.generators.social.karate_club_graph

karate_club_graph()
Returns Zachary’s Karate Club graph.
Each node in the returned graph has a node attribute ‘club’ that indicates the name of the club to which the
member represented by that node belongs, either ‘Mr. Hi’ or ‘Officer’.

Examples

To get the name of the club to which a node belongs:

>>> import networkx as nx


>>> G = nx.karate_club_graph()
(continues on next page)

562 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> G.nodes[5]['club']
'Mr. Hi'
>>> G.nodes[9]['club']
'Officer'

References

5.16.2 networkx.generators.social.davis_southern_women_graph

davis_southern_women_graph()
Returns Davis Southern women social network.
This is a bipartite graph.

References

5.16.3 networkx.generators.social.florentine_families_graph

florentine_families_graph()
Returns Florentine families graph.

References

5.16.4 networkx.generators.social.les_miserables_graph

les_miserables_graph()
Returns coappearance network of characters in the novel Les Miserables.

References

5.17 Community

Generators for classes of graphs used in studying social networks.

caveman_graph(l, k) Returns a caveman graph of l cliques of size k.


connected_caveman_graph(l, k) Returns a connected caveman graph of l cliques of size
k.
relaxed_caveman_graph(l, k, p[, seed]) Returns a relaxed caveman graph.
random_partition_graph(sizes, p_in, p_out[, Returns the random partition graph with a partition of
. . . ]) sizes.
planted_partition_graph(l, k, p_in, p_out[, Returns the planted l-partition graph.
. . . ])
gaussian_random_partition_graph(n, s, v, Generate a Gaussian random partition graph.
...)
ring_of_cliques(num_cliques, clique_size) Defines a “ring of cliques” graph.
stochastic_block_model(sizes, p[, nodelist, Returns a stochastic block model graph.
. . . ])
Continued on next page

5.17. Community 563


NetworkX Reference, Release 2.3rc1.dev20190222214247

Table 17 – continued from previous page


windmill_graph(n, k) Generate a windmill graph.

5.17.1 networkx.generators.community.caveman_graph

caveman_graph(l, k)
Returns a caveman graph of l cliques of size k.
Parameters
• l (int) – Number of cliques
• k (int) – Size of cliques
Returns G – caveman graph
Return type NetworkX Graph

Notes

This returns an undirected graph, it can be converted to a directed graph using nx.to_directed(), or a
multigraph using nx.MultiGraph(nx.caveman_graph(l, k)). Only the undirected version is de-
scribed in1 and it is unclear which of the directed generalizations is most useful.

Examples

>>> G = nx.caveman_graph(3, 3)

See also:
connected_caveman_graph()

References

5.17.2 networkx.generators.community.connected_caveman_graph

connected_caveman_graph(l, k)
Returns a connected caveman graph of l cliques of size k.
The connected caveman graph is formed by creating n cliques of size k, then a single edge in each clique is
rewired to a node in an adjacent clique.
Parameters
• l (int) – number of cliques
• k (int) – size of cliques
Returns G – connected caveman graph
Return type NetworkX Graph
1 Watts, D. J. ‘Networks, Dynamics, and the Small-World Phenomenon.’ Amer. J. Soc. 105, 493-527, 1999.

564 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This returns an undirected graph, it can be converted to a directed graph using nx.to_directed(), or a
multigraph using nx.MultiGraph(nx.caveman_graph(l, k)). Only the undirected version is de-
scribed in1 and it is unclear which of the directed generalizations is most useful.

Examples

>>> G = nx.connected_caveman_graph(3, 3)

References

5.17.3 networkx.generators.community.relaxed_caveman_graph

relaxed_caveman_graph(l, k, p, seed=None)
Returns a relaxed caveman graph.
A relaxed caveman graph starts with l cliques of size k. Edges are then randomly rewired with probability p to
link different cliques.
Parameters
• l (int) – Number of groups
• k (int) – Size of cliques
• p (float) – Probabilty of rewiring each edge.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G – Relaxed Caveman Graph
Return type NetworkX Graph
Raises NetworkXError: – If p is not in [0,1]

Examples

>>> G = nx.relaxed_caveman_graph(2, 3, 0.1, seed=42)

References

5.17.4 networkx.generators.community.random_partition_graph

random_partition_graph(sizes, p_in, p_out, seed=None, directed=False)


Returns the random partition graph with a partition of sizes.
A partition graph is a graph of communities with sizes defined by s in sizes. Nodes in the same group are
connected with probability p_in and nodes of different groups are connected with probability p_out.
Parameters
• sizes (list of ints) – Sizes of groups
1 Watts, D. J. ‘Networks, Dynamics, and the Small-World Phenomenon.’ Amer. J. Soc. 105, 493-527, 1999.

5.17. Community 565


NetworkX Reference, Release 2.3rc1.dev20190222214247

• p_in (float) – probability of edges with in groups


• p_out (float) – probability of edges between groups
• directed (boolean optional, default=False) – Whether to create a directed graph
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G – random partition graph of size sum(gs)
Return type NetworkX Graph or DiGraph
Raises NetworkXError – If p_in or p_out is not in [0,1]

Examples

>>> G = nx.random_partition_graph([10,10,10],.25,.01)
>>> len(G)
30
>>> partition = G.graph['partition']
>>> len(partition)
3

Notes

This is a generalization of the planted-l-partition described in1 . It allows for the creation of groups of any size.
The partition is store as a graph attribute ‘partition’.

References

5.17.5 networkx.generators.community.planted_partition_graph

planted_partition_graph(l, k, p_in, p_out, seed=None, directed=False)


Returns the planted l-partition graph.
This model partitions a graph with n=l*k vertices in l groups with k vertices each. Vertices of the same group
are linked with a probability p_in, and vertices of different groups are linked with probability p_out.
Parameters
• l (int) – Number of groups
• k (int) – Number of vertices in each group
• p_in (float) – probability of connecting vertices within a group
• p_out (float) – probability of connected vertices between groups
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• directed (bool,optional (default=False)) – If True return a directed graph
Returns G – planted l-partition graph
Return type NetworkX Graph or DiGraph
1 Santo Fortunato ‘Community Detection in Graphs’ Physical Reports Volume 486, Issue 3-5 p. 75-174. https://arxiv.org/abs/0906.0612

566 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

Raises NetworkXError: – If p_in,p_out are not in [0,1] or

Examples

>>> G = nx.planted_partition_graph(4, 3, 0.5, 0.1, seed=42)

See also:
random_partition_model()

References

5.17.6 networkx.generators.community.gaussian_random_partition_graph

gaussian_random_partition_graph(n, s, v, p_in, p_out, directed=False, seed=None)


Generate a Gaussian random partition graph.
A Gaussian random partition graph is created by creating k partitions each with a size drawn from a normal
distribution with mean s and variance s/v. Nodes are connected within clusters with probability p_in and between
clusters with probability p_out[1]
Parameters
• n (int) – Number of nodes in the graph
• s (float) – Mean cluster size
• v (float) – Shape parameter. The variance of cluster size distribution is s/v.
• p_in (float) – Probabilty of intra cluster connection.
• p_out (float) – Probability of inter cluster connection.
• directed (boolean, optional default=False) – Whether to create a directed graph or not
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G – gaussian random partition graph
Return type NetworkX Graph or DiGraph
Raises NetworkXError – If s is > n If p_in or p_out is not in [0,1]

Notes

Note the number of partitions is dependent on s,v and n, and that the last partition may be considerably smaller,
as it is sized to simply fill out the nodes [1]
See also:
random_partition_graph()

Examples

>>> G = nx.gaussian_random_partition_graph(100,10,10,.25,.1)
>>> len(G)
100

5.17. Community 567


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

5.17.7 networkx.generators.community.ring_of_cliques

ring_of_cliques(num_cliques, clique_size)
Defines a “ring of cliques” graph.
A ring of cliques graph is consisting of cliques, connected through single links. Each clique is a complete graph.
Parameters
• num_cliques (int) – Number of cliques
• clique_size (int) – Size of cliques
Returns G – ring of cliques graph
Return type NetworkX Graph
Raises NetworkXError – If the number of cliques is lower than 2 or if the size of cliques is
smaller than 2.

Examples

>>> G = nx.ring_of_cliques(8, 4)

See also:
connected_caveman_graph()

Notes

The connected_caveman_graph graph removes a link from each clique to connect it with the next clique.
Instead, the ring_of_cliques graph simply adds the link without removing any link from the cliques.

5.17.8 networkx.generators.community.stochastic_block_model

stochastic_block_model(sizes, p, nodelist=None, seed=None, directed=False, selfloops=False,


sparse=True)
Returns a stochastic block model graph.
This model partitions the nodes in blocks of arbitrary sizes, and places edges between pairs of nodes indepen-
dently, with a probability that depends on the blocks.
Parameters
• sizes (list of ints) – Sizes of blocks
• p (list of list of floats) – Element (r,s) gives the density of edges going from the nodes of
group r to nodes of group s. p must match the number of groups (len(sizes) == len(p)), and
it must be symmetric if the graph is undirected.
• nodelist (list, optional) – The block tags are assigned according to the node identifiers in
nodelist. If nodelist is None, then the ordering is the range [0,sum(sizes)-1].
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
• directed (boolean optional, default=False) – Whether to create a directed graph or not.

568 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

• selfloops (boolean optional, default=False) – Whether to include self-loops or not.


• sparse (boolean optional, default=True) – Use the sparse heuristic to speed up the generator.
Returns g – Stochastic block model graph of size sum(sizes)
Return type NetworkX Graph or DiGraph
Raises NetworkXError – If probabilities are not in [0,1]. If the probability matrix is not square
(directed case). If the probability matrix is not symmetric (undirected case). If the sizes list does
not match nodelist or the probability matrix. If nodelist contains duplicate.

Examples

>>> sizes = [75, 75, 300]


>>> probs = [[0.25, 0.05, 0.02],
... [0.05, 0.35, 0.07],
... [0.02, 0.07, 0.40]]
>>> g = nx.stochastic_block_model(sizes, probs, seed=0)
>>> len(g)
450
>>> H = nx.quotient_graph(g, g.graph['partition'], relabel=True)
>>> for v in H.nodes(data=True):
... print(round(v[1]['density'], 3))
...
0.245
0.348
0.405
>>> for v in H.edges(data=True):
... print(round(1.0 * v[2]['weight'] / (sizes[v[0]] * sizes[v[1]]), 3))
...
0.051
0.022
0.07

See also:
random_partition_graph(), planted_partition_graph(), gaussian_random_partition_graph(),
gnp_random_graph()

References

5.17.9 networkx.generators.community.windmill_graph

windmill_graph(n, k)
Generate a windmill graph. A windmill graph is a graph of n cliques each of size k that are all joined at one
node. It can be thought of as taking a disjoint union of n cliques of size k, selecting one point from each, and
contracting all of the selected points. Alternatively, one could generate n cliques of size k-1 and one node that
is connected to all other nodes in the graph.
Parameters
• n (int) – Number of cliques
• k (int) – Size of cliques
Returns G – windmill graph with n cliques of size k
Return type NetworkX Graph

5.17. Community 569


NetworkX Reference, Release 2.3rc1.dev20190222214247

Raises NetworkXError – If the number of cliques is less than two If the size of the cliques are
less than two

Examples

>>> G = nx.windmill_graph(4, 5)

Notes

The node labeled 0 will be the node connected to all other nodes. Note that windmill graphs are usually denoted
Wd(k,n), so the parameters are in the opposite order as the parameters of this method.

5.18 Spectral

Generates graphs with a given eigenvector structure

spectral_graph_forge(G, alpha[, . . . ]) Returns a random simple graph with spectrum resem-


bling that of G

5.18.1 networkx.generators.spectral_graph_forge.spectral_graph_forge

spectral_graph_forge(G, alpha, transformation=’identity’, seed=None)


Returns a random simple graph with spectrum resembling that of G
This algorithm, called Spectral Graph Forge (SGF), computes the eigenvectors of a given graph adjacency ma-
trix, filters them and builds a random graph with a similar eigenstructure. SGF has been proved to be particularly
useful for synthesizing realistic social networks and it can also be used to anonymize graph sensitive data.
Parameters
• G (Graph)
• alpha (float) – Ratio representing the percentage of eigenvectors of G to consider, values in
[0,1].
• transformation (string, optional) – Represents the intended matrix linear transformation,
possible values are ‘identity’ and ‘modularity’
• seed (integer, random_state, or None (default)) – Indicator of numpy random number gen-
eration state. See Randomness.
Returns H – A graph with a similar eigenvector structure of the input one.
Return type Graph
Raises NetworkXError – If transformation has a value different from ‘identity’ or ‘modularity’

Notes

Spectral Graph Forge (SGF) generates a random simple graph resembling the global properties of the given
one. It leverages the low-rank approximation of the associated adjacency matrix driven by the alpha precision

570 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

parameter. SGF preserves the number of nodes of the input graph and their ordering. This way, nodes of output
graphs resemble the properties of the input one and attributes can be directly mapped.
It considers the graph adjacency matrices which can optionally be transformed to other symmetric real matrices
(currently transformation options include identity and modularity). The modularity transformation, in the sense
of Newman’s modularity matrix allows the focusing on community structure related properties of the graph.
SGF applies a low-rank approximation whose fixed rank is computed from the ratio alpha of the input graph
adjacency matrix dimension. This step performs a filtering on the input eigenvectors similar to the low pass
filtering common in telecommunications.
The filtered values (after truncation) are used as input to a Bernoulli sampling for constructing a random adja-
cency matrix.

References

Examples

>>> import networkx as nx


>>> G = nx.karate_club_graph()
>>> H = nx.spectral_graph_forge(G, 0.3)
>>>

5.19 Trees

Functions for generating trees.

random_tree(n[, seed]) Returns a uniformly random tree on n nodes.


prefix_tree(paths) Creates a directed prefix tree from the given list of iter-
ables.

5.19.1 networkx.generators.trees.random_tree

random_tree(n, seed=None)
Returns a uniformly random tree on n nodes.
Parameters
• n (int) – A positive integer representing the number of nodes in the tree.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns A tree, given as an undirected graph, whose nodes are numbers in the set {0, . . . , n - 1}.
Return type NetworkX graph
Raises NetworkXPointlessConcept – If n is zero (because the null graph is not a tree).

Notes

The current implementation of this function generates a uniformly random Prüfer sequence then converts that
to a tree via the from_prufer_sequence() function. Since there is a bijection between Prüfer sequences

5.19. Trees 571


NetworkX Reference, Release 2.3rc1.dev20190222214247

of length n - 2 and trees on n nodes, the tree is chosen uniformly at random from the set of all trees on n nodes.

5.19.2 networkx.generators.trees.prefix_tree

prefix_tree(paths)
Creates a directed prefix tree from the given list of iterables.
Parameters paths (iterable of lists) – An iterable over “paths”, which are themselves lists of nodes.
Common prefixes among these paths are converted into common initial segments in the gener-
ated tree.
Most commonly, this may be an iterable over lists of integers, or an iterable over Python strings.
Returns
• T (DiGraph) – A directed graph representing an arborescence consisting of the prefix tree
generated by paths. Nodes are directed “downward”, from parent to child. A special
“synthetic” root node is added to be the parent of the first node in each path. A special
“synthetic” leaf node, the “nil” node, is added to be the child of all nodes representing
the last element in a path. (The addition of this nil node technically makes this not an
arborescence but a directed acyclic graph; removing the nil node makes it an arborescence.)
Each node has an attribute ‘source’ whose value is the original element of the path to which
this node corresponds. The ‘source’ of the root node is None, and the ‘source’ of the nil
node is NIL.
The root node is the only node of in-degree zero in the graph, and the nil node is the only
node of out-degree zero. For convenience, the nil node can be accessed via the NIL attribute;
for example:

>>> from networkx.generators.trees import NIL


>>> paths = ['ab', 'abs', 'ad']
>>> T, root = nx.prefix_tree(paths)
>>> T.predecessors(NIL)

• root (string) – The randomly generated uuid of the root node.

Notes

The prefix tree is also known as a trie.

Examples

Create a prefix tree from a list of strings with some common prefixes:

>>> strings = ['ab', 'abs', 'ad']


>>> T, root = nx.prefix_tree(strings)

Continuing the above example, to recover the original paths that generated the prefix tree, traverse up the tree
from the NIL node to the root:

>>> from networkx.generators.trees import NIL


>>>
>>> strings = ['ab', 'abs', 'ad']
>>> T, root = nx.prefix_tree(strings)
(continues on next page)

572 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> recovered = []
>>> for v in T.predecessors(NIL):
... s = ''
... while v != root:
... # Prepend the character `v` to the accumulator `s`.
... s = str(T.node[v]['source']) + s
... # Each non-nil, non-root node has exactly one parent.
... v = next(T.predecessors(v))
... recovered.append(s)
>>> sorted(recovered)
['ab', 'abs', 'ad']

5.20 Non Isomorphic Trees

Implementation of the Wright, Richmond, Odlyzko and McKay (WROM) algorithm for the enumeration of all non-
isomorphic free trees of a given order. Rooted trees are represented by level sequences, i.e., lists in which the i-th
element specifies the distance of vertex i to the root.

nonisomorphic_trees(order[, create]) Returns a list of nonisomporphic trees


number_of_nonisomorphic_trees(order) Returns the number of nonisomorphic trees

5.20.1 networkx.generators.nonisomorphic_trees.nonisomorphic_trees

nonisomorphic_trees(order, create=’graph’)
Returns a list of nonisomporphic trees
Parameters
• order (int) – order of the desired tree(s)
• create (graph or matrix (default=”Graph)) – If graph is selected a list of trees will be
returned, if matrix is selected a list of adjancency matrix will be returned
Returns
• G (List of NetworkX Graphs)
• M (List of Adjacency matrices)

References

5.20.2 networkx.generators.nonisomorphic_trees.number_of_nonisomorphic_trees

number_of_nonisomorphic_trees(order)
Returns the number of nonisomorphic trees
Parameters order (int) – order of the desired tree(s)
Returns length
Return type Number of nonisomorphic graphs for the given order

5.20. Non Isomorphic Trees 573


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

5.21 Triads

Functions that generate the triad graphs, that is, the possible digraphs on three nodes.

triad_graph(triad_name) Returns the triad graph with the given name.

5.21.1 networkx.generators.triads.triad_graph

triad_graph(triad_name)
Returns the triad graph with the given name.
Each string in the following tuple is a valid triad name:

('003', '012', '102', '021D', '021U', '021C', '111D', '111U',


'030T', '030C', '201', '120D', '120U', '120C', '210', '300')

Each triad name corresponds to one of the possible valid digraph on three nodes.
Parameters triad_name (string) – The name of a triad, as described above.
Returns The digraph on three nodes with the given name. The nodes of the graph are the single-
character strings ‘a’, ‘b’, and ‘c’.
Return type DiGraph
Raises ValueError – If triad_name is not the name of a triad.
See also:
triadic_census()

5.22 Joint Degree Sequence

Generate graphs with a given joint degree

is_valid_joint_degree(joint_degrees) Checks whether the given joint degree dictionary is re-


alizable as a simple graph.
joint_degree_graph(joint_degrees[, seed]) Generates a random simple graph with the given joint
degree dictionary.

5.22.1 networkx.generators.joint_degree_seq.is_valid_joint_degree

is_valid_joint_degree(joint_degrees)
Checks whether the given joint degree dictionary is realizable as a simple graph.
A joint degree dictionary is a dictionary of dictionaries, in which entry joint_degrees[k][l] is an integer
representing the number of edges joining nodes of degree k with nodes of degree l. Such a dictionary is realizable
as a simple graph if and only if the following conditions are satisfied.
• each entry must be an integer,

574 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

• the total number of nodes of degree k, computed by sum(joint_degrees[k].values()) / k,


must be an integer,
• the total number of edges joining nodes of degree k with nodes of degree l cannot exceed the total
number of possible edges,
• each diagonal entry joint_degrees[k][k] must be even (this is a convention assumed by the
joint_degree_graph() function).

Parameters joint_degrees (dictionary of dictionary of integers) – A joint degree dictionary in which


entry joint_degrees[k][l] is the number of edges joining nodes of degree k with nodes
of degree l.
Returns Whether the given joint degree dictionary is realizable as a simple graph.
Return type bool

References

5.22.2 networkx.generators.joint_degree_seq.joint_degree_graph

joint_degree_graph(joint_degrees, seed=None)
Generates a random simple graph with the given joint degree dictionary.
Parameters
• joint_degrees (dictionary of dictionary of integers) – A joint degree dictionary in which
entry joint_degrees[k][l] is the number of edges joining nodes of degree k with
nodes of degree l.
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns G – A graph with the specified joint degree dictionary.
Return type Graph
Raises NetworkXError – If joint_degrees dictionary is not realizable.

Notes

In each iteration of the “while loop” the algorithm picks two disconnected nodes v and w, of degree k and l
correspondingly, for which joint_degrees[k][l] has not reached its target yet. It then adds edge (v, w)
and increases the number of edges in graph G by one.
The intelligence of the algorithm lies in the fact that it is always possible to add an edge between such discon-
nected nodes v and w, even if one or both nodes do not have free stubs. That is made possible by executing a
“neighbor switch”, an edge rewiring move that releases a free stub while keeping the joint degree of G the same.
The algorithm continues for E (number of edges) iterations of the “while loop”, at the which point all entries of
the given joint_degrees[k][l] have reached their target values and the construction is complete.

5.22. Joint Degree Sequence 575


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

Examples

>>> import networkx as nx


>>> joint_degrees = {1: {4: 1},
... 2: {2: 2, 3: 2, 4: 2},
... 3: {2: 2, 4: 1},
... 4: {1: 1, 2: 2, 3: 1}}
>>> G=nx.joint_degree_graph(joint_degrees)
>>>

5.23 Mycielski

Functions related to the Mycielski Operation and the Mycielskian family of graphs.

mycielskian(G[, iterations]) Returns the Mycielskian of a simple, undirected graph


G
mycielski_graph(n) Generator for the n_th Mycielski Graph.

5.23.1 networkx.generators.mycielski.mycielskian

mycielskian(G, iterations=1)
Returns the Mycielskian of a simple, undirected graph G
The Mycielskian of graph preserves a graph’s triangle free property while increasing the chromatic number by
1.
The Mycielski Operation on a graph, 𝐺 = (𝑉, 𝐸), constructs a new graph with 2|𝑉 | + 1 nodes and 3|𝐸| + |𝑉 |
edges.
The construction is as follows:
Let 𝑉 = 0, ..., 𝑛⋃︀− 1.⋃︀ Construct another vertex set 𝑈 = 𝑛, ..., 2𝑛 and a vertex, w. Construct a new graph, M,
with vertices 𝑈 𝑉 𝑤. For edges, (𝑢, 𝑣) ∈ 𝐸 add edges (𝑢, 𝑣), (𝑢, 𝑣 + 𝑛), and (𝑢 + 𝑛, 𝑣) to M. Finally, for
all vertices 𝑢 ∈ 𝑈 , add edge (𝑢, 𝑤) to M.
The Mycielski Operation can be done multiple times by repeating the above process iteratively.
More information can be found at https://en.wikipedia.org/wiki/Mycielskian
Parameters
• G (graph) – A simple, undirected NetworkX graph
• iterations (int) – The number of iterations of the Mycielski operation to perform on G.
Defaults to 1. Must be a non-negative integer.
Returns M – The Mycielskian of G after the specified number of iterations.
Return type graph

Notes

Graph, node, and edge data are not necessarily propagated to the new graph.

576 Chapter 5. Graph generators


NetworkX Reference, Release 2.3rc1.dev20190222214247

5.23.2 networkx.generators.mycielski.mycielski_graph

mycielski_graph(n)
Generator for the n_th Mycielski Graph.
The Mycielski family of graphs is an infinite set of graphs. 𝑀1 is the singleton graph, 𝑀2 is two vertices with
an edge, and, for 𝑖 > 2, 𝑀𝑖 is the Mycielskian of 𝑀𝑖−1 .
More information can be found at http://mathworld.wolfram.com/MycielskiGraph.html
Parameters n (int) – The desired Mycielski Graph.
Returns M – The n_th Mycielski Graph
Return type graph

Notes

The first graph in the Mycielski sequence is the singleton graph. The Mycielskian of this graph is not the 𝑃2
graph, but rather the 𝑃2 graph with an extra, isolated vertex. The second Mycielski graph is the 𝑃2 graph, so the
first two are hard coded. The remaining graphs are generated using the Mycielski operation.

5.23. Mycielski 577


NetworkX Reference, Release 2.3rc1.dev20190222214247

578 Chapter 5. Graph generators


CHAPTER

SIX

LINEAR ALGEBRA

6.1 Graph Matrix

Adjacency matrix and incidence matrix of graphs.

adjacency_matrix(G[, nodelist, weight]) Returns adjacency matrix of G.


incidence_matrix(G[, nodelist, edgelist, . . . ]) Returns incidence matrix of G.

6.1.1 networkx.linalg.graphmatrix.adjacency_matrix

adjacency_matrix(G, nodelist=None, weight=’weight’)


Returns adjacency matrix of G.
Parameters
• G (graph) – A NetworkX graph
• nodelist (list, optional) – The rows and columns are ordered according to the nodes in
nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• weight (string or None, optional (default=’weight’)) – The edge data key used to provide
each value in the matrix. If None, then each edge has weight 1.
Returns A – Adjacency matrix representation of G.
Return type SciPy sparse matrix

Notes

For directed graphs, entry i,j corresponds to an edge from i to j.


If you want a pure Python adjacency matrix representation try networkx.convert.to_dict_of_dicts which will
return a dictionary-of-dictionaries format that can be addressed as a sparse matrix.
For MultiGraph/MultiDiGraph with parallel edges the weights are summed. See to_numpy_matrix for other
options.
The convention used for self-loop edges in graphs is to assign the diagonal matrix entry value to the edge weight
attribute (or the number 1 if the edge has no weight attribute). If the alternate convention of doubling the edge
weight is desired the resulting Scipy sparse matrix can be modified as follows:

579
NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> import scipy as sp


>>> G = nx.Graph([(1,1)])
>>> A = nx.adjacency_matrix(G)
>>> print(A.todense())
[[1]]
>>> A.setdiag(A.diagonal()*2)
>>> print(A.todense())
[[2]]

See also:
to_numpy_matrix(), to_scipy_sparse_matrix(), to_dict_of_dicts()

6.1.2 networkx.linalg.graphmatrix.incidence_matrix

incidence_matrix(G, nodelist=None, edgelist=None, oriented=False, weight=None)


Returns incidence matrix of G.
The incidence matrix assigns each row to a node and each column to an edge. For a standard incidence matrix a
1 appears wherever a row’s node is incident on the column’s edge. For an oriented incidence matrix each edge
is assigned an orientation (arbitrarily for undirected and aligning to direction for directed). A -1 appears for the
tail of an edge and 1 for the head of the edge. The elements are zero otherwise.
Parameters
• G (graph) – A NetworkX graph
• nodelist (list, optional (default= all nodes in G)) – The rows are ordered according to the
nodes in nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• edgelist (list, optional (default= all edges in G)) – The columns are ordered according to
the edges in edgelist. If edgelist is None, then the ordering is produced by G.edges().
• oriented (bool, optional (default=False)) – If True, matrix elements are +1 or -1 for the
head or tail node respectively of each edge. If False, +1 occurs at both nodes.
• weight (string or None, optional (default=None)) – The edge data key used to provide each
value in the matrix. If None, then each edge has weight 1. Edge weights, if used, should be
positive so that the orientation can provide the sign.
Returns A – The incidence matrix of G.
Return type SciPy sparse matrix

Notes

For MultiGraph/MultiDiGraph, the edges in edgelist should be (u,v,key) 3-tuples.


“Networks are the best discrete model for so many problems in applied mathematics”1 .

References

6.2 Laplacian Matrix

Laplacian matrix of graphs.


1 Gil Strang, Network applications: A = incidence matrix, http://academicearth.org/lectures/network-applications-incidence-matrix

580 Chapter 6. Linear algebra


NetworkX Reference, Release 2.3rc1.dev20190222214247

laplacian_matrix(G[, nodelist, weight]) Returns the Laplacian matrix of G.


normalized_laplacian_matrix(G[, nodelist, Returns the normalized Laplacian matrix of G.
. . . ])
directed_laplacian_matrix(G[, nodelist, Returns the directed Laplacian matrix of G.
. . . ])

6.2.1 networkx.linalg.laplacianmatrix.laplacian_matrix

laplacian_matrix(G, nodelist=None, weight=’weight’)


Returns the Laplacian matrix of G.
The graph Laplacian is the matrix L = D - A, where A is the adjacency matrix and D is the diagonal matrix of
node degrees.
Parameters
• G (graph) – A NetworkX graph
• nodelist (list, optional) – The rows and columns are ordered according to the nodes in
nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• weight (string or None, optional (default=’weight’)) – The edge data key used to compute
each value in the matrix. If None, then each edge has weight 1.
Returns L – The Laplacian matrix of G.
Return type SciPy sparse matrix

Notes

For MultiGraph/MultiDiGraph, the edges weights are summed.


See also:
to_numpy_matrix(), normalized_laplacian_matrix()

6.2.2 networkx.linalg.laplacianmatrix.normalized_laplacian_matrix

normalized_laplacian_matrix(G, nodelist=None, weight=’weight’)


Returns the normalized Laplacian matrix of G.
The normalized graph Laplacian is the matrix

𝑁 = 𝐷−1/2 𝐿𝐷−1/2

where L is the graph Laplacian and D is the diagonal matrix of node degrees.
Parameters
• G (graph) – A NetworkX graph
• nodelist (list, optional) – The rows and columns are ordered according to the nodes in
nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• weight (string or None, optional (default=’weight’)) – The edge data key used to compute
each value in the matrix. If None, then each edge has weight 1.
Returns N – The normalized Laplacian matrix of G.

6.2. Laplacian Matrix 581


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type NumPy matrix

Notes

For MultiGraph/MultiDiGraph, the edges weights are summed. See to_numpy_matrix for other options.
If the Graph contains selfloops, D is defined as diag(sum(A,1)), where A is the adjacency matrix2 .
See also:
laplacian_matrix()

References

6.2.3 networkx.linalg.laplacianmatrix.directed_laplacian_matrix

directed_laplacian_matrix(G, nodelist=None, weight=’weight’, walk_type=None, alpha=0.95)


Returns the directed Laplacian matrix of G.
The graph directed Laplacian is the matrix

𝐿 = 𝐼 − (Φ1/2 𝑃 Φ−1/2 + Φ−1/2 𝑃 𝑇 Φ1/2 )/2

where I is the identity matrix, P is the transition matrix of the graph, and Phi a matrix with the Perron vector
of P in the diagonal and zeros elsewhere.
Depending on the value of walk_type, P can be the transition matrix induced by a random walk, a lazy random
walk, or a random walk with teleportation (PageRank).
Parameters
• G (DiGraph) – A NetworkX graph
• nodelist (list, optional) – The rows and columns are ordered according to the nodes in
nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• weight (string or None, optional (default=’weight’)) – The edge data key used to compute
each value in the matrix. If None, then each edge has weight 1.
• walk_type (string or None, optional (default=None)) – If None, P is selected depending on
the properties of the graph. Otherwise is one of ‘random’, ‘lazy’, or ‘pagerank’
• alpha (real) – (1 - alpha) is the teleportation probability used with pagerank
Returns L – Normalized Laplacian of G.
Return type NumPy array

Notes

Only implemented for DiGraphs


See also:
laplacian_matrix()
2Steve Butler, Interlacing For Weighted Graphs Using The Normalized Laplacian, Electronic Journal of Linear Algebra, Volume 16, pp. 90-98,
March 2007.

582 Chapter 6. Linear algebra


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

6.3 Spectrum

Eigenvalue spectrum of graphs.

laplacian_spectrum(G[, weight]) Returns eigenvalues of the Laplacian of G


adjacency_spectrum(G[, weight]) Returns eigenvalues of the adjacency matrix of G.
modularity_spectrum(G) Returns eigenvalues of the modularity matrix of G.

6.3.1 networkx.linalg.spectrum.laplacian_spectrum

laplacian_spectrum(G, weight=’weight’)
Returns eigenvalues of the Laplacian of G
Parameters
• G (graph) – A NetworkX graph
• weight (string or None, optional (default=’weight’)) – The edge data key used to compute
each value in the matrix. If None, then each edge has weight 1.
Returns evals – Eigenvalues
Return type NumPy array

Notes

For MultiGraph/MultiDiGraph, the edges weights are summed. See to_numpy_matrix for other options.
See also:
laplacian_matrix()

6.3.2 networkx.linalg.spectrum.adjacency_spectrum

adjacency_spectrum(G, weight=’weight’)
Returns eigenvalues of the adjacency matrix of G.
Parameters
• G (graph) – A NetworkX graph
• weight (string or None, optional (default=’weight’)) – The edge data key used to compute
each value in the matrix. If None, then each edge has weight 1.
Returns evals – Eigenvalues
Return type NumPy array

Notes

For MultiGraph/MultiDiGraph, the edges weights are summed. See to_numpy_matrix for other options.
See also:

6.3. Spectrum 583


NetworkX Reference, Release 2.3rc1.dev20190222214247

adjacency_matrix()

6.3.3 networkx.linalg.spectrum.modularity_spectrum

modularity_spectrum(G)
Returns eigenvalues of the modularity matrix of G.
Parameters G (Graph) – A NetworkX Graph or DiGraph
Returns evals – Eigenvalues
Return type NumPy array
See also:
modularity_matrix()

References

6.4 Algebraic Connectivity

Algebraic connectivity and Fiedler vectors of undirected graphs.

algebraic_connectivity(G[, weight, . . . ]) Returns the algebraic connectivity of an undirected


graph.
fiedler_vector(G[, weight, normalized, tol, . . . ]) Returns the Fiedler vector of a connected undirected
graph.
spectral_ordering(G[, weight, normalized, . . . ]) Compute the spectral_ordering of a graph.

6.4.1 networkx.linalg.algebraicconnectivity.algebraic_connectivity

algebraic_connectivity(G, weight=’weight’, normalized=False, tol=1e-08, method=’tracemin_pcg’,


seed=None)
Returns the algebraic connectivity of an undirected graph.
The algebraic connectivity of a connected undirected graph is the second smallest eigenvalue of its Laplacian
matrix.
Parameters
• G (NetworkX graph) – An undirected graph.
• weight (object, optional (default: None)) – The data key used to determine the weight of
each edge. If None, then each edge has unit weight.
• normalized (bool, optional (default: False)) – Whether the normalized Laplacian matrix is
used.
• tol (float, optional (default: 1e-8)) – Tolerance of relative residual in eigenvalue computa-
tion.
• method (string, optional (default: ‘tracemin_pcg’)) – Method of eigenvalue computation. It
must be one of the tracemin options shown below (TraceMIN), ‘lanczos’ (Lanczos iteration)
or ‘lobpcg’ (LOBPCG).
The TraceMIN algorithm uses a linear system solver. The following values allow specifying
the solver to be used.

584 Chapter 6. Linear algebra


NetworkX Reference, Release 2.3rc1.dev20190222214247

Value Solver
‘tracemin_pcg’ Preconditioned conjugate gradient method
‘tracemin_chol’ Cholesky factorization
‘tracemin_lu’ LU factorization

• seed (integer, random_state, or None (default)) – Indicator of random number generation


state. See Randomness.
Returns algebraic_connectivity – Algebraic connectivity.
Return type float
Raises
• NetworkXNotImplemented – If G is directed.
• NetworkXError – If G has less than two nodes.

Notes

Edge weights are interpreted by their absolute values. For MultiGraph’s, weights of parallel edges are summed.
Zero-weighted edges are ignored.
To use Cholesky factorization in the TraceMIN algorithm, the scikits.sparse package must be installed.
See also:
laplacian_matrix()

6.4.2 networkx.linalg.algebraicconnectivity.fiedler_vector

fiedler_vector(G, weight=’weight’, normalized=False, tol=1e-08, method=’tracemin_pcg’,


seed=None)
Returns the Fiedler vector of a connected undirected graph.
The Fiedler vector of a connected undirected graph is the eigenvector corresponding to the second smallest
eigenvalue of the Laplacian matrix of of the graph.
Parameters
• G (NetworkX graph) – An undirected graph.
• weight (object, optional (default: None)) – The data key used to determine the weight of
each edge. If None, then each edge has unit weight.
• normalized (bool, optional (default: False)) – Whether the normalized Laplacian matrix is
used.
• tol (float, optional (default: 1e-8)) – Tolerance of relative residual in eigenvalue computa-
tion.
• method (string, optional (default: ‘tracemin_pcg’)) – Method of eigenvalue computation. It
must be one of the tracemin options shown below (TraceMIN), ‘lanczos’ (Lanczos iteration)
or ‘lobpcg’ (LOBPCG).
The TraceMIN algorithm uses a linear system solver. The following values allow specifying
the solver to be used.

6.4. Algebraic Connectivity 585


NetworkX Reference, Release 2.3rc1.dev20190222214247

Value Solver
‘tracemin_pcg’ Preconditioned conjugate gradient method
‘tracemin_chol’ Cholesky factorization
‘tracemin_lu’ LU factorization

• seed (integer, random_state, or None (default)) – Indicator of random number generation


state. See Randomness.
Returns fiedler_vector – Fiedler vector.
Return type NumPy array of floats.
Raises
• NetworkXNotImplemented – If G is directed.
• NetworkXError – If G has less than two nodes or is not connected.

Notes

Edge weights are interpreted by their absolute values. For MultiGraph’s, weights of parallel edges are summed.
Zero-weighted edges are ignored.
To use Cholesky factorization in the TraceMIN algorithm, the scikits.sparse package must be installed.
See also:
laplacian_matrix()

6.4.3 networkx.linalg.algebraicconnectivity.spectral_ordering

spectral_ordering(G, weight=’weight’, normalized=False, tol=1e-08, method=’tracemin_pcg’,


seed=None)
Compute the spectral_ordering of a graph.
The spectral ordering of a graph is an ordering of its nodes where nodes in the same weakly connected compo-
nents appear contiguous and ordered by their corresponding elements in the Fiedler vector of the component.
Parameters
• G (NetworkX graph) – A graph.
• weight (object, optional (default: None)) – The data key used to determine the weight of
each edge. If None, then each edge has unit weight.
• normalized (bool, optional (default: False)) – Whether the normalized Laplacian matrix is
used.
• tol (float, optional (default: 1e-8)) – Tolerance of relative residual in eigenvalue computa-
tion.
• method (string, optional (default: ‘tracemin_pcg’)) – Method of eigenvalue computation. It
must be one of the tracemin options shown below (TraceMIN), ‘lanczos’ (Lanczos iteration)
or ‘lobpcg’ (LOBPCG).
The TraceMIN algorithm uses a linear system solver. The following values allow specifying
the solver to be used.

586 Chapter 6. Linear algebra


NetworkX Reference, Release 2.3rc1.dev20190222214247

Value Solver
‘tracemin_pcg’ Preconditioned conjugate gradient method
‘tracemin_chol’ Cholesky factorization
‘tracemin_lu’ LU factorization

• seed (integer, random_state, or None (default)) – Indicator of random number generation


state. See Randomness.
Returns spectral_ordering – Spectral ordering of nodes.
Return type NumPy array of floats.
Raises NetworkXError – If G is empty.

Notes

Edge weights are interpreted by their absolute values. For MultiGraph’s, weights of parallel edges are summed.
Zero-weighted edges are ignored.
To use Cholesky factorization in the TraceMIN algorithm, the scikits.sparse package must be installed.
See also:
laplacian_matrix()

6.5 Attribute Matrices

Functions for constructing matrix-like objects from graph attributes.

attr_matrix(G[, edge_attr, node_attr, . . . ]) Returns a NumPy matrix using attributes from G.


attr_sparse_matrix(G[, edge_attr, . . . ]) Returns a SciPy sparse matrix using attributes from G.

6.5.1 networkx.linalg.attrmatrix.attr_matrix

attr_matrix(G, edge_attr=None, node_attr=None, normalized=False, rc_order=None, dtype=None, or-


der=None)
Returns a NumPy matrix using attributes from G.
If only G is passed in, then the adjacency matrix is constructed.
Let A be a discrete set of values for the node attribute node_attr. Then the elements of A represent the rows
and columns of the constructed matrix. Now, iterate through every edge e=(u,v) in G and consider the value
of the edge attribute edge_attr. If ua and va are the values of the node attribute node_attr for u and v,
respectively, then the value of the edge attribute is added to the matrix element at (ua, va).
Parameters
• G (graph) – The NetworkX graph used to construct the NumPy matrix.
• edge_attr (str, optional) – Each element of the matrix represents a running total of the
specified edge attribute for edges whose node attributes correspond to the rows/cols of the
matirx. The attribute must be present for all edges in the graph. If no attribute is specified,
then we just count the number of edges whose node attributes correspond to the matrix
element.

6.5. Attribute Matrices 587


NetworkX Reference, Release 2.3rc1.dev20190222214247

• node_attr (str, optional) – Each row and column in the matrix represents a particular value
of the node attribute. The attribute must be present for all nodes in the graph. Note, the
values of this attribute should be reliably hashable. So, float values are not recommended.
If no attribute is specified, then the rows and columns will be the nodes of the graph.
• normalized (bool, optional) – If True, then each row is normalized by the summation of its
values.
• rc_order (list, optional) – A list of the node attribute values. This list specifies the ordering
of rows and columns of the array. If no ordering is provided, then the ordering will be
random (and also, a return value).
Other Parameters
• dtype (NumPy data-type, optional) – A valid NumPy dtype used to initialize the array.
Keep in mind certain dtypes can yield unexpected results if the array is to be normalized.
The parameter is passed to numpy.zeros(). If unspecified, the NumPy default is used.
• order ({‘C’, ‘F’}, optional) – Whether to store multidimensional data in C- or Fortran-
contiguous (row- or column-wise) order in memory. This parameter is passed to
numpy.zeros(). If unspecified, the NumPy default is used.
Returns
• M (NumPy matrix) – The attribute matrix.
• ordering (list) – If rc_order was specified, then only the matrix is returned. However, if
rc_order was None, then the ordering used to construct the matrix is returned as well.

Examples

Construct an adjacency matrix:

>>> import numpy as np


>>> try:
... np.set_printoptions(legacy="1.13")
... except TypeError:
... pass
>>> G = nx.Graph()
>>> G.add_edge(0, 1, thickness=1, weight=3)
>>> G.add_edge(0, 2, thickness=2)
>>> G.add_edge(1, 2, thickness=3)
>>> nx.attr_matrix(G, rc_order=[0, 1, 2])
matrix([[ 0., 1., 1.],
[ 1., 0., 1.],
[ 1., 1., 0.]])

Alternatively, we can obtain the matrix describing edge thickness.

>>> nx.attr_matrix(G, edge_attr='thickness', rc_order=[0, 1, 2])


matrix([[ 0., 1., 2.],
[ 1., 0., 3.],
[ 2., 3., 0.]])

We can also color the nodes and ask for the probability distribution over all edges (u,v) describing:
Pr(v has color Y | u has color X)

588 Chapter 6. Linear algebra


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G.nodes[0]['color'] = 'red'


>>> G.nodes[1]['color'] = 'red'
>>> G.nodes[2]['color'] = 'blue'
>>> rc = ['red', 'blue']
>>> nx.attr_matrix(G, node_attr='color', normalized=True, rc_order=rc)
matrix([[ 0.33333333, 0.66666667],
[ 1. , 0. ]])

For example, the above tells us that for all edges (u,v):
Pr( v is red | u is red) = 1/3 Pr( v is blue | u is red) = 2/3
Pr( v is red | u is blue) = 1 Pr( v is blue | u is blue) = 0
Finally, we can obtain the total weights listed by the node colors.
>>> nx.attr_matrix(G, edge_attr='weight', node_attr='color', rc_order=rc)
matrix([[ 3., 2.],
[ 2., 0.]])

Thus, the total weight over all edges (u,v) with u and v having colors:
(red, red) is 3 # the sole contribution is from edge (0,1) (red, blue) is 2 # contributions from edges
(0,2) and (1,2) (blue, red) is 2 # same as (red, blue) since graph is undirected (blue, blue) is 0 # there
are no edges with blue endpoints

6.5.2 networkx.linalg.attrmatrix.attr_sparse_matrix

attr_sparse_matrix(G, edge_attr=None, node_attr=None, normalized=False, rc_order=None,


dtype=None)
Returns a SciPy sparse matrix using attributes from G.
If only G is passed in, then the adjacency matrix is constructed.
Let A be a discrete set of values for the node attribute node_attr. Then the elements of A represent the rows
and columns of the constructed matrix. Now, iterate through every edge e=(u,v) in G and consider the value
of the edge attribute edge_attr. If ua and va are the values of the node attribute node_attr for u and v,
respectively, then the value of the edge attribute is added to the matrix element at (ua, va).
Parameters
• G (graph) – The NetworkX graph used to construct the NumPy matrix.
• edge_attr (str, optional) – Each element of the matrix represents a running total of the
specified edge attribute for edges whose node attributes correspond to the rows/cols of the
matirx. The attribute must be present for all edges in the graph. If no attribute is specified,
then we just count the number of edges whose node attributes correspond to the matrix
element.
• node_attr (str, optional) – Each row and column in the matrix represents a particular value
of the node attribute. The attribute must be present for all nodes in the graph. Note, the
values of this attribute should be reliably hashable. So, float values are not recommended.
If no attribute is specified, then the rows and columns will be the nodes of the graph.
• normalized (bool, optional) – If True, then each row is normalized by the summation of its
values.
• rc_order (list, optional) – A list of the node attribute values. This list specifies the ordering
of rows and columns of the array. If no ordering is provided, then the ordering will be
random (and also, a return value).

6.5. Attribute Matrices 589


NetworkX Reference, Release 2.3rc1.dev20190222214247

Other Parameters dtype (NumPy data-type, optional) – A valid NumPy dtype used to initialize the
array. Keep in mind certain dtypes can yield unexpected results if the array is to be normalized.
The parameter is passed to numpy.zeros(). If unspecified, the NumPy default is used.
Returns
• M (SciPy sparse matrix) – The attribute matrix.
• ordering (list) – If rc_order was specified, then only the matrix is returned. However, if
rc_order was None, then the ordering used to construct the matrix is returned as well.

Examples

Construct an adjacency matrix:

>>> import numpy as np


>>> try:
... np.set_printoptions(legacy="1.13")
... except TypeError:
... pass
>>> G = nx.Graph()
>>> G.add_edge(0,1,thickness=1,weight=3)
>>> G.add_edge(0,2,thickness=2)
>>> G.add_edge(1,2,thickness=3)
>>> M = nx.attr_sparse_matrix(G, rc_order=[0,1,2])
>>> M.todense()
matrix([[ 0., 1., 1.],
[ 1., 0., 1.],
[ 1., 1., 0.]])

Alternatively, we can obtain the matrix describing edge thickness.

>>> M = nx.attr_sparse_matrix(G, edge_attr='thickness', rc_order=[0,1,2])


>>> M.todense()
matrix([[ 0., 1., 2.],
[ 1., 0., 3.],
[ 2., 3., 0.]])

We can also color the nodes and ask for the probability distribution over all edges (u,v) describing:
Pr(v has color Y | u has color X)

>>> G.nodes[0]['color'] = 'red'


>>> G.nodes[1]['color'] = 'red'
>>> G.nodes[2]['color'] = 'blue'
>>> rc = ['red', 'blue']
>>> M = nx.attr_sparse_matrix(G, node_attr='color',
˓→ normalized=True, rc_order=rc)
>>> M.todense()
matrix([[ 0.33333333, 0.66666667],
[ 1. , 0. ]])

For example, the above tells us that for all edges (u,v):
Pr( v is red | u is red) = 1/3 Pr( v is blue | u is red) = 2/3
Pr( v is red | u is blue) = 1 Pr( v is blue | u is blue) = 0
Finally, we can obtain the total weights listed by the node colors.

590 Chapter 6. Linear algebra


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> M = nx.attr_sparse_matrix(G, edge_attr='weight',


˓→ node_attr='color', rc_order=rc)
>>> M.todense()
matrix([[ 3., 2.],
[ 2., 0.]])

Thus, the total weight over all edges (u,v) with u and v having colors:
(red, red) is 3 # the sole contribution is from edge (0,1) (red, blue) is 2 # contributions from edges
(0,2) and (1,2) (blue, red) is 2 # same as (red, blue) since graph is undirected (blue, blue) is 0 # there
are no edges with blue endpoints

6.6 Modularity Matrices

Modularity matrix of graphs.

modularity_matrix(G[, nodelist, weight]) Returns the modularity matrix of G.


directed_modularity_matrix(G[, nodelist, Returns the directed modularity matrix of G.
weight])

6.6.1 networkx.linalg.modularitymatrix.modularity_matrix

modularity_matrix(G, nodelist=None, weight=None)


Returns the modularity matrix of G.
The modularity matrix is the matrix B = A - <A>, where A is the adjacency matrix and <A> is the average
adjacency matrix, assuming that the graph is described by the configuration model.
More specifically, the element B_ij of B is defined as

𝑘𝑖 𝑘𝑗
𝐴𝑖𝑗 −
2𝑚
where k_i is the degree of node i, and where m is the number of edges in the graph. When weight is set to a
name of an attribute edge, Aij, k_i, k_j and m are computed using its value.
Parameters
• G (Graph) – A NetworkX graph
• nodelist (list, optional) – The rows and columns are ordered according to the nodes in
nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used for the edge weight. If None then all edge weights are 1.
Returns B – The modularity matrix of G.
Return type Numpy matrix

Examples

>>> import networkx as nx


>>> k =[3, 2, 2, 1, 0]
(continues on next page)

6.6. Modularity Matrices 591


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> G = nx.havel_hakimi_graph(k)
>>> B = nx.modularity_matrix(G)

See also:
to_numpy_matrix(), adjacency_matrix(), laplacian_matrix(),
directed_modularity_matrix()

References

6.6.2 networkx.linalg.modularitymatrix.directed_modularity_matrix

directed_modularity_matrix(G, nodelist=None, weight=None)


Returns the directed modularity matrix of G.
The modularity matrix is the matrix B = A - <A>, where A is the adjacency matrix and <A> is the expected
adjacency matrix, assuming that the graph is described by the configuration model.
More specifically, the element B_ij of B is defined as

𝐵𝑖𝑗 = 𝐴𝑖𝑗 − 𝑘𝑖𝑜𝑢𝑡 𝑘𝑗𝑖𝑛 /𝑚

where 𝑘𝑖𝑖𝑛 is the in degree of node i, and 𝑘𝑗𝑜𝑢𝑡 is the out degree of node j, with m the number of edges in the
graph. When weight is set to a name of an attribute edge, Aij, k_i, k_j and m are computed using its value.
Parameters
• G (DiGraph) – A NetworkX DiGraph
• nodelist (list, optional) – The rows and columns are ordered according to the nodes in
nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• weight (string or None, optional (default=None)) – The edge attribute that holds the numer-
ical value used for the edge weight. If None then all edge weights are 1.
Returns B – The modularity matrix of G.
Return type Numpy matrix

Examples

>>> import networkx as nx


>>> G = nx.DiGraph()
>>> G.add_edges_from(((1,2), (1,3), (3,1), (3,2), (3,5), (4,5), (4,6),
... (5,4), (5,6), (6,4)))
>>> B = nx.directed_modularity_matrix(G)

Notes

NetworkX defines the element A_ij of the adjacency matrix as 1 if there is a link going from node i to node j.
Leicht and Newman use the opposite definition. This explains the different expression for B_ij.
See also:
to_numpy_matrix(), adjacency_matrix(), laplacian_matrix(),
modularity_matrix()

592 Chapter 6. Linear algebra


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

6.6. Modularity Matrices 593


NetworkX Reference, Release 2.3rc1.dev20190222214247

594 Chapter 6. Linear algebra


CHAPTER

SEVEN

CONVERTING TO AND FROM OTHER DATA FORMATS

7.1 To NetworkX Graph

Functions to convert NetworkX graphs to and from other formats.


The preferred way of converting data to a NetworkX graph is through the graph constructor. The constructor calls the
to_networkx_graph() function which attempts to guess the input type and convert it automatically.

Examples

Create a graph with a single edge from a dictionary of dictionaries

>>> d={0: {1: 1}} # dict-of-dicts single edge (0,1)


>>> G=nx.Graph(d)

See also:
nx_agraph, nx_pydot

to_networkx_graph(data[, create_using, . . . ]) Make a NetworkX graph from a known data structure.

7.1.1 networkx.convert.to_networkx_graph

to_networkx_graph(data, create_using=None, multigraph_input=False)


Make a NetworkX graph from a known data structure.
The preferred way to call this is automatically from the class constructor

>>> d = {0: {1: {'weight':1}}} # dict-of-dicts single edge (0,1)


>>> G = nx.Graph(d)

instead of the equivalent

>>> G = nx.from_dict_of_dicts(d)

Parameters
• data (object to be converted) –
Current known types are: any NetworkX graph dict-of-dicts dict-of-lists list of edges
Pandas DataFrame (row per edge) numpy matrix numpy ndarray scipy sparse matrix py-
graphviz agraph

595
NetworkX Reference, Release 2.3rc1.dev20190222214247

• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to


create. If graph instance, then cleared before populated.
• multigraph_input (bool (default False)) – If True and data is a dict_of_dicts, try to create
a multigraph assuming dict_of_dict_of_lists. If data and create_using are both multigraphs
then create a multigraph from a multigraph.

7.2 Dictionaries

to_dict_of_dicts(G[, nodelist, edge_data]) Returns adjacency representation of graph as a dictio-


nary of dictionaries.
from_dict_of_dicts(d[, create_using, . . . ]) Returns a graph from a dictionary of dictionaries.

7.2.1 networkx.convert.to_dict_of_dicts

to_dict_of_dicts(G, nodelist=None, edge_data=None)


Returns adjacency representation of graph as a dictionary of dictionaries.
Parameters
• G (graph) – A NetworkX graph
• nodelist (list) – Use only nodes specified in nodelist
• edge_data (list, optional) – If provided, the value of the dictionary will be set to edge_data
for all edges. This is useful to make an adjacency matrix type representation with 1 as
the edge data. If edgedata is None, the edgedata in G is used to fill the values. If G is a
multigraph, the edgedata is a dict for each pair (u,v).

7.2.2 networkx.convert.from_dict_of_dicts

from_dict_of_dicts(d, create_using=None, multigraph_input=False)


Returns a graph from a dictionary of dictionaries.
Parameters
• d (dictionary of dictionaries) – A dictionary of dictionaries adjacency representation.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
• multigraph_input (bool (default False)) – When True, the values of the inner dict are as-
sumed to be containers of edge data for multiple edges. Otherwise this routine assumes the
edge data are singletons.

Examples

>>> dod = {0: {1: {'weight': 1}}} # single edge (0,1)


>>> G = nx.from_dict_of_dicts(dod)

or

>>> G = nx.Graph(dod) # use Graph constructor

596 Chapter 7. Converting to and from other data formats


NetworkX Reference, Release 2.3rc1.dev20190222214247

7.3 Lists

to_dict_of_lists(G[, nodelist]) Returns adjacency representation of graph as a dictio-


nary of lists.
from_dict_of_lists(d[, create_using]) Returns a graph from a dictionary of lists.
to_edgelist(G[, nodelist]) Returns a list of edges in the graph.
from_edgelist(edgelist[, create_using]) Returns a graph from a list of edges.

7.3.1 networkx.convert.to_dict_of_lists

to_dict_of_lists(G, nodelist=None)
Returns adjacency representation of graph as a dictionary of lists.
Parameters
• G (graph) – A NetworkX graph
• nodelist (list) – Use only nodes specified in nodelist

Notes

Completely ignores edge data for MultiGraph and MultiDiGraph.

7.3.2 networkx.convert.from_dict_of_lists

from_dict_of_lists(d, create_using=None)
Returns a graph from a dictionary of lists.
Parameters
• d (dictionary of lists) – A dictionary of lists adjacency representation.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.

Examples

>>> dol = {0: [1]} # single edge (0,1)


>>> G = nx.from_dict_of_lists(dol)

or

>>> G = nx.Graph(dol) # use Graph constructor

7.3.3 networkx.convert.to_edgelist

to_edgelist(G, nodelist=None)
Returns a list of edges in the graph.
Parameters
• G (graph) – A NetworkX graph

7.3. Lists 597


NetworkX Reference, Release 2.3rc1.dev20190222214247

• nodelist (list) – Use only nodes specified in nodelist

7.3.4 networkx.convert.from_edgelist

from_edgelist(edgelist, create_using=None)
Returns a graph from a list of edges.
Parameters
• edgelist (list or iterator) – Edge tuples
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.

Examples

>>> edgelist = [(0, 1)] # single edge (0,1)


>>> G = nx.from_edgelist(edgelist)

or

>>> G = nx.Graph(edgelist) # use Graph constructor

7.4 Numpy

Functions to convert NetworkX graphs to and from numpy/scipy matrices.


The preferred way of converting data to a NetworkX graph is through the graph constructor. The constructor calls the
to_networkx_graph() function which attempts to guess the input type and convert it automatically.

Examples

Create a 10 node random graph from a numpy matrix

>>> import numpy as np


>>> a = np.random.randint(0, 2, size=(10, 10))
>>> D = nx.DiGraph(a)

or equivalently

>>> D = nx.to_networkx_graph(a, create_using=nx.DiGraph)

See also:
nx_agraph, nx_pydot

to_numpy_matrix(G[, nodelist, dtype, order, . . . ]) Returns the graph adjacency matrix as a NumPy matrix.
to_numpy_array(G[, nodelist, dtype, order, . . . ]) Returns the graph adjacency matrix as a NumPy array.
to_numpy_recarray(G[, nodelist, dtype, order]) Returns the graph adjacency matrix as a NumPy recar-
ray.
from_numpy_matrix(A[, parallel_edges, . . . ]) Returns a graph from numpy matrix.
from_numpy_array(A[, parallel_edges, . . . ]) Returns a graph from NumPy array.

598 Chapter 7. Converting to and from other data formats


NetworkX Reference, Release 2.3rc1.dev20190222214247

7.4.1 networkx.convert_matrix.to_numpy_matrix

to_numpy_matrix(G, nodelist=None, dtype=None, order=None, multigraph_weight=<built-in function


sum>, weight=’weight’, nonedge=0.0)
Returns the graph adjacency matrix as a NumPy matrix.
Parameters
• G (graph) – The NetworkX graph used to construct the NumPy matrix.
• nodelist (list, optional) – The rows and columns are ordered according to the nodes in
nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• dtype (NumPy data type, optional) – A valid single NumPy data type used to initialize the
array. This must be a simple type such as int or numpy.float64 and not a compound data
type (see to_numpy_recarray) If None, then the NumPy default is used.
• order ({‘C’, ‘F’}, optional) – Whether to store multidimensional data in C- or Fortran-
contiguous (row- or column-wise) order in memory. If None, then the NumPy default is
used.
• multigraph_weight ({sum, min, max}, optional) – An operator that determines how weights
in multigraphs are handled. The default is to sum the weights of the multiple edges.
• weight (string or None optional (default = ‘weight’)) – The edge attribute that holds the
numerical value used for the edge weight. If an edge does not have that attribute, then the
value 1 is used instead.
• nonedge (float (default = 0.0)) – The matrix values corresponding to nonedges are typically
set to zero. However, this could be undesirable if there are matrix values corresponding to
actual edges that also have the value zero. If so, one might prefer nonedges to have some
other value, such as nan.
Returns M – Graph adjacency matrix
Return type NumPy matrix
See also:
to_numpy_recarray(), from_numpy_matrix()

Notes

The matrix entries are assigned to the weight edge attribute. When an edge does not have a weight attribute, the
value of the entry is set to the number 1. For multiple (parallel) edges, the values of the entries are determined
by the multigraph_weight parameter. The default is to sum the weight attributes for each of the parallel
edges.
When nodelist does not contain every node in G, the matrix is built from the subgraph of G that is induced
by the nodes in nodelist.
The convention used for self-loop edges in graphs is to assign the diagonal matrix entry value to the weight
attribute of the edge (or the number 1 if the edge has no weight attribute). If the alternate convention of doubling
the edge weight is desired the resulting Numpy matrix can be modified as follows:

>>> import numpy as np


>>> try:
... np.set_printoptions(legacy="1.13")
... except TypeError:
... pass
(continues on next page)

7.4. Numpy 599


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> G = nx.Graph([(1, 1)])
>>> A = nx.to_numpy_matrix(G)
>>> A
matrix([[ 1.]])
>>> A.A[np.diag_indices_from(A)] *= 2
>>> A
matrix([[ 2.]])

Examples

>>> G = nx.MultiDiGraph()
>>> G.add_edge(0, 1, weight=2)
0
>>> G.add_edge(1, 0)
0
>>> G.add_edge(2, 2, weight=3)
0
>>> G.add_edge(2, 2)
1
>>> nx.to_numpy_matrix(G, nodelist=[0, 1, 2])
matrix([[ 0., 2., 0.],
[ 1., 0., 0.],
[ 0., 0., 4.]])

7.4.2 networkx.convert_matrix.to_numpy_array

to_numpy_array(G, nodelist=None, dtype=None, order=None, multigraph_weight=<built-in function


sum>, weight=’weight’, nonedge=0.0)
Returns the graph adjacency matrix as a NumPy array.
Parameters
• G (graph) – The NetworkX graph used to construct the NumPy array.
• nodelist (list, optional) – The rows and columns are ordered according to the nodes in
nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• dtype (NumPy data type, optional) – A valid single NumPy data type used to initialize the
array. This must be a simple type such as int or numpy.float64 and not a compound data
type (see to_numpy_recarray) If None, then the NumPy default is used.
• order ({‘C’, ‘F’}, optional) – Whether to store multidimensional data in C- or Fortran-
contiguous (row- or column-wise) order in memory. If None, then the NumPy default is
used.
• multigraph_weight ({sum, min, max}, optional) – An operator that determines how weights
in multigraphs are handled. The default is to sum the weights of the multiple edges.
• weight (string or None optional (default = ‘weight’)) – The edge attribute that holds the
numerical value used for the edge weight. If an edge does not have that attribute, then the
value 1 is used instead.
• nonedge (float (default = 0.0)) – The array values corresponding to nonedges are typically
set to zero. However, this could be undesirable if there are array values corresponding to
actual edges that also have the value zero. If so, one might prefer nonedges to have some
other value, such as nan.

600 Chapter 7. Converting to and from other data formats


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns A – Graph adjacency matrix


Return type NumPy ndarray
See also:
from_numpy_array()

Notes

Entries in the adjacency matrix are assigned to the weight edge attribute. When an edge does not have a weight
attribute, the value of the entry is set to the number 1. For multiple (parallel) edges, the values of the entries are
determined by the multigraph_weight parameter. The default is to sum the weight attributes for each of
the parallel edges.
When nodelist does not contain every node in G, the adjacency matrix is built from the subgraph of G that is
induced by the nodes in nodelist.
The convention used for self-loop edges in graphs is to assign the diagonal array entry value to the weight
attribute of the edge (or the number 1 if the edge has no weight attribute). If the alternate convention of doubling
the edge weight is desired the resulting NumPy array can be modified as follows:

>>> import numpy as np


>>> try:
... np.set_printoptions(legacy="1.13")
... except TypeError:
... pass
>>> G = nx.Graph([(1, 1)])
>>> A = nx.to_numpy_array(G)
>>> A
array([[ 1.]])
>>> A[np.diag_indices_from(A)] *= 2
>>> A
array([[ 2.]])

Examples

>>> G = nx.MultiDiGraph()
>>> G.add_edge(0, 1, weight=2)
0
>>> G.add_edge(1, 0)
0
>>> G.add_edge(2, 2, weight=3)
0
>>> G.add_edge(2, 2)
1
>>> nx.to_numpy_array(G, nodelist=[0, 1, 2])
array([[ 0., 2., 0.],
[ 1., 0., 0.],
[ 0., 0., 4.]])

7.4.3 networkx.convert_matrix.to_numpy_recarray

to_numpy_recarray(G, nodelist=None, dtype=None, order=None)


Returns the graph adjacency matrix as a NumPy recarray.

7.4. Numpy 601


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (graph) – The NetworkX graph used to construct the NumPy matrix.
• nodelist (list, optional) – The rows and columns are ordered according to the nodes in
nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• dtype (NumPy data-type, optional) – A valid NumPy named dtype used to initialize the
NumPy recarray. The data type names are assumed to be keys in the graph edge attribute
dictionary.
• order ({‘C’, ‘F’}, optional) – Whether to store multidimensional data in C- or Fortran-
contiguous (row- or column-wise) order in memory. If None, then the NumPy default is
used.
Returns M – The graph with specified edge data as a Numpy recarray
Return type NumPy recarray

Notes

When nodelist does not contain every node in G, the matrix is built from the subgraph of G that is induced
by the nodes in nodelist.

Examples

>>> G = nx.Graph()
>>> G.add_edge(1, 2, weight=7.0, cost=5)
>>> A = nx.to_numpy_recarray(G, dtype=[('weight', float), ('cost', int)])
>>> print(A.weight)
[[ 0. 7.]
[ 7. 0.]]
>>> print(A.cost)
[[0 5]
[5 0]]

7.4.4 networkx.convert_matrix.from_numpy_matrix

from_numpy_matrix(A, parallel_edges=False, create_using=None)


Returns a graph from numpy matrix.
The numpy matrix is interpreted as an adjacency matrix for the graph.
Parameters
• A (numpy matrix) – An adjacency matrix representation of a graph
• parallel_edges (Boolean) – If True, create_using is a multigraph, and A is an integer
matrix, then entry (i, j) in the matrix is interpreted as the number of parallel edges joining
vertices i and j in the graph. If False, then the entries in the adjacency matrix are interpreted
as the weight of a single edge joining the vertices.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.

602 Chapter 7. Converting to and from other data formats


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

If create_using is networkx.MultiGraph or networkx.MultiDiGraph, parallel_edges


is True, and the entries of A are of type int, then this function returns a multigraph (constructed from
create_using) with parallel edges.
If create_using indicates an undirected multigraph, then only the edges indicated by the upper triangle of
the matrix A will be added to the graph.
If the numpy matrix has a single data type for each matrix entry it will be converted to an appropriate Python
data type.
If the numpy matrix has a user-specified compound data type the names of the data fields will be used as attribute
keys in the resulting NetworkX graph.
See also:
to_numpy_matrix(), to_numpy_recarray()

Examples

Simple integer weights on edges:

>>> import numpy as np


>>> A = np.matrix([[1, 1], [2, 1]])
>>> G = nx.from_numpy_matrix(A)

If create_using indicates a multigraph and the matrix has only integer entries and parallel_edges is
False, then the entries will be treated as weights for edges joining the nodes (without creating parallel edges):

>>> A = np.matrix([[1, 1], [1, 2]])


>>> G = nx.from_numpy_matrix(A, create_using=nx.MultiGraph)
>>> G[1][1]
AtlasView({0: {'weight': 2}})

If create_using indicates a multigraph and the matrix has only integer entries and parallel_edges is
True, then the entries will be treated as the number of parallel edges joining those two vertices:

>>> A = np.matrix([[1, 1], [1, 2]])


>>> temp = nx.MultiGraph()
>>> G = nx.from_numpy_matrix(A, parallel_edges=True, create_using=temp)
>>> G[1][1]
AtlasView({0: {'weight': 1}, 1: {'weight': 1}})

User defined compound data type on edges:

>>> dt = [('weight', float), ('cost', int)]


>>> A = np.matrix([[(1.0, 2)]], dtype=dt)
>>> G = nx.from_numpy_matrix(A)
>>> list(G.edges())
[(0, 0)]
>>> G[0][0]['cost']
2
>>> G[0][0]['weight']
1.0

7.4. Numpy 603


NetworkX Reference, Release 2.3rc1.dev20190222214247

7.4.5 networkx.convert_matrix.from_numpy_array

from_numpy_array(A, parallel_edges=False, create_using=None)


Returns a graph from NumPy array.
The NumPy array is interpreted as an adjacency matrix for the graph.
Parameters
• A (NumPy ndarray) – An adjacency matrix representation of a graph
• parallel_edges (Boolean) – If this is True, create_using is a multigraph, and A is an
integer array, then entry (i, j) in the array is interpreted as the number of parallel edges
joining vertices i and j in the graph. If it is False, then the entries in the array are interpreted
as the weight of a single edge joining the vertices.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.

Notes

If create_using is networkx.MultiGraph or networkx.MultiDiGraph, parallel_edges


is True, and the entries of A are of type int, then this function returns a multigraph (of the same type as
create_using) with parallel edges.
If create_using indicates an undirected multigraph, then only the edges indicated by the upper triangle of
the array A will be added to the graph.
If the NumPy array has a single data type for each array entry it will be converted to an appropriate Python data
type.
If the NumPy array has a user-specified compound data type the names of the data fields will be used as attribute
keys in the resulting NetworkX graph.
See also:
to_numpy_array()

Examples

Simple integer weights on edges:

>>> import numpy as np


>>> A = np.array([[1, 1], [2, 1]])
>>> G = nx.from_numpy_array(A)
>>> G.edges(data=True)
EdgeDataView([(0, 0, {'weight': 1}), (0, 1, {'weight': 2}), (1, 1, {'weight': 1}
˓→)])

If create_using indicates a multigraph and the array has only integer entries and parallel_edges is
False, then the entries will be treated as weights for edges joining the nodes (without creating parallel edges):

>>> A = np.array([[1, 1], [1, 2]])


>>> G = nx.from_numpy_array(A, create_using=nx.MultiGraph)
>>> G[1][1]
AtlasView({0: {'weight': 2}})

If create_using indicates a multigraph and the array has only integer entries and parallel_edges is
True, then the entries will be treated as the number of parallel edges joining those two vertices:

604 Chapter 7. Converting to and from other data formats


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> A = np.array([[1, 1], [1, 2]])


>>> temp = nx.MultiGraph()
>>> G = nx.from_numpy_array(A, parallel_edges=True, create_using=temp)
>>> G[1][1]
AtlasView({0: {'weight': 1}, 1: {'weight': 1}})

User defined compound data type on edges:

>>> dt = [('weight', float), ('cost', int)]


>>> A = np.array([[(1.0, 2)]], dtype=dt)
>>> G = nx.from_numpy_array(A)
>>> G.edges()
EdgeView([(0, 0)])
>>> G[0][0]['cost']
2
>>> G[0][0]['weight']
1.0

7.5 Scipy

to_scipy_sparse_matrix(G[, nodelist, dtype, Returns the graph adjacency matrix as a SciPy sparse
. . . ]) matrix.
from_scipy_sparse_matrix(A[, . . . ]) Creates a new graph from an adjacency matrix given as
a SciPy sparse matrix.

7.5.1 networkx.convert_matrix.to_scipy_sparse_matrix

to_scipy_sparse_matrix(G, nodelist=None, dtype=None, weight=’weight’, format=’csr’)


Returns the graph adjacency matrix as a SciPy sparse matrix.
Parameters
• G (graph) – The NetworkX graph used to construct the NumPy matrix.
• nodelist (list, optional) – The rows and columns are ordered according to the nodes in
nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• dtype (NumPy data-type, optional) – A valid NumPy dtype used to initialize the array. If
None, then the NumPy default is used.
• weight (string or None optional (default=’weight’)) – The edge attribute that holds the
numerical value used for the edge weight. If None then all edge weights are 1.
• format (str in {‘bsr’, ‘csr’, ‘csc’, ‘coo’, ‘lil’, ‘dia’, ‘dok’}) – The type of the matrix to be
returned (default ‘csr’). For some algorithms different implementations of sparse matrices
can perform better. See1 for details.
Returns M – Graph adjacency matrix.
Return type SciPy sparse matrix
1 Scipy Dev. References, “Sparse Matrices”, https://docs.scipy.org/doc/scipy/reference/sparse.html

7.5. Scipy 605


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

The matrix entries are populated using the edge attribute held in parameter weight. When an edge does not have
that attribute, the value of the entry is 1.
For multiple edges the matrix values are the sums of the edge weights.
When nodelist does not contain every node in G, the matrix is built from the subgraph of G that is induced
by the nodes in nodelist.
Uses coo_matrix format. To convert to other formats specify the format= keyword.
The convention used for self-loop edges in graphs is to assign the diagonal matrix entry value to the weight
attribute of the edge (or the number 1 if the edge has no weight attribute). If the alternate convention of doubling
the edge weight is desired the resulting Scipy sparse matrix can be modified as follows:
>>> import scipy as sp
>>> G = nx.Graph([(1, 1)])
>>> A = nx.to_scipy_sparse_matrix(G)
>>> print(A.todense())
[[1]]
>>> A.setdiag(A.diagonal() * 2)
>>> print(A.todense())
[[2]]

Examples

>>> G = nx.MultiDiGraph()
>>> G.add_edge(0, 1, weight=2)
0
>>> G.add_edge(1, 0)
0
>>> G.add_edge(2, 2, weight=3)
0
>>> G.add_edge(2, 2)
1
>>> S = nx.to_scipy_sparse_matrix(G, nodelist=[0, 1, 2])
>>> print(S.todense())
[[0 2 0]
[1 0 0]
[0 0 4]]

References

7.5.2 networkx.convert_matrix.from_scipy_sparse_matrix

from_scipy_sparse_matrix(A, parallel_edges=False, create_using=None, edge_attribute=’weight’)


Creates a new graph from an adjacency matrix given as a SciPy sparse matrix.
Parameters
• A (scipy sparse matrix) – An adjacency matrix representation of a graph
• parallel_edges (Boolean) – If this is True, create_using is a multigraph, and A is an
integer matrix, then entry (i, j) in the matrix is interpreted as the number of parallel edges
joining vertices i and j in the graph. If it is False, then the entries in the matrix are interpreted
as the weight of a single edge joining the vertices.

606 Chapter 7. Converting to and from other data formats


NetworkX Reference, Release 2.3rc1.dev20190222214247

• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to


create. If graph instance, then cleared before populated.
• edge_attribute (string) – Name of edge attribute to store matrix numeric value. The data
will have the same type as the matrix entry (int, float, (real,imag)).

Notes

If create_using is networkx.MultiGraph or networkx.MultiDiGraph, parallel_edges


is True, and the entries of A are of type int, then this function returns a multigraph (constructed from
create_using) with parallel edges. In this case, edge_attribute will be ignored.
If create_using indicates an undirected multigraph, then only the edges indicated by the upper triangle of
the matrix A will be added to the graph.

Examples

>>> import scipy as sp


>>> A = sp.sparse.eye(2, 2, 1)
>>> G = nx.from_scipy_sparse_matrix(A)

If create_using indicates a multigraph and the matrix has only integer entries and parallel_edges is
False, then the entries will be treated as weights for edges joining the nodes (without creating parallel edges):

>>> A = sp.sparse.csr_matrix([[1, 1], [1, 2]])


>>> G = nx.from_scipy_sparse_matrix(A, create_using=nx.MultiGraph)
>>> G[1][1]
AtlasView({0: {'weight': 2}})

If create_using indicates a multigraph and the matrix has only integer entries and parallel_edges is
True, then the entries will be treated as the number of parallel edges joining those two vertices:

>>> A = sp.sparse.csr_matrix([[1, 1], [1, 2]])


>>> G = nx.from_scipy_sparse_matrix(A, parallel_edges=True,
... create_using=nx.MultiGraph)
>>> G[1][1]
AtlasView({0: {'weight': 1}, 1: {'weight': 1}})

7.6 Pandas

to_pandas_adjacency(G[, nodelist, dtype, . . . ]) Returns the graph adjacency matrix as a Pandas


DataFrame.
from_pandas_adjacency(df[, create_using]) Returns a graph from Pandas DataFrame.
to_pandas_edgelist(G[, source, target, . . . ]) Returns the graph edge list as a Pandas DataFrame.
from_pandas_edgelist(df[, source, target, . . . ]) Returns a graph from Pandas DataFrame containing an
edge list.

7.6. Pandas 607


NetworkX Reference, Release 2.3rc1.dev20190222214247

7.6.1 networkx.convert_matrix.to_pandas_adjacency

to_pandas_adjacency(G, nodelist=None, dtype=None, order=None, multigraph_weight=<built-in


function sum>, weight=’weight’, nonedge=0.0)
Returns the graph adjacency matrix as a Pandas DataFrame.
Parameters
• G (graph) – The NetworkX graph used to construct the Pandas DataFrame.
• nodelist (list, optional) – The rows and columns are ordered according to the nodes in
nodelist. If nodelist is None, then the ordering is produced by G.nodes().
• multigraph_weight ({sum, min, max}, optional) – An operator that determines how weights
in multigraphs are handled. The default is to sum the weights of the multiple edges.
• weight (string or None, optional) – The edge attribute that holds the numerical value used
for the edge weight. If an edge does not have that attribute, then the value 1 is used instead.
• nonedge (float, optional) – The matrix values corresponding to nonedges are typically set to
zero. However, this could be undesirable if there are matrix values corresponding to actual
edges that also have the value zero. If so, one might prefer nonedges to have some other
value, such as nan.
Returns df – Graph adjacency matrix
Return type Pandas DataFrame

Notes

The DataFrame entries are assigned to the weight edge attribute. When an edge does not have a weight attribute,
the value of the entry is set to the number 1. For multiple (parallel) edges, the values of the entries are determined
by the ‘multigraph_weight’ parameter. The default is to sum the weight attributes for each of the parallel edges.
When nodelist does not contain every node in G, the matrix is built from the subgraph of G that is induced
by the nodes in nodelist.
The convention used for self-loop edges in graphs is to assign the diagonal matrix entry value to the weight
attribute of the edge (or the number 1 if the edge has no weight attribute). If the alternate convention of doubling
the edge weight is desired the resulting Pandas DataFrame can be modified as follows:

>>> import pandas as pd


>>> pd.options.display.max_columns = 20
>>> import numpy as np
>>> G = nx.Graph([(1, 1)])
>>> df = nx.to_pandas_adjacency(G, dtype=int)
>>> df
1
1 1
>>> df.values[np.diag_indices_from(df)] *= 2
>>> df
1
1 2

Examples

608 Chapter 7. Converting to and from other data formats


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.MultiDiGraph()
>>> G.add_edge(0, 1, weight=2)
0
>>> G.add_edge(1, 0)
0
>>> G.add_edge(2, 2, weight=3)
0
>>> G.add_edge(2, 2)
1
>>> nx.to_pandas_adjacency(G, nodelist=[0, 1, 2], dtype=int)
0 1 2
0 0 2 0
1 1 0 0
2 0 0 4

7.6.2 networkx.convert_matrix.from_pandas_adjacency

from_pandas_adjacency(df, create_using=None)
Returns a graph from Pandas DataFrame.
The Pandas DataFrame is interpreted as an adjacency matrix for the graph.
Parameters
• df (Pandas DataFrame) – An adjacency matrix representation of a graph
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.

Notes

If the numpy matrix has a single data type for each matrix entry it will be converted to an appropriate Python
data type.
If the numpy matrix has a user-specified compound data type the names of the data fields will be used as attribute
keys in the resulting NetworkX graph.
See also:
to_pandas_adjacency()

Examples

Simple integer weights on edges:


>>> import pandas as pd
>>> pd.options.display.max_columns = 20
>>> df = pd.DataFrame([[1, 1], [2, 1]])
>>> df
0 1
0 1 1
1 2 1
>>> G = nx.from_pandas_adjacency(df)
>>> G.name = 'Graph from pandas adjacency matrix'
>>> print(nx.info(G))
Name: Graph from pandas adjacency matrix
(continues on next page)

7.6. Pandas 609


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


Type: Graph
Number of nodes: 2
Number of edges: 3
Average degree: 3.0000

7.6.3 networkx.convert_matrix.to_pandas_edgelist

to_pandas_edgelist(G, source=’source’, target=’target’, nodelist=None, dtype=None, order=None)


Returns the graph edge list as a Pandas DataFrame.
Parameters
• G (graph) – The NetworkX graph used to construct the Pandas DataFrame.
• source (str or int, optional) – A valid column name (string or integer) for the source nodes
(for the directed case).
• target (str or int, optional) – A valid column name (string or integer) for the target nodes
(for the directed case).
• nodelist (list, optional) – Use only nodes specified in nodelist
Returns df – Graph edge list
Return type Pandas DataFrame

Examples

>>> G = nx.Graph([('A', 'B', {'cost': 1, 'weight': 7}),


... ('C', 'E', {'cost': 9, 'weight': 10})])
>>> df = nx.to_pandas_edgelist(G, nodelist=['A', 'C'])
>>> df[['source', 'target', 'cost', 'weight']]
source target cost weight
0 A B 1 7
1 C E 9 10

7.6.4 networkx.convert_matrix.from_pandas_edgelist

from_pandas_edgelist(df, source=’source’, target=’target’, edge_attr=None, create_using=None)


Returns a graph from Pandas DataFrame containing an edge list.
The Pandas DataFrame should contain at least two columns of node names and zero or more columns of edge
attributes. Each row will be processed as one edge instance.
Note: This function iterates over DataFrame.values, which is not guaranteed to retain the data type across
columns in the row. This is only a problem if your row is entirely numeric and a mix of ints and floats. In that
case, all values will be returned as floats. See the DataFrame.iterrows documentation for an example.
Parameters
• df (Pandas DataFrame) – An edge list representation of a graph
• source (str or int) – A valid column name (string or integer) for the source nodes (for the
directed case).

610 Chapter 7. Converting to and from other data formats


NetworkX Reference, Release 2.3rc1.dev20190222214247

• target (str or int) – A valid column name (string or integer) for the target nodes (for the
directed case).
• edge_attr (str or int, iterable, True) – A valid column name (str or integer) or list of column
names that will be used to retrieve items from the row and add them to the graph as edge
attributes. If True, all of the remaining columns will be added.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
See also:
to_pandas_edgelist()

Examples

Simple integer weights on edges:

>>> import pandas as pd


>>> pd.options.display.max_columns = 20
>>> import numpy as np
>>> rng = np.random.RandomState(seed=5)
>>> ints = rng.randint(1, 11, size=(3,2))
>>> a = ['A', 'B', 'C']
>>> b = ['D', 'A', 'E']
>>> df = pd.DataFrame(ints, columns=['weight', 'cost'])
>>> df[0] = a
>>> df['b'] = b
>>> df[['weight', 'cost', 0, 'b']]
weight cost 0 b
0 4 7 A D
1 7 1 B A
2 10 9 C E
>>> G = nx.from_pandas_edgelist(df, 0, 'b', ['weight', 'cost'])
>>> G['E']['C']['weight']
10
>>> G['E']['C']['cost']
9
>>> edges = pd.DataFrame({'source': [0, 1, 2],
... 'target': [2, 2, 3],
... 'weight': [3, 4, 5],
... 'color': ['red', 'blue', 'blue']})
>>> G = nx.from_pandas_edgelist(edges, edge_attr=True)
>>> G[0][2]['color']
'red'

7.6. Pandas 611


NetworkX Reference, Release 2.3rc1.dev20190222214247

612 Chapter 7. Converting to and from other data formats


CHAPTER

EIGHT

RELABELING NODES

8.1 Relabeling

convert_node_labels_to_integers(G[, Returns a copy of the graph G with the nodes relabeled


. . . ]) using consecutive integers.
relabel_nodes(G, mapping[, copy]) Relabel the nodes of the graph G.

8.1.1 networkx.relabel.convert_node_labels_to_integers

convert_node_labels_to_integers(G, first_label=0, ordering=’default’, label_attribute=None)


Returns a copy of the graph G with the nodes relabeled using consecutive integers.
Parameters
• G (graph) – A NetworkX graph
• first_label (int, optional (default=0)) – An integer specifying the starting offset in number-
ing nodes. The new integer labels are numbered first_label, . . . , n-1+first_label.
• ordering (string) – “default” : inherit node ordering from G.nodes() “sorted” : inherit node
ordering from sorted(G.nodes()) “increasing degree” : nodes are sorted by increasing degree
“decreasing degree” : nodes are sorted by decreasing degree
• label_attribute (string, optional (default=None)) – Name of node attribute to store old
label. If None no attribute is created.

Notes

Node and edge attribute data are copied to the new (relabeled) graph.
There is no guarantee that the relabeling of nodes to integers will give the same two integers for two (even
identical graphs). Use the ordering argument to try to preserve the order.
See also:
relabel_nodes()

8.1.2 networkx.relabel.relabel_nodes

relabel_nodes(G, mapping, copy=True)


Relabel the nodes of the graph G.
Parameters

613
NetworkX Reference, Release 2.3rc1.dev20190222214247

• G (graph) – A NetworkX graph


• mapping (dictionary) – A dictionary with the old labels as keys and new labels as values.
A partial mapping is allowed.
• copy (bool (optional, default=True)) – If True return a copy, or if False relabel the nodes in
place.

Examples

To create a new graph with nodes relabeled according to a given dictionary:

>>> G = nx.path_graph(3)
>>> sorted(G)
[0, 1, 2]
>>> mapping = {0: 'a', 1: 'b', 2: 'c'}
>>> H = nx.relabel_nodes(G, mapping)
>>> sorted(H)
['a', 'b', 'c']

Nodes can be relabeled with any hashable object, including numbers and strings:

>>> import string


>>> G = nx.path_graph(26) # nodes are integers 0 through 25
>>> sorted(G)[:3]
[0, 1, 2]
>>> mapping = dict(zip(G, string.ascii_lowercase))
>>> G = nx.relabel_nodes(G, mapping) # nodes are characters a through z
>>> sorted(G)[:3]
['a', 'b', 'c']
>>> mapping = dict(zip(G, range(1, 27)))
>>> G = nx.relabel_nodes(G, mapping) # nodes are integers 1 through 26
>>> sorted(G)[:3]
[1, 2, 3]

To perform a partial in-place relabeling, provide a dictionary mapping only a subset of the nodes, and set the
copy keyword argument to False:

>>> G = nx.path_graph(3) # nodes 0-1-2


>>> mapping = {0: 'a', 1: 'b'} # 0->'a' and 1->'b'
>>> G = nx.relabel_nodes(G, mapping, copy=False)
>>> sorted(G, key=str)
[2, 'a', 'b']

A mapping can also be given as a function:

>>> G = nx.path_graph(3)
>>> H = nx.relabel_nodes(G, lambda x: x ** 2)
>>> list(H)
[0, 1, 4]

Notes

Only the nodes specified in the mapping will be relabeled.


The keyword setting copy=False modifies the graph in place. Relabel_nodes avoids naming collisions by build-
ing a directed graph from mapping which specifies the order of relabelings. Naming collisions, such as a->b,

614 Chapter 8. Relabeling nodes


NetworkX Reference, Release 2.3rc1.dev20190222214247

b->c, are ordered such that “b” gets renamed to “c” before “a” gets renamed “b”. In cases of circular mappings
(e.g. a->b, b->a), modifying the graph is not possible in-place and an exception is raised. In that case, use
copy=True.
See also:
convert_node_labels_to_integers()

8.1. Relabeling 615


NetworkX Reference, Release 2.3rc1.dev20190222214247

616 Chapter 8. Relabeling nodes


CHAPTER

NINE

READING AND WRITING GRAPHS

9.1 Adjacency List

9.1.1 Adjacency List

Read and write NetworkX graphs as adjacency lists.


Adjacency list format is useful for graphs without data associated with nodes or edges and for nodes that can be
meaningfully represented as strings.

Format

The adjacency list format consists of lines with node labels. The first label in a line is the source node. Further labels
in the line are considered target nodes and are added to the graph along with an edge between the source node and
target node.
The graph with edges a-b, a-c, d-e can be represented as the following adjacency list (anything following the # in a
line is a comment):

a b c # source target target


d e

read_adjlist(path[, comments, delimiter, . . . ]) Read graph in adjacency list format from path.
write_adjlist(G, path[, comments, . . . ]) Write graph G in single-line adjacency-list format to
path.
parse_adjlist(lines[, comments, delimiter, . . . ]) Parse lines of a graph adjacency list representation.
generate_adjlist(G[, delimiter]) Generate a single line of the graph G in adjacency list
format.

9.1.2 networkx.readwrite.adjlist.read_adjlist

read_adjlist(path, comments=’#’, delimiter=None, create_using=None, nodetype=None, encoding=’utf-


8’)
Read graph in adjacency list format from path.
Parameters
• path (string or file) – Filename or file handle to read. Filenames ending in .gz or .bz2 will
be uncompressed.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to

617
NetworkX Reference, Release 2.3rc1.dev20190222214247

create. If graph instance, then cleared before populated.


• nodetype (Python type, optional) – Convert nodes to this type.
• comments (string, optional) – Marker for comment lines
• delimiter (string, optional) – Separator for node labels. The default is whitespace.
Returns G – The graph corresponding to the lines in adjacency list format.
Return type NetworkX graph

Examples

>>> G=nx.path_graph(4)
>>> nx.write_adjlist(G, "test.adjlist")
>>> G=nx.read_adjlist("test.adjlist")

The path can be a filehandle or a string with the name of the file. If a filehandle is provided, it has to be opened
in ‘rb’ mode.

>>> fh=open("test.adjlist", 'rb')


>>> G=nx.read_adjlist(fh)

Filenames ending in .gz or .bz2 will be compressed.

>>> nx.write_adjlist(G,"test.adjlist.gz")
>>> G=nx.read_adjlist("test.adjlist.gz")

The optional nodetype is a function to convert node strings to nodetype.


For example

>>> G=nx.read_adjlist("test.adjlist", nodetype=int)

will attempt to convert all nodes to integer type.


Since nodes must be hashable, the function nodetype must return hashable types (e.g. int, float, str, frozenset -
or tuples of those, etc.)
The optional create_using parameter indicates the type of NetworkX graph created. The default is nx.Graph,
an undirected graph. To read the data as a directed graph use

>>> G=nx.read_adjlist("test.adjlist", create_using=nx.DiGraph)

Notes

This format does not store graph or node data.


See also:
write_adjlist()

9.1.3 networkx.readwrite.adjlist.write_adjlist

write_adjlist(G, path, comments=’#’, delimiter=’ ’, encoding=’utf-8’)


Write graph G in single-line adjacency-list format to path.

618 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (NetworkX graph)
• path (string or file) – Filename or file handle for data output. Filenames ending in .gz or
.bz2 will be compressed.
• comments (string, optional) – Marker for comment lines
• delimiter (string, optional) – Separator for node labels
• encoding (string, optional) – Text encoding.

Examples

>>> G=nx.path_graph(4)
>>> nx.write_adjlist(G,"test.adjlist")

The path can be a filehandle or a string with the name of the file. If a filehandle is provided, it has to be opened
in ‘wb’ mode.

>>> fh=open("test.adjlist",'wb')
>>> nx.write_adjlist(G, fh)

Notes

This format does not store graph, node, or edge data.


See also:
read_adjlist(), generate_adjlist()

9.1.4 networkx.readwrite.adjlist.parse_adjlist

parse_adjlist(lines, comments=’#’, delimiter=None, create_using=None, nodetype=None)


Parse lines of a graph adjacency list representation.
Parameters
• lines (list or iterator of strings) – Input data in adjlist format
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
• nodetype (Python type, optional) – Convert nodes to this type.
• comments (string, optional) – Marker for comment lines
• delimiter (string, optional) – Separator for node labels. The default is whitespace.
Returns G – The graph corresponding to the lines in adjacency list format.
Return type NetworkX graph

Examples

9.1. Adjacency List 619


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> lines = ['1 2 5',


... '2 3 4',
... '3 5',
... '4',
... '5']
>>> G = nx.parse_adjlist(lines, nodetype=int)
>>> nodes = [1, 2, 3, 4, 5]
>>> all(node in G for node in nodes)
True
>>> edges = [(1, 2), (1, 5), (2, 3), (2, 4), (3, 5)]
>>> all((u, v) in G.edges() or (v, u) in G.edges() for (u, v) in edges)
True

See also:
read_adjlist()

9.1.5 networkx.readwrite.adjlist.generate_adjlist

generate_adjlist(G, delimiter=’ ’)
Generate a single line of the graph G in adjacency list format.
Parameters
• G (NetworkX graph)
• delimiter (string, optional) – Separator for node labels
Returns lines – Lines of data in adjlist format.
Return type string

Examples

>>> G = nx.lollipop_graph(4, 3)
>>> for line in nx.generate_adjlist(G):
... print(line)
0 1 2 3
1 2 3
2 3
3 4
4 5
5 6
6

See also:
write_adjlist(), read_adjlist()

9.2 Multiline Adjacency List

9.2.1 Multi-line Adjacency List

Read and write NetworkX graphs as multi-line adjacency lists.

620 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

The multi-line adjacency list format is useful for graphs with nodes that can be meaningfully represented as strings.
With this format simple edge data can be stored but node or graph data is not.

Format

The first label in a line is the source node label followed by the node degree d. The next d lines are target node labels
and optional edge data. That pattern repeats for all nodes in the graph.
The graph with edges a-b, a-c, d-e can be represented as the following adjacency list (anything following the # in a
line is a comment):

# example.multiline-adjlist
a 2
b
c
d 1
e

read_multiline_adjlist(path[, comments, Read graph in multi-line adjacency list format from


. . . ]) path.
write_multiline_adjlist(G, path[, . . . ]) Write the graph G in multiline adjacency list format to
path
parse_multiline_adjlist(lines[, comments, Parse lines of a multiline adjacency list representation
. . . ]) of a graph.
generate_multiline_adjlist(G[, delimiter]) Generate a single line of the graph G in multiline adja-
cency list format.

9.2.2 networkx.readwrite.multiline_adjlist.read_multiline_adjlist

read_multiline_adjlist(path, comments=’#’, delimiter=None, create_using=None, nodetype=None,


edgetype=None, encoding=’utf-8’)
Read graph in multi-line adjacency list format from path.
Parameters
• path (string or file) – Filename or file handle to read. Filenames ending in .gz or .bz2 will
be uncompressed.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
• nodetype (Python type, optional) – Convert nodes to this type.
• edgetype (Python type, optional) – Convert edge data to this type.
• comments (string, optional) – Marker for comment lines
• delimiter (string, optional) – Separator for node labels. The default is whitespace.
Returns G
Return type NetworkX graph

9.2. Multiline Adjacency List 621


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G=nx.path_graph(4)
>>> nx.write_multiline_adjlist(G,"test.adjlist")
>>> G=nx.read_multiline_adjlist("test.adjlist")

The path can be a file or a string with the name of the file. If a file s provided, it has to be opened in ‘rb’ mode.

>>> fh=open("test.adjlist", 'rb')


>>> G=nx.read_multiline_adjlist(fh)

Filenames ending in .gz or .bz2 will be compressed.

>>> nx.write_multiline_adjlist(G,"test.adjlist.gz")
>>> G=nx.read_multiline_adjlist("test.adjlist.gz")

The optional nodetype is a function to convert node strings to nodetype.


For example

>>> G=nx.read_multiline_adjlist("test.adjlist", nodetype=int)

will attempt to convert all nodes to integer type.


The optional edgetype is a function to convert edge data strings to edgetype.

>>> G=nx.read_multiline_adjlist("test.adjlist")

The optional create_using parameter is a NetworkX graph container. The default is Graph(), an undirected
graph. To read the data as a directed graph use

>>> G=nx.read_multiline_adjlist("test.adjlist", create_using=nx.DiGraph)

Notes

This format does not store graph, node, or edge data.


See also:
write_multiline_adjlist()

9.2.3 networkx.readwrite.multiline_adjlist.write_multiline_adjlist

write_multiline_adjlist(G, path, delimiter=’ ’, comments=’#’, encoding=’utf-8’)


Write the graph G in multiline adjacency list format to path
Parameters
• G (NetworkX graph)
• comments (string, optional) – Marker for comment lines
• delimiter (string, optional) – Separator for node labels
• encoding (string, optional) – Text encoding.

622 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G=nx.path_graph(4)
>>> nx.write_multiline_adjlist(G,"test.adjlist")

The path can be a file handle or a string with the name of the file. If a file handle is provided, it has to be opened
in ‘wb’ mode.
>>> fh=open("test.adjlist",'wb')
>>> nx.write_multiline_adjlist(G,fh)

Filenames ending in .gz or .bz2 will be compressed.


>>> nx.write_multiline_adjlist(G,"test.adjlist.gz")

See also:
read_multiline_adjlist()

9.2.4 networkx.readwrite.multiline_adjlist.parse_multiline_adjlist

parse_multiline_adjlist(lines, comments=’#’, delimiter=None, create_using=None, node-


type=None, edgetype=None)
Parse lines of a multiline adjacency list representation of a graph.
Parameters
• lines (list or iterator of strings) – Input data in multiline adjlist format
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
• nodetype (Python type, optional) – Convert nodes to this type.
• comments (string, optional) – Marker for comment lines
• delimiter (string, optional) – Separator for node labels. The default is whitespace.
Returns G – The graph corresponding to the lines in multiline adjacency list format.
Return type NetworkX graph

Examples

>>> lines = ['1 2',


... "2 {'weight':3, 'name': 'Frodo'}",
... "3 {}",
... "2 1",
... "5 {'weight':6, 'name': 'Saruman'}"]
>>> G = nx.parse_multiline_adjlist(iter(lines), nodetype=int)
>>> list(G)
[1, 2, 3, 5]

9.2.5 networkx.readwrite.multiline_adjlist.generate_multiline_adjlist

generate_multiline_adjlist(G, delimiter=’ ’)
Generate a single line of the graph G in multiline adjacency list format.

9.2. Multiline Adjacency List 623


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (NetworkX graph)
• delimiter (string, optional) – Separator for node labels
Returns lines – Lines of data in multiline adjlist format.
Return type string

Examples

>>> G = nx.lollipop_graph(4, 3)
>>> for line in nx.generate_multiline_adjlist(G):
... print(line)
0 3
1 {}
2 {}
3 {}
1 2
2 {}
3 {}
2 1
3 {}
3 1
4 {}
4 1
5 {}
5 1
6 {}
6 0

See also:
write_multiline_adjlist(), read_multiline_adjlist()

9.3 Edge List

9.3.1 Edge Lists

Read and write NetworkX graphs as edge lists.


The multi-line adjacency list format is useful for graphs with nodes that can be meaningfully represented as strings.
With the edgelist format simple edge data can be stored but node or graph data is not. There is no way of representing
isolated nodes unless the node has a self-loop edge.

Format

You can read or write three formats of edge lists with these functions.
Node pairs with no data:

1 2

Python dictionary as data:

624 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

1 2 {'weight':7, 'color':'green'}

Arbitrary data:

1 2 7 green

read_edgelist(path[, comments, delimiter, . . . ]) Read a graph from a list of edges.


write_edgelist(G, path[, comments, . . . ]) Write graph as a list of edges.
read_weighted_edgelist(path[, comments, Read a graph as list of edges with numeric weights.
. . . ])
write_weighted_edgelist(G, path[, comments, Write graph G as a list of edges with numeric weights.
. . . ])
generate_edgelist(G[, delimiter, data]) Generate a single line of the graph G in edge list format.
parse_edgelist(lines[, comments, delimiter, . . . ]) Parse lines of an edge list representation of a graph.

9.3.2 networkx.readwrite.edgelist.read_edgelist

read_edgelist(path, comments=’#’, delimiter=None, create_using=None, nodetype=None, data=True,


edgetype=None, encoding=’utf-8’)
Read a graph from a list of edges.
Parameters
• path (file or string) – File or filename to read. If a file is provided, it must be opened in ‘rb’
mode. Filenames ending in .gz or .bz2 will be uncompressed.
• comments (string, optional) – The character used to indicate the start of a comment.
• delimiter (string, optional) – The string used to separate values. The default is whitespace.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
• nodetype (int, float, str, Python type, optional) – Convert node data from strings to specified
type
• data (bool or list of (label,type) tuples) – Tuples specifying dictionary key names and types
for edge data
• edgetype (int, float, str, Python type, optional OBSOLETE) – Convert edge data from strings
to specified type and use as ‘weight’
• encoding (string, optional) – Specify which encoding to use when reading file.
Returns G – A networkx Graph or other type specified with create_using
Return type graph

Examples

>>> nx.write_edgelist(nx.path_graph(4), "test.edgelist")


>>> G=nx.read_edgelist("test.edgelist")

>>> fh=open("test.edgelist", 'rb')


>>> G=nx.read_edgelist(fh)
>>> fh.close()

9.3. Edge List 625


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G=nx.read_edgelist("test.edgelist", nodetype=int)


>>> G=nx.read_edgelist("test.edgelist",create_using=nx.DiGraph)

Edgelist with data in a list:


>>> textline = '1 2 3'
>>> fh = open('test.edgelist','w')
>>> d = fh.write(textline)
>>> fh.close()
>>> G = nx.read_edgelist('test.edgelist', nodetype=int, data=(('weight',float),))
>>> list(G)
[1, 2]
>>> list(G.edges(data=True))
[(1, 2, {'weight': 3.0})]

See parse_edgelist() for more examples of formatting.


See also:
parse_edgelist()

Notes

Since nodes must be hashable, the function nodetype must return hashable types (e.g. int, float, str, frozenset -
or tuples of those, etc.)

9.3.3 networkx.readwrite.edgelist.write_edgelist

write_edgelist(G, path, comments=’#’, delimiter=’ ’, data=True, encoding=’utf-8’)


Write graph as a list of edges.
Parameters
• G (graph) – A NetworkX graph
• path (file or string) – File or filename to write. If a file is provided, it must be opened in
‘wb’ mode. Filenames ending in .gz or .bz2 will be compressed.
• comments (string, optional) – The character used to indicate the start of a comment
• delimiter (string, optional) – The string used to separate values. The default is whitespace.
• data (bool or list, optional) – If False write no edge data. If True write a string representation
of the edge data dictionary.. If a list (or other iterable) is provided, write the keys specified
in the list.
• encoding (string, optional) – Specify which encoding to use when writing file.

Examples

>>> G=nx.path_graph(4)
>>> nx.write_edgelist(G, "test.edgelist")
>>> G=nx.path_graph(4)
>>> fh=open("test.edgelist",'wb')
>>> nx.write_edgelist(G, fh)
>>> nx.write_edgelist(G, "test.edgelist.gz")
>>> nx.write_edgelist(G, "test.edgelist.gz", data=False)

626 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G=nx.Graph()
>>> G.add_edge(1,2,weight=7,color='red')
>>> nx.write_edgelist(G,'test.edgelist',data=False)
>>> nx.write_edgelist(G,'test.edgelist',data=['color'])
>>> nx.write_edgelist(G,'test.edgelist',data=['color','weight'])

See also:
write_edgelist(), write_weighted_edgelist()

9.3.4 networkx.readwrite.edgelist.read_weighted_edgelist

read_weighted_edgelist(path, comments=’#’, delimiter=None, create_using=None, nodetype=None,


encoding=’utf-8’)
Read a graph as list of edges with numeric weights.
Parameters
• path (file or string) – File or filename to read. If a file is provided, it must be opened in ‘rb’
mode. Filenames ending in .gz or .bz2 will be uncompressed.
• comments (string, optional) – The character used to indicate the start of a comment.
• delimiter (string, optional) – The string used to separate values. The default is whitespace.
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
• nodetype (int, float, str, Python type, optional) – Convert node data from strings to specified
type
• encoding (string, optional) – Specify which encoding to use when reading file.
Returns G – A networkx Graph or other type specified with create_using
Return type graph

Notes

Since nodes must be hashable, the function nodetype must return hashable types (e.g. int, float, str, frozenset -
or tuples of those, etc.)
Example edgelist file format.
With numeric edge data:

# read with
# >>> G=nx.read_weighted_edgelist(fh)
# source target data
a b 1
a c 3.14159
d e 42

9.3.5 networkx.readwrite.edgelist.write_weighted_edgelist

write_weighted_edgelist(G, path, comments=’#’, delimiter=’ ’, encoding=’utf-8’)


Write graph G as a list of edges with numeric weights.

9.3. Edge List 627


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (graph) – A NetworkX graph
• path (file or string) – File or filename to write. If a file is provided, it must be opened in
‘wb’ mode. Filenames ending in .gz or .bz2 will be compressed.
• comments (string, optional) – The character used to indicate the start of a comment
• delimiter (string, optional) – The string used to separate values. The default is whitespace.
• encoding (string, optional) – Specify which encoding to use when writing file.

Examples

>>> G=nx.Graph()
>>> G.add_edge(1,2,weight=7)
>>> nx.write_weighted_edgelist(G, 'test.weighted.edgelist')

See also:
read_edgelist(), write_edgelist(), write_weighted_edgelist()

9.3.6 networkx.readwrite.edgelist.generate_edgelist

generate_edgelist(G, delimiter=’ ’, data=True)


Generate a single line of the graph G in edge list format.
Parameters
• G (NetworkX graph)
• delimiter (string, optional) – Separator for node labels
• data (bool or list of keys) – If False generate no edge data. If True use a dictionary repre-
sentation of edge data. If a list of keys use a list of data values corresponding to the keys.
Returns lines – Lines of data in adjlist format.
Return type string

Examples

>>> G = nx.lollipop_graph(4, 3)
>>> G[1][2]['weight'] = 3
>>> G[3][4]['capacity'] = 12
>>> for line in nx.generate_edgelist(G, data=False):
... print(line)
0 1
0 2
0 3
1 2
1 3
2 3
3 4
4 5
5 6

628 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> for line in nx.generate_edgelist(G):


... print(line)
0 1 {}
0 2 {}
0 3 {}
1 2 {'weight': 3}
1 3 {}
2 3 {}
3 4 {'capacity': 12}
4 5 {}
5 6 {}

>>> for line in nx.generate_edgelist(G,data=['weight']):


... print(line)
0 1
0 2
0 3
1 2 3
1 3
2 3
3 4
4 5
5 6

See also:
write_adjlist(), read_adjlist()

9.3.7 networkx.readwrite.edgelist.parse_edgelist

parse_edgelist(lines, comments=’#’, delimiter=None, create_using=None, nodetype=None,


data=True)
Parse lines of an edge list representation of a graph.
Parameters
• lines (list or iterator of strings) – Input data in edgelist format
• comments (string, optional) – Marker for comment lines
• delimiter (string, optional) – Separator for node labels
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.
• nodetype (Python type, optional) – Convert nodes to this type.
• data (bool or list of (label,type) tuples) – If False generate no edge data or if True use a
dictionary representation of edge data or a list tuples specifying dictionary key names and
types for edge data.
Returns G – The graph corresponding to lines
Return type NetworkX Graph

Examples

Edgelist with no data:

9.3. Edge List 629


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> lines = ["1 2",


... "2 3",
... "3 4"]
>>> G = nx.parse_edgelist(lines, nodetype = int)
>>> list(G)
[1, 2, 3, 4]
>>> list(G.edges())
[(1, 2), (2, 3), (3, 4)]

Edgelist with data in Python dictionary representation:

>>> lines = ["1 2 {'weight':3}",


... "2 3 {'weight':27}",
... "3 4 {'weight':3.0}"]
>>> G = nx.parse_edgelist(lines, nodetype = int)
>>> list(G)
[1, 2, 3, 4]
>>> list(G.edges(data=True))
[(1, 2, {'weight': 3}), (2, 3, {'weight': 27}), (3, 4, {'weight': 3.0})]

Edgelist with data in a list:

>>> lines = ["1 2 3",


... "2 3 27",
... "3 4 3.0"]
>>> G = nx.parse_edgelist(lines, nodetype = int, data=(('weight',float),))
>>> list(G)
[1, 2, 3, 4]
>>> list(G.edges(data=True))
[(1, 2, {'weight': 3.0}), (2, 3, {'weight': 27.0}), (3, 4, {'weight': 3.0})]

See also:
read_weighted_edgelist()

9.4 GEXF

Read and write graphs in GEXF format.


GEXF (Graph Exchange XML Format) is a language for describing complex network structures, their associated data
and dynamics.
This implementation does not support mixed graphs (directed and undirected edges together).

9.4.1 Format

GEXF is an XML format. See https://gephi.org/gexf/format/schema.html for the specification and https://gephi.org/
gexf/format/basic.html for examples.

read_gexf(path[, node_type, relabel, version]) Read graph in GEXF format from path.
write_gexf(G, path[, encoding, prettyprint, . . . ]) Write G in GEXF format to path.
generate_gexf(G[, encoding, prettyprint, . . . ]) Generate lines of GEXF format representation of G.
relabel_gexf_graph(G) Relabel graph using “label” node keyword for node la-
bel.

630 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

9.4.2 networkx.readwrite.gexf.read_gexf

read_gexf(path, node_type=None, relabel=False, version=’1.2draft’)


Read graph in GEXF format from path.
“GEXF (Graph Exchange XML Format) is a language for describing complex networks structures, their associ-
ated data and dynamics”1 .
Parameters
• path (file or string) – File or file name to write. File names ending in .gz or .bz2 will be
compressed.
• node_type (Python type (default: None)) – Convert node ids to this type if not None.
• relabel (bool (default: False)) – If True relabel the nodes to use the GEXF node “label”
attribute instead of the node “id” attribute as the NetworkX node label.
• version (string (default: 1.2draft)) – Version of GEFX File Format (see https://gephi.org/
gexf/format/schema.html). Supported values: “1.1draft”, “1.2draft”
Returns graph – If no parallel edges are found a Graph or DiGraph is returned. Otherwise a Multi-
Graph or MultiDiGraph is returned.
Return type NetworkX graph

Notes

This implementation does not support mixed graphs (directed and undirected edges together).

References

9.4.3 networkx.readwrite.gexf.write_gexf

write_gexf(G, path, encoding=’utf-8’, prettyprint=True, version=’1.2draft’)


Write G in GEXF format to path.
“GEXF (Graph Exchange XML Format) is a language for describing complex networks structures, their associ-
ated data and dynamics”1 .
Node attributes are checked according to the version of the GEXF schemas used for parameters which are not
user defined, e.g. visualization ‘viz’2 . See example for usage.
Parameters
• G (graph) – A NetworkX graph
• path (file or string) – File or file name to write. File names ending in .gz or .bz2 will be
compressed.
• encoding (string (optional, default: ‘utf-8’)) – Encoding for text data.
• prettyprint (bool (optional, default: True)) – If True use line breaks and indenting in output
XML.
1 GEXF File Format, https://gephi.org/gexf/format/
1 GEXF File Format, https://gephi.org/gexf/format/
2 GEXF viz schema 1.1, https://gephi.org/gexf/1.1draft/viz

9.4. GEXF 631


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(4)
>>> nx.write_gexf(G, "test.gexf")

# visualization data >>> G.nodes[0][‘viz’] = {‘size’: 54} >>> G.nodes[0][‘viz’][‘position’] = {‘x’ : 0, ‘y’ : 1}
>>> G.nodes[0][‘viz’][‘color’] = {‘r’ : 0, ‘g’ : 0, ‘b’ : 256}

Notes

This implementation does not support mixed graphs (directed and undirected edges together).
The node id attribute is set to be the string of the node label. If you want to specify an id use set it as node data,
e.g. node[‘a’][‘id’]=1 to set the id of node ‘a’ to 1.

References

9.4.4 networkx.readwrite.gexf.generate_gexf

generate_gexf(G, encoding=’utf-8’, prettyprint=True, version=’1.2draft’)


Generate lines of GEXF format representation of G.
“GEXF (Graph Exchange XML Format) is a language for describing complex networks structures, their associ-
ated data and dynamics”1 .
Parameters
• G (graph) – A NetworkX graph
• encoding (string (optional, default: ‘utf-8’)) – Encoding for text data.
• prettyprint (bool (optional, default: True)) – If True use line breaks and indenting in output
XML.
• version (string (default: 1.2draft)) – Version of GEFX File Format (see https://gephi.org/
gexf/format/schema.html). Supported values: “1.1draft”, “1.2draft”

Examples

>>> G = nx.path_graph(4)
>>> linefeed = chr(10) # linefeed=

>>> s = linefeed.join(nx.generate_gexf(G)) # doctest: +SKIP


>>> for line in nx.generate_gexf(G): # doctest: +SKIP
... print line

Notes

This implementation does not support mixed graphs (directed and undirected edges together).
The node id attribute is set to be the string of the node label. If you want to specify an id use set it as node data,
e.g. node[‘a’][‘id’]=1 to set the id of node ‘a’ to 1.
1 GEXF File Format, https://gephi.org/gexf/format/

632 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

9.4.5 networkx.readwrite.gexf.relabel_gexf_graph

relabel_gexf_graph(G)
Relabel graph using “label” node keyword for node label.
Parameters G (graph) – A NetworkX graph read from GEXF data
Returns H – A NetworkX graph with relabed nodes
Return type graph
Raises NetworkXError – If node labels are missing or not unique while relabel=True.

Notes

This function relabels the nodes in a NetworkX graph with the “label” attribute. It also handles relabeling the
specific GEXF node attributes “parents”, and “pid”.

9.5 GML

Read graphs in GML format.


“GML, the G>raph Modelling Language, is our proposal for a portable file format for graphs. GML’s key features are
portability, simple syntax, extensibility and flexibility. A GML file consists of a hierarchical key-value lists. Graphs
can be annotated with arbitrary data structures. The idea for a common file format was born at the GD‘95; this proposal
is the outcome of many discussions. GML is the standard file format in the Graphlet graph editor system. It has been
overtaken and adapted by several other systems for drawing graphs.”
GML files are stored using a 7-bit ASCII encoding with any extended ASCII characters (iso8859-1) appearing as
HTML character entities. You will need to give some thought into how the exported data should interact with different
languages and even different Python versions. Re-importing from gml is also a concern.
Without specifying a stringizer/destringizer, the code is capable of handling
int/float/str/dict/list data as required by the GML specification. For other data types, you need to
explicitly supply a stringizer/destringizer.
For better interoperability of data generated by Python 2 and Python 3, we’ve provided literal_stringizer
and literal_destringizer.
For additional documentation on the GML file format, please see the GML website.
Several example graphs in GML format may be found on Mark Newman’s Network data page.

read_gml(path[, label, destringizer]) Read graph in GML format from path.


write_gml(G, path[, stringizer]) Write a graph G in GML format to the file or file handle
path.
parse_gml(lines[, label, destringizer]) Parse GML graph from a string or iterable.
generate_gml(G[, stringizer]) Generate a single entry of the graph G in GML format.
literal_destringizer(rep) Convert a Python literal to the value it represents.
literal_stringizer(value) Convert a value to a Python literal in GML represen-
tation.

9.5. GML 633


NetworkX Reference, Release 2.3rc1.dev20190222214247

9.5.1 networkx.readwrite.gml.read_gml

read_gml(path, label=’label’, destringizer=None)


Read graph in GML format from path.
Parameters
• path (filename or filehandle) – The filename or filehandle to read from.
• label (string, optional) – If not None, the parsed nodes will be renamed according to node
attributes indicated by label. Default value: ‘label’.
• destringizer (callable, optional) – A destringizer that recovers values stored as strings
in GML. If it cannot convert a string to a value, a ValueError is raised. Default value :
None.
Returns G – The parsed graph.
Return type NetworkX graph
Raises NetworkXError – If the input cannot be parsed.
See also:
write_gml(), parse_gml(), literal_destringizer()

Notes

GML files are stored using a 7-bit ASCII encoding with any extended ASCII characters (iso8859-1) appearing
as HTML character entities. Without specifying a stringizer/destringizer, the code is capable of
handling int/float/str/dict/list data as required by the GML specification. For other data types, you
need to explicitly supply a stringizer/destringizer.
For additional documentation on the GML file format, please see the GML website.
See the module docstring networkx.readwrite.gml for more details.

Examples

>>> G = nx.path_graph(4)
>>> nx.write_gml(G, 'test.gml')
>>> H = nx.read_gml('test.gml')

9.5.2 networkx.readwrite.gml.write_gml

write_gml(G, path, stringizer=None)


Write a graph G in GML format to the file or file handle path.
Parameters
• G (NetworkX graph) – The graph to be converted to GML.
• path (filename or filehandle) – The filename or filehandle to write. Files whose names end
with .gz or .bz2 will be compressed.
• stringizer (callable, optional) – A stringizer which converts non-int/non-float/non-dict
values into strings. If it cannot convert a value into a string, it should raise a ValueError
to indicate that. Default value: None.

634 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

Raises NetworkXError – If stringizer cannot convert a value into a string, or the value to
convert is not a string while stringizer is None.
See also:
read_gml(), generate_gml(), literal_stringizer()

Notes

Graph attributes named ‘directed’, ‘multigraph’, ‘node’ or ‘edge’, node attributes named ‘id’ or ‘label’, edge
attributes named ‘source’ or ‘target’ (or ‘key’ if G is a multigraph) are ignored because these attribute names are
used to encode the graph structure.
GML files are stored using a 7-bit ASCII encoding with any extended ASCII characters (iso8859-1) appearing
as HTML character entities. Without specifying a stringizer/destringizer, the code is capable of
handling int/float/str/dict/list data as required by the GML specification. For other data types, you
need to explicitly supply a stringizer/destringizer.
Note that while we allow non-standard GML to be read from a file, we make sure to write GML format. In
particular, underscores are not allowed in attribute names. For additional documentation on the GML file format,
please see the GML website.
See the module docstring networkx.readwrite.gml for more details.

Examples

>>> G = nx.path_graph(4)
>>> nx.write_gml(G, "test.gml")

Filenames ending in .gz or .bz2 will be compressed.

>>> nx.write_gml(G, "test.gml.gz")

9.5.3 networkx.readwrite.gml.parse_gml

parse_gml(lines, label=’label’, destringizer=None)


Parse GML graph from a string or iterable.
Parameters
• lines (string or iterable of strings) – Data in GML format.
• label (string, optional) – If not None, the parsed nodes will be renamed according to node
attributes indicated by label. Default value: ‘label’.
• destringizer (callable, optional) – A destringizer that recovers values stored as strings
in GML. If it cannot convert a string to a value, a ValueError is raised. Default value :
None.
Returns G – The parsed graph.
Return type NetworkX graph
Raises NetworkXError – If the input cannot be parsed.
See also:
write_gml(), read_gml(), literal_destringizer()

9.5. GML 635


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This stores nested GML attributes as dictionaries in the NetworkX graph, node, and edge attribute structures.
GML files are stored using a 7-bit ASCII encoding with any extended ASCII characters (iso8859-1) appearing
as HTML character entities. Without specifying a stringizer/destringizer, the code is capable of
handling int/float/str/dict/list data as required by the GML specification. For other data types, you
need to explicitly supply a stringizer/destringizer.
For additional documentation on the GML file format, please see the GML website.
See the module docstring networkx.readwrite.gml for more details.

9.5.4 networkx.readwrite.gml.generate_gml

generate_gml(G, stringizer=None)
Generate a single entry of the graph G in GML format.
Parameters
• G (NetworkX graph) – The graph to be converted to GML.
• stringizer (callable, optional) – A stringizer which converts non-int/non-float/non-dict
values into strings. If it cannot convert a value into a string, it should raise a ValueError
to indicate that. Default value: None.
Returns lines – Lines of GML data. Newlines are not appended.
Return type generator of strings
Raises NetworkXError – If stringizer cannot convert a value into a string, or the value to
convert is not a string while stringizer is None.
See also:
literal_stringizer()

Notes

Graph attributes named ‘directed’, ‘multigraph’, ‘node’ or ‘edge’, node attributes named ‘id’ or ‘label’, edge
attributes named ‘source’ or ‘target’ (or ‘key’ if G is a multigraph) are ignored because these attribute names are
used to encode the graph structure.
GML files are stored using a 7-bit ASCII encoding with any extended ASCII characters (iso8859-1) appearing
as HTML character entities. Without specifying a stringizer/destringizer, the code is capable of
handling int/float/str/dict/list data as required by the GML specification. For other data types, you
need to explicitly supply a stringizer/destringizer.
For additional documentation on the GML file format, please see the GML website.
See the module docstring networkx.readwrite.gml for more details.

Examples

>>> G = nx.Graph()
>>> G.add_node("1")
>>> print("\n".join(nx.generate_gml(G)))
graph [
(continues on next page)

636 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


node [
id 0
label "1"
]
]
>>> G = nx.OrderedMultiGraph([("a", "b"), ("a", "b")])
>>> print("\n".join(nx.generate_gml(G)))
graph [
multigraph 1
node [
id 0
label "a"
]
node [
id 1
label "b"
]
edge [
source 0
target 1
key 0
]
edge [
source 0
target 1
key 1
]
]

9.5.5 networkx.readwrite.gml.literal_destringizer

literal_destringizer(rep)
Convert a Python literal to the value it represents.
Parameters rep (string) – A Python literal.
Returns value – The value of the Python literal.
Return type object
Raises ValueError – If rep is not a Python literal.

9.5.6 networkx.readwrite.gml.literal_stringizer

literal_stringizer(value)
Convert a value to a Python literal in GML representation.
Parameters value (object) – The value to be converted to GML representation.
Returns rep – A double-quoted Python literal representing value. Unprintable characters are re-
placed by XML character references.
Return type string
Raises ValueError – If value cannot be converted to GML.

9.5. GML 637


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

literal_stringizer is largely the same as repr in terms of functionality but attempts prefix unicode
and bytes literals with u and b to provide better interoperability of data generated by Python 2 and Python 3.
The original value can be recovered using the networkx.readwrite.gml.
literal_destringizer() function.

9.6 Pickle

9.6.1 Pickled Graphs

Read and write NetworkX graphs as Python pickles.


“The pickle module implements a fundamental, but powerful algorithm for serializing and de-serializing a Python
object structure. “Pickling” is the process whereby a Python object hierarchy is converted into a byte stream, and
“unpickling” is the inverse operation, whereby a byte stream is converted back into an object hierarchy.”
Note that NetworkX graphs can contain any hashable Python object as node (not just integers and strings). For arbitrary
data types it may be difficult to represent the data as text. In that case using Python pickles to store the graph data can
be used.

Format

See https://docs.python.org/2/library/pickle.html

read_gpickle(path) Read graph object in Python pickle format.


write_gpickle(G, path[, protocol]) Write graph in Python pickle format.

9.6.2 networkx.readwrite.gpickle.read_gpickle

read_gpickle(path)
Read graph object in Python pickle format.
Pickles are a serialized byte stream of a Python object1 . This format will preserve Python objects used as nodes
or edges.
Parameters path (file or string) – File or filename to write. Filenames ending in .gz or .bz2 will be
uncompressed.
Returns G – A NetworkX graph
Return type graph

Examples

>>> G = nx.path_graph(4)
>>> nx.write_gpickle(G, "test.gpickle")
>>> G = nx.read_gpickle("test.gpickle")

1 https://docs.python.org/2/library/pickle.html

638 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

9.6.3 networkx.readwrite.gpickle.write_gpickle

write_gpickle(G, path, protocol=4)


Write graph in Python pickle format.
Pickles are a serialized byte stream of a Python object1 . This format will preserve Python objects used as nodes
or edges.
Parameters
• G (graph) – A NetworkX graph
• path (file or string) – File or filename to write. Filenames ending in .gz or .bz2 will be
compressed.
• protocol (integer) – Pickling protocol to use. Default value: pickle.
HIGHEST_PROTOCOL.

Examples

>>> G = nx.path_graph(4)
>>> nx.write_gpickle(G, "test.gpickle")

References

9.7 GraphML

9.7.1 GraphML

Read and write graphs in GraphML format.


This implementation does not support mixed graphs (directed and unidirected edges together), hyperedges, nested
graphs, or ports.
“GraphML is a comprehensive and easy-to-use file format for graphs. It consists of a language core to describe the
structural properties of a graph and a flexible extension mechanism to add application-specific data. Its main features
include support of
• directed, undirected, and mixed graphs,
• hypergraphs,
• hierarchical graphs,
• graphical representations,
• references to external data,
• application-specific attribute data, and
• light-weight parsers.
1 https://docs.python.org/2/library/pickle.html

9.7. GraphML 639


NetworkX Reference, Release 2.3rc1.dev20190222214247

Unlike many other file formats for graphs, GraphML does not use a custom syntax. Instead, it is based on XML and
hence ideally suited as a common denominator for all kinds of services generating, archiving, or processing graphs.”
http://graphml.graphdrawing.org/

Format

GraphML is an XML format. See http://graphml.graphdrawing.org/specification.html for the specification and http:
//graphml.graphdrawing.org/primer/graphml-primer.html for examples.

read_graphml(path[, node_type, edge_key_type]) Read graph in GraphML format from path.


write_graphml(G, path[, encoding, . . . ]) Write G in GraphML XML format to path
generate_graphml(G[, encoding, prettyprint]) Generate GraphML lines for G
parse_graphml(graphml_string[, node_type]) Read graph in GraphML format from string.

9.7.2 networkx.readwrite.graphml.read_graphml

read_graphml(path, node_type=<class ’str’>, edge_key_type=<class ’int’>)


Read graph in GraphML format from path.
Parameters
• path (file or string) – File or filename to write. Filenames ending in .gz or .bz2 will be
compressed.
• node_type (Python type (default: str)) – Convert node ids to this type
• edge_key_type (Python type (default: int)) – Convert graphml edge ids to this type as key
of multi-edges
Returns graph – If no parallel edges are found a Graph or DiGraph is returned. Otherwise a Multi-
Graph or MultiDiGraph is returned.
Return type NetworkX graph

Notes

Default node and edge attributes are not propagated to each node and edge. They can be obtained from G.
graph and applied to node and edge attributes if desired using something like this:

>>> default_color = G.graph['node_default']['color'] # doctest: +SKIP


>>> for node, data in G.nodes(data=True): # doctest: +SKIP
... if 'color' not in data:
... data['color']=default_color
>>> default_color = G.graph['edge_default']['color'] # doctest: +SKIP
>>> for u, v, data in G.edges(data=True): # doctest: +SKIP
... if 'color' not in data:
... data['color']=default_color

This implementation does not support mixed graphs (directed and unidirected edges together), hypergraphs,
nested graphs, or ports.
For multigraphs the GraphML edge “id” will be used as the edge key. If not specified then they “key” attribute
will be used. If there is no “key” attribute a default NetworkX multigraph edge key will be provided.
Files with the yEd “yfiles” extension will can be read but the graphics information is discarded.

640 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

yEd compressed files (“file.graphmlz” extension) can be read by renaming the file to “file.graphml.gz”.

9.7.3 networkx.readwrite.graphml.write_graphml

write_graphml(G, path, encoding=’utf-8’, prettyprint=True, infer_numeric_types=False)


Write G in GraphML XML format to path
This function uses the LXML framework and should be faster than the version using the xml library.
Parameters
• G (graph) – A networkx graph
• path (file or string) – File or filename to write. Filenames ending in .gz or .bz2 will be
compressed.
• encoding (string (optional)) – Encoding for text data.
• prettyprint (bool (optional)) – If True use line breaks and indenting in output XML.
• infer_numeric_types (boolean) – Determine if numeric types should be generalized. For
example, if edges have both int and float ‘weight’ attributes, we infer in GraphML that both
are floats.

Examples

>>> G = nx.path_graph(4)
>>> nx.write_graphml_lxml(G, "fourpath.graphml") # doctest: +SKIP

Notes

This implementation does not support mixed graphs (directed and unidirected edges together) hyperedges,
nested graphs, or ports.

9.7.4 networkx.readwrite.graphml.generate_graphml

generate_graphml(G, encoding=’utf-8’, prettyprint=True)


Generate GraphML lines for G
Parameters
• G (graph) – A networkx graph
• encoding (string (optional)) – Encoding for text data.
• prettyprint (bool (optional)) – If True use line breaks and indenting in output XML.

Examples

>>> G = nx.path_graph(4)
>>> linefeed = chr(10) # linefeed =

>>> s = linefeed.join(nx.generate_graphml(G)) # doctest: +SKIP


>>> for line in nx.generate_graphml(G): # doctest: +SKIP
... print(line)

9.7. GraphML 641


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

This implementation does not support mixed graphs (directed and unidirected edges together) hyperedges,
nested graphs, or ports.

9.7.5 networkx.readwrite.graphml.parse_graphml

parse_graphml(graphml_string, node_type=<class ’str’>)


Read graph in GraphML format from string.
Parameters
• graphml_string (string) – String containing graphml information (e.g., contents of a
graphml file).
• node_type (Python type (default: str)) – Convert node ids to this type
Returns graph – If no parallel edges are found a Graph or DiGraph is returned. Otherwise a Multi-
Graph or MultiDiGraph is returned.
Return type NetworkX graph

Examples

>>> G = nx.path_graph(4)
>>> linefeed = chr(10) # linefeed =

>>> s = linefeed.join(nx.generate_graphml(G))
>>> H = nx.parse_graphml(s)

Notes

Default node and edge attributes are not propagated to each node and edge. They can be obtained from G.
graph and applied to node and edge attributes if desired using something like this:

>>> default_color = G.graph['node_default']['color'] # doctest: +SKIP


>>> for node, data in G.nodes(data=True): # doctest: +SKIP
... if 'color' not in data:
... data['color']=default_color
>>> default_color = G.graph['edge_default']['color'] # doctest: +SKIP
>>> for u, v, data in G.edges(data=True): # doctest: +SKIP
... if 'color' not in data:
... data['color']=default_color

This implementation does not support mixed graphs (directed and unidirected edges together), hypergraphs,
nested graphs, or ports.
For multigraphs the GraphML edge “id” will be used as the edge key. If not specified then they “key” attribute
will be used. If there is no “key” attribute a default NetworkX multigraph edge key will be provided.

642 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

9.8 JSON

9.8.1 JSON data

Generate and parse JSON serializable data for NetworkX graphs.


These formats are suitable for use with the d3.js examples https://d3js.org/
The three formats that you can generate with NetworkX are:
• node-link like in the d3.js example https://bl.ocks.org/mbostock/4062045
• tree like in the d3.js example https://bl.ocks.org/mbostock/4063550
• adjacency like in the d3.js example https://bost.ocks.org/mike/miserables/

node_link_data(G[, attrs]) Returns data in node-link format that is suitable for


JSON serialization and use in Javascript documents.
node_link_graph(data[, directed, . . . ]) Returns graph from node-link data format.
adjacency_data(G[, attrs]) Returns data in adjacency format that is suitable for
JSON serialization and use in Javascript documents.
adjacency_graph(data[, directed, . . . ]) Returns graph from adjacency data format.
cytoscape_data(G[, attrs]) Returns data in Cytoscape JSON format (cyjs).
cytoscape_graph(data[, attrs])
tree_data(G, root[, attrs]) Returns data in tree format that is suitable for JSON se-
rialization and use in Javascript documents.
tree_graph(data[, attrs]) Returns graph from tree data format.
jit_data(G[, indent]) Returns data in JIT JSON format.
jit_graph(data[, create_using]) Read a graph from JIT JSON.

9.8.2 networkx.readwrite.json_graph.node_link_data

node_link_data(G, attrs=None)
Returns data in node-link format that is suitable for JSON serialization and use in Javascript documents.
Parameters
• G (NetworkX graph)
• attrs (dict) – A dictionary that contains five keys ‘source’, ‘target’, ‘name’, ‘key’ and ‘link’.
The corresponding values provide the attribute names for storing NetworkX-internal graph
data. The values should be unique. Default value:

dict(source='source', target='target', name='id',


key='key', link='links')

If some user-defined graph data use these attribute names as data keys, they may be silently
dropped.
Returns data – A dictionary with node-link formatted data.
Return type dict
Raises NetworkXError – If values in attrs are not unique.

9.8. JSON 643


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> from networkx.readwrite import json_graph


>>> G = nx.Graph([('A', 'B')])
>>> data1 = json_graph.node_link_data(G)
>>> H = nx.gn_graph(2)
>>> data2 = json_graph.node_link_data(H, {'link': 'edges', 'source': 'from',
˓→'target': 'to'})

To serialize with json

>>> import json


>>> s1 = json.dumps(data1)
>>> s2 = json.dumps(data2, default={'link': 'edges', 'source': 'from', 'target':
˓→'to'})

Notes

Graph, node, and link attributes are stored in this format. Note that attribute keys will be converted to strings in
order to comply with JSON.
Attribute ‘key’ is only used for multigraphs.
See also:
node_link_graph(), adjacency_data(), tree_data()

9.8.3 networkx.readwrite.json_graph.node_link_graph

node_link_graph(data, directed=False, multigraph=True, attrs=None)


Returns graph from node-link data format.
Parameters
• data (dict) – node-link formatted graph data
• directed (bool) – If True, and direction not specified in data, return a directed graph.
• multigraph (bool) – If True, and multigraph not specified in data, return a multigraph.
• attrs (dict) – A dictionary that contains five keys ‘source’, ‘target’, ‘name’, ‘key’ and ‘link’.
The corresponding values provide the attribute names for storing NetworkX-internal graph
data. Default value:
dict(source=’source’, target=’target’, name=’id’, key=’key’, link=’links’)
Returns G – A NetworkX graph object
Return type NetworkX graph

Examples

>>> from networkx.readwrite import json_graph


>>> G = nx.Graph([('A', 'B')])
>>> data = json_graph.node_link_data(G)
>>> H = json_graph.node_link_graph(data)

644 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

Attribute ‘key’ is only used for multigraphs.


See also:
node_link_data(), adjacency_data(), tree_data()

9.8.4 networkx.readwrite.json_graph.adjacency_data

adjacency_data(G, attrs={’id’: ’id’, ’key’: ’key’})


Returns data in adjacency format that is suitable for JSON serialization and use in Javascript documents.
Parameters
• G (NetworkX graph)
• attrs (dict) – A dictionary that contains two keys ‘id’ and ‘key’. The corresponding values
provide the attribute names for storing NetworkX-internal graph data. The values should be
unique. Default value: dict(id='id', key='key').
If some user-defined graph data use these attribute names as data keys, they may be silently
dropped.
Returns data – A dictionary with adjacency formatted data.
Return type dict
Raises NetworkXError – If values in attrs are not unique.

Examples

>>> from networkx.readwrite import json_graph


>>> G = nx.Graph([(1,2)])
>>> data = json_graph.adjacency_data(G)

To serialize with json

>>> import json


>>> s = json.dumps(data)

Notes

Graph, node, and link attributes will be written when using this format but attribute keys must be strings if you
want to serialize the resulting data with JSON.
The default value of attrs will be changed in a future release of NetworkX.
See also:
adjacency_graph(), node_link_data(), tree_data()

9.8.5 networkx.readwrite.json_graph.adjacency_graph

adjacency_graph(data, directed=False, multigraph=True, attrs={’id’: ’id’, ’key’: ’key’})


Returns graph from adjacency data format.

9.8. JSON 645


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters data (dict) – Adjacency list formatted graph data


Returns
• G (NetworkX graph) – A NetworkX graph object
• directed (bool) – If True, and direction not specified in data, return a directed graph.
• multigraph (bool) – If True, and multigraph not specified in data, return a multigraph.
• attrs (dict) – A dictionary that contains two keys ‘id’ and ‘key’. The corresponding values
provide the attribute names for storing NetworkX-internal graph data. The values should be
unique. Default value: dict(id='id', key='key').

Examples

>>> from networkx.readwrite import json_graph


>>> G = nx.Graph([(1,2)])
>>> data = json_graph.adjacency_data(G)
>>> H = json_graph.adjacency_graph(data)

Notes

The default value of attrs will be changed in a future release of NetworkX.


See also:
adjacency_graph(), node_link_data(), tree_data()

9.8.6 networkx.readwrite.json_graph.cytoscape_data

cytoscape_data(G, attrs=None)
Returns data in Cytoscape JSON format (cyjs).
Parameters G (NetworkX Graph)
Returns data – A dictionary with cyjs formatted data.
Return type dict
Raises NetworkXError – If values in attrs are not unique.

9.8.7 networkx.readwrite.json_graph.cytoscape_graph

cytoscape_graph(data, attrs=None)

9.8.8 networkx.readwrite.json_graph.tree_data

tree_data(G, root, attrs={’children’: ’children’, ’id’: ’id’})


Returns data in tree format that is suitable for JSON serialization and use in Javascript documents.
Parameters
• G (NetworkX graph) – G must be an oriented tree
• root (node) – The root of the tree

646 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

• attrs (dict) – A dictionary that contains two keys ‘id’ and ‘children’. The corresponding
values provide the attribute names for storing NetworkX-internal graph data. The values
should be unique. Default value: dict(id='id', children='children').
If some user-defined graph data use these attribute names as data keys, they may be silently
dropped.
Returns data – A dictionary with node-link formatted data.
Return type dict
Raises NetworkXError – If values in attrs are not unique.

Examples

>>> from networkx.readwrite import json_graph


>>> G = nx.DiGraph([(1,2)])
>>> data = json_graph.tree_data(G,root=1)

To serialize with json

>>> import json


>>> s = json.dumps(data)

Notes

Node attributes are stored in this format but keys for attributes must be strings if you want to serialize with
JSON.
Graph and edge attributes are not stored.
The default value of attrs will be changed in a future release of NetworkX.
See also:
tree_graph(), node_link_data(), node_link_data()

9.8.9 networkx.readwrite.json_graph.tree_graph

tree_graph(data, attrs={’children’: ’children’, ’id’: ’id’})


Returns graph from tree data format.
Parameters data (dict) – Tree formatted graph data
Returns
• G (NetworkX DiGraph)
• attrs (dict) – A dictionary that contains two keys ‘id’ and ‘children’. The corresponding
values provide the attribute names for storing NetworkX-internal graph data. The values
should be unique. Default value: dict(id='id', children='children').

Examples

9.8. JSON 647


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> from networkx.readwrite import json_graph


>>> G = nx.DiGraph([(1,2)])
>>> data = json_graph.tree_data(G,root=1)
>>> H = json_graph.tree_graph(data)

Notes

The default value of attrs will be changed in a future release of NetworkX.


See also:
tree_graph(), node_link_data(), adjacency_data()

9.8.10 networkx.readwrite.json_graph.jit_data

jit_data(G, indent=None)
Returns data in JIT JSON format.
Parameters
• G (NetworkX Graph)
• indent (optional, default=None) – If indent is a non-negative integer, then JSON array el-
ements and object members will be pretty-printed with that indent level. An indent level
of 0, or negative, will only insert newlines. None (the default) selects the most compact
representation.
Returns data
Return type JIT JSON string

9.8.11 networkx.readwrite.json_graph.jit_graph

jit_graph(data, create_using=None)
Read a graph from JIT JSON.
Parameters
• data (JSON Graph Object)
• create_using (Networkx Graph, optional (default: Graph())) – Return graph of this type.
The provided instance will be cleared.
Returns G
Return type NetworkX Graph built from create_using if provided.

9.9 LEDA

Read graphs in LEDA format.


LEDA is a C++ class library for efficient data types and algorithms.

648 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

9.9.1 Format

See http://www.algorithmic-solutions.info/leda_guide/graphs/leda_native_graph_fileformat.html

read_leda(path[, encoding]) Read graph in LEDA format from path.


parse_leda(lines) Read graph in LEDA format from string or iterable.

9.9.2 networkx.readwrite.leda.read_leda

read_leda(path, encoding=’UTF-8’)
Read graph in LEDA format from path.
Parameters path (file or string) – File or filename to read. Filenames ending in .gz or .bz2 will be
uncompressed.
Returns G
Return type NetworkX graph

Examples

G=nx.read_leda(‘file.leda’)

References

9.9.3 networkx.readwrite.leda.parse_leda

parse_leda(lines)
Read graph in LEDA format from string or iterable.
Parameters lines (string or iterable) – Data in LEDA format.
Returns G
Return type NetworkX graph

Examples

G=nx.parse_leda(string)

References

9.10 YAML

9.10.1 YAML

Read and write NetworkX graphs in YAML format.


“YAML is a data serialization format designed for human readability and interaction with scripting languages.” See
http://www.yaml.org for documentation.

9.10. YAML 649


NetworkX Reference, Release 2.3rc1.dev20190222214247

Format

http://pyyaml.org/wiki/PyYAML

read_yaml(path) Read graph in YAML format from path.


write_yaml(G_to_be_yaml, . . . ) Write graph G in YAML format to path.

9.10.2 networkx.readwrite.nx_yaml.read_yaml

read_yaml(path)
Read graph in YAML format from path.
YAML is a data serialization format designed for human readability and interaction with scripting languages1 .
Parameters path (file or string) – File or filename to read. Filenames ending in .gz or .bz2 will be
uncompressed.
Returns G
Return type NetworkX graph

Examples

>>> G=nx.path_graph(4)
>>> nx.write_yaml(G,'test.yaml')
>>> G=nx.read_yaml('test.yaml')

References

9.10.3 networkx.readwrite.nx_yaml.write_yaml

write_yaml(G_to_be_yaml, path_for_yaml_output, **kwds)


Write graph G in YAML format to path.
YAML is a data serialization format designed for human readability and interaction with scripting languages1 .
Parameters
• G (graph) – A NetworkX graph
• path (file or string) – File or filename to write. Filenames ending in .gz or .bz2 will be
compressed.

Notes

To use encoding on the output file include e.g. encoding='utf-8' in the keyword arguments.
1 http://www.yaml.org
1 http://www.yaml.org

650 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G=nx.path_graph(4)
>>> nx.write_yaml(G,'test.yaml')

References

9.11 SparseGraph6

Functions for reading and writing graphs in the graph6 or sparse6 file formats.
According to the author of these formats,
graph6 and sparse6 are formats for storing undirected graphs in a compact manner, using only printable
ASCII characters. Files in these formats have text type and contain one line per graph.
graph6 is suitable for small graphs, or large dense graphs. sparse6 is more space-efficient for large sparse
graphs.
—graph6 and sparse6 homepage

9.11.1 Graph6

Functions for reading and writing graphs in the graph6 format.


The graph6 file format is suitable for small graphs or large dense graphs. For large sparse graphs, use the sparse6
format.
For more information, see the graph6 homepage.

from_graph6_bytes(string) Read a simple undirected graph in graph6 format from


string.
read_graph6(path) Read simple undirected graphs in graph6 format from
path.
to_graph6_bytes(G[, nodes, header]) Convert a simple undirected graph to bytes in graph6
format.
write_graph6(G, path[, nodes, header]) Write a simple undirected graph to a path in graph6 for-
mat.

networkx.readwrite.graph6.from_graph6_bytes

from_graph6_bytes(string)
Read a simple undirected graph in graph6 format from string.
Parameters string (string) – Data in graph6 format, without a trailing newline.
Returns G
Return type Graph
Raises
• NetworkXError – If the string is unable to be parsed in graph6 format

9.11. SparseGraph6 651


NetworkX Reference, Release 2.3rc1.dev20190222214247

• ValueError – If any character c in the input string does not satisfy 63 <= ord(c) <
127.

Examples

>>> G = nx.from_graph6_bytes(b'A_')
>>> sorted(G.edges())
[(0, 1)]

See also:
read_graph6(), write_graph6()

References

networkx.readwrite.graph6.read_graph6

read_graph6(path)
Read simple undirected graphs in graph6 format from path.
Parameters path (file or string) – File or filename to write.
Returns G – If the file contains multiple lines then a list of graphs is returned
Return type Graph or list of Graphs
Raises NetworkXError – If the string is unable to be parsed in graph6 format

Examples

You can read a graph6 file by giving the path to the file:

>>> import tempfile


>>> with tempfile.NamedTemporaryFile() as f:
... _ = f.write(b'>>graph6<<A_\n')
... _ = f.seek(0)
... G = nx.read_graph6(f.name)
>>> list(G.edges())
[(0, 1)]

You can also read a graph6 file by giving an open file-like object:

>>> import tempfile


>>> with tempfile.NamedTemporaryFile() as f:
... _ = f.write(b'>>graph6<<A_\n')
... _ = f.seek(0)
... G = nx.read_graph6(f)
>>> list(G.edges())
[(0, 1)]

See also:
from_graph6_bytes(), write_graph6()

652 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

References

networkx.readwrite.graph6.to_graph6_bytes

to_graph6_bytes(G, nodes=None, header=True)


Convert a simple undirected graph to bytes in graph6 format.
Parameters
• G (Graph (undirected))
• nodes (list or iterable) – Nodes are labeled 0. . . n-1 in the order provided. If None the
ordering given by G.nodes() is used.
• header (bool) – If True add ‘>>graph6<<’ bytes to head of data.
Raises
• NetworkXNotImplemented – If the graph is directed or is a multigraph.
• ValueError – If the graph has at least 2 ** 36 nodes; the graph6 format is only defined
for graphs of order less than 2 ** 36.

Examples

>>> nx.to_graph6_bytes(nx.path_graph(2)) # doctest: +SKIP


b'>>graph6<<A_\n'

See also:
from_graph6_bytes(), read_graph6(), write_graph6_bytes()

Notes

The returned bytes end with a newline character.


The format does not support edge or node labels, parallel edges or self loops. If self loops are present they are
silently ignored.

References

networkx.readwrite.graph6.write_graph6

write_graph6(G, path, nodes=None, header=True)


Write a simple undirected graph to a path in graph6 format.
Parameters
• G (Graph (undirected))
• path (str) – The path naming the file to which to write the graph.
• nodes (list or iterable) – Nodes are labeled 0. . . n-1 in the order provided. If None the
ordering given by G.nodes() is used.
• header (bool) – If True add ‘>>graph6<<’ string to head of data
Raises
• NetworkXNotImplemented – If the graph is directed or is a multigraph.

9.11. SparseGraph6 653


NetworkX Reference, Release 2.3rc1.dev20190222214247

• ValueError – If the graph has at least 2 ** 36 nodes; the graph6 format is only defined
for graphs of order less than 2 ** 36.

Examples

You can write a graph6 file by giving the path to a file:

>>> import tempfile


>>> with tempfile.NamedTemporaryFile() as f:
... nx.write_graph6(nx.path_graph(2), f.name)
... _ = f.seek(0)
... print(f.read())
b'>>graph6<<A_\n'

See also:
from_graph6_bytes(), read_graph6()

Notes

The function writes a newline character after writing the encoding of the graph.
The format does not support edge or node labels, parallel edges or self loops. If self loops are present they are
silently ignored.

References

9.11.2 Sparse6

Functions for reading and writing graphs in the sparse6 format.


The sparse6 file format is a space-efficient format for large sparse graphs. For small graphs or large dense graphs, use
the graph6 file format.
For more information, see the sparse6 homepage.

from_sparse6_bytes(string) Read an undirected graph in sparse6 format from string.


read_sparse6(path) Read an undirected graph in sparse6 format from path.
to_sparse6_bytes(G[, nodes, header]) Convert an undirected graph to bytes in sparse6 format.
write_sparse6(G, path[, nodes, header]) Write graph G to given path in sparse6 format.

networkx.readwrite.sparse6.from_sparse6_bytes

from_sparse6_bytes(string)
Read an undirected graph in sparse6 format from string.
Parameters string (string) – Data in sparse6 format
Returns G
Return type Graph
Raises NetworkXError – If the string is unable to be parsed in sparse6 format

654 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.from_sparse6_bytes(b':A_')
>>> sorted(G.edges())
[(0, 1), (0, 1), (0, 1)]

See also:
read_sparse6(), write_sparse6()

References

networkx.readwrite.sparse6.read_sparse6

read_sparse6(path)
Read an undirected graph in sparse6 format from path.
Parameters path (file or string) – File or filename to write.
Returns G – If the file contains multiple lines then a list of graphs is returned
Return type Graph/Multigraph or list of Graphs/MultiGraphs
Raises NetworkXError – If the string is unable to be parsed in sparse6 format

Examples

You can read a sparse6 file by giving the path to the file:
>>> import tempfile
>>> with tempfile.NamedTemporaryFile() as f:
... _ = f.write(b'>>sparse6<<:An\n')
... _ = f.seek(0)
... G = nx.read_sparse6(f.name)
>>> list(G.edges())
[(0, 1)]

You can also read a sparse6 file by giving an open file-like object:
>>> import tempfile
>>> with tempfile.NamedTemporaryFile() as f:
... _ = f.write(b'>>sparse6<<:An\n')
... _ = f.seek(0)
... G = nx.read_sparse6(f)
>>> list(G.edges())
[(0, 1)]

See also:
read_sparse6(), from_sparse6_bytes()

References

networkx.readwrite.sparse6.to_sparse6_bytes

to_sparse6_bytes(G, nodes=None, header=True)


Convert an undirected graph to bytes in sparse6 format.

9.11. SparseGraph6 655


NetworkX Reference, Release 2.3rc1.dev20190222214247

Parameters
• G (Graph (undirected))
• nodes (list or iterable) – Nodes are labeled 0. . . n-1 in the order provided. If None the
ordering given by G.nodes() is used.
• header (bool) – If True add ‘>>sparse6<<’ bytes to head of data.
Raises
• NetworkXNotImplemented – If the graph is directed.
• ValueError – If the graph has at least 2 ** 36 nodes; the sparse6 format is only de-
fined for graphs of order less than 2 ** 36.

Examples

>>> nx.to_sparse6_bytes(nx.path_graph(2)) # doctest: +SKIP


b'>>sparse6<<:An\n'

See also:
to_sparse6_bytes(), read_sparse6(), write_sparse6_bytes()

Notes

The returned bytes end with a newline character.


The format does not support edge or node labels.

References

networkx.readwrite.sparse6.write_sparse6

write_sparse6(G, path, nodes=None, header=True)


Write graph G to given path in sparse6 format.
Parameters
• G (Graph (undirected))
• path (file or string) – File or filename to write
• nodes (list or iterable) – Nodes are labeled 0. . . n-1 in the order provided. If None the
ordering given by G.nodes() is used.
• header (bool) – If True add ‘>>sparse6<<’ string to head of data
Raises NetworkXError – If the graph is directed

Examples

You can write a sparse6 file by giving the path to the file:

656 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> import tempfile


>>> with tempfile.NamedTemporaryFile() as f:
... nx.write_sparse6(nx.path_graph(2), f.name)
... print(f.read())
b'>>sparse6<<:An\n'

You can also write a sparse6 file by giving an open file-like object:

>>> with tempfile.NamedTemporaryFile() as f:


... nx.write_sparse6(nx.path_graph(2), f)
... _ = f.seek(0)
... print(f.read())
b'>>sparse6<<:An\n'

See also:
read_sparse6(), from_sparse6_bytes()

Notes

The format does not support edge or node labels.

References

9.12 Pajek

9.12.1 Pajek

Read graphs in Pajek format.


This implementation handles directed and undirected graphs including those with self loops and parallel edges.

Format

See http://vlado.fmf.uni-lj.si/pub/networks/pajek/doc/draweps.htm for format information.

read_pajek(path[, encoding]) Read graph in Pajek format from path.


write_pajek(G, path[, encoding]) Write graph in Pajek format to path.
parse_pajek(lines) Parse Pajek format graph from string or iterable.
generate_pajek(G) Generate lines in Pajek graph format.

9.12.2 networkx.readwrite.pajek.read_pajek

read_pajek(path, encoding=’UTF-8’)
Read graph in Pajek format from path.
Parameters path (file or string) – File or filename to write. Filenames ending in .gz or .bz2 will be
uncompressed.
Returns G
Return type NetworkX MultiGraph or MultiDiGraph.

9.12. Pajek 657


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G=nx.path_graph(4)
>>> nx.write_pajek(G, "test.net")
>>> G=nx.read_pajek("test.net")

To create a Graph instead of a MultiGraph use

>>> G1=nx.Graph(G)

References

See http://vlado.fmf.uni-lj.si/pub/networks/pajek/doc/draweps.htm for format information.

9.12.3 networkx.readwrite.pajek.write_pajek

write_pajek(G, path, encoding=’UTF-8’)


Write graph in Pajek format to path.
Parameters
• G (graph) – A Networkx graph
• path (file or string) – File or filename to write. Filenames ending in .gz or .bz2 will be
compressed.

Examples

>>> G=nx.path_graph(4)
>>> nx.write_pajek(G, "test.net")

Warning: Optional node attributes and edge attributes must be non-empty strings. Otherwise it will not be
written into the file. You will need to convert those attributes to strings if you want to keep them.

References

See http://vlado.fmf.uni-lj.si/pub/networks/pajek/doc/draweps.htm for format information.

9.12.4 networkx.readwrite.pajek.parse_pajek

parse_pajek(lines)
Parse Pajek format graph from string or iterable.
Parameters lines (string or iterable) – Data in Pajek format.
Returns G
Return type NetworkX graph
See also:
read_pajek()

658 Chapter 9. Reading and writing graphs


NetworkX Reference, Release 2.3rc1.dev20190222214247

9.12.5 networkx.readwrite.pajek.generate_pajek

generate_pajek(G)
Generate lines in Pajek graph format.
Parameters G (graph) – A Networkx graph

References

See http://vlado.fmf.uni-lj.si/pub/networks/pajek/doc/draweps.htm for format information.

9.13 GIS Shapefile

9.13.1 Shapefile

Generates a networkx.DiGraph from point and line shapefiles.


“The Esri Shapefile or simply a shapefile is a popular geospatial vector data format for geographic information systems
software. It is developed and regulated by Esri as a (mostly) open specification for data interoperability among Esri
and other software products.” See https://en.wikipedia.org/wiki/Shapefile for additional information.

read_shp(path[, simplify, geom_attrs, strict]) Generates a networkx.DiGraph from shapefiles.


write_shp(G, outdir) Writes a networkx.DiGraph to two shapefiles, edges and
nodes.

9.13.2 networkx.readwrite.nx_shp.read_shp

read_shp(path, simplify=True, geom_attrs=True, strict=True)


Generates a networkx.DiGraph from shapefiles. Point geometries are translated into nodes, lines into edges.
Coordinate tuples are used as keys. Attributes are preserved, line geometries are simplified into start and end
coordinates. Accepts a single shapefile or directory of many shapefiles.
“The Esri Shapefile or simply a shapefile is a popular geospatial vector data format for geographic information
systems software1 .”
Parameters
• path (file or string) – File, directory, or filename to read.
• simplify (bool) – If True, simplify line geometries to start and end coordinates. If False, and
line feature geometry has multiple segments, the non-geometric attributes for that feature
will be repeated for each edge comprising that feature.
• geom_attrs (bool) – If True, include the Wkb, Wkt and Json geometry attributes with each
edge.
NOTE: if these attributes are available, write_shp will use them to write the geometry. If
nodes store the underlying coordinates for the edge geometry as well (as they do when they
are read via this method) and they change, your geomety will be out of sync.
• strict (bool) – If True, raise NetworkXError when feature geometry is missing or Geom-
etryType is not supported. If False, silently ignore missing or unsupported geometry in
features.
1 https://en.wikipedia.org/wiki/Shapefile

9.13. GIS Shapefile 659


NetworkX Reference, Release 2.3rc1.dev20190222214247

Returns G
Return type NetworkX graph
Raises
• ImportError – If ogr module is not available.
• RuntimeError – If file cannot be open or read.
• NetworkXError – If strict=True and feature is missing geometry or GeometryType is not
supported.

Examples

>>> G=nx.read_shp('test.shp') # doctest: +SKIP

References

9.13.3 networkx.readwrite.nx_shp.write_shp

write_shp(G, outdir)
Writes a networkx.DiGraph to two shapefiles, edges and nodes. Nodes and edges are expected to have a Well
Known Binary (Wkb) or Well Known Text (Wkt) key in order to generate geometries. Also acceptable are nodes
with a numeric tuple key (x,y).
“The Esri Shapefile or simply a shapefile is a popular geospatial vector data format for geographic information
systems software1 .”
Parameters outdir (directory path) – Output directory for the two shapefiles.
Returns
Return type None

Examples

nx.write_shp(digraph, ‘/shapefiles’) # doctest +SKIP

References

1 https://en.wikipedia.org/wiki/Shapefile

660 Chapter 9. Reading and writing graphs


CHAPTER

TEN

DRAWING

NetworkX provides basic functionality for visualizing graphs, but its main goal is to enable graph analysis rather than
perform graph visualization. In the future, graph visualization functionality may be removed from NetworkX or only
available as an add-on package.
Proper graph visualization is hard, and we highly recommend that people visualize their graphs with tools dedicated to
that task. Notable examples of dedicated and fully-featured graph visualization tools are Cytoscape, Gephi, Graphviz
and, for LaTeX typesetting, PGF/TikZ. To use these and other such tools, you should export your NetworkX graph into
a format that can be read by those tools. For example, Cytoscape can read the GraphML format, and so, networkx.
write_graphml(G) might be an appropriate choice.

10.1 Matplotlib

10.1.1 Matplotlib

Draw networks with matplotlib.


See also:
matplotlib http://matplotlib.org/
pygraphviz http://pygraphviz.github.io/

draw(G[, pos, ax]) Draw the graph G with Matplotlib.


draw_networkx(G[, pos, arrows, with_labels]) Draw the graph G using Matplotlib.
draw_networkx_nodes(G, pos[, nodelist, . . . ]) Draw the nodes of the graph G.
draw_networkx_edges(G, pos[, edgelist, . . . ]) Draw the edges of the graph G.
draw_networkx_labels(G, pos[, labels, . . . ]) Draw node labels on the graph G.
draw_networkx_edge_labels(G, pos[, . . . ]) Draw edge labels.
draw_circular(G, **kwargs) Draw the graph G with a circular layout.
draw_kamada_kawai(G, **kwargs) Draw the graph G with a Kamada-Kawai force-directed
layout.
draw_random(G, **kwargs) Draw the graph G with a random layout.
draw_spectral(G, **kwargs) Draw the graph G with a spectral 2D layout.
draw_spring(G, **kwargs) Draw the graph G with a spring layout.
draw_shell(G, **kwargs) Draw networkx graph with shell layout.

661
NetworkX Reference, Release 2.3rc1.dev20190222214247

10.1.2 networkx.drawing.nx_pylab.draw

draw(G, pos=None, ax=None, **kwds)


Draw the graph G with Matplotlib.
Draw the graph as a simple representation with no node labels or edge labels and using the full Matplotlib figure
area and no axis labels by default. See draw_networkx() for more full-featured drawing that allows title, axis
labels etc.
Parameters
• G (graph) – A networkx graph
• pos (dictionary, optional) – A dictionary with nodes as keys and positions as values. If
not specified a spring layout positioning will be computed. See networkx.drawing.
layout for functions that compute node positions.
• ax (Matplotlib Axes object, optional) – Draw the graph in specified Matplotlib axes.
• kwds (optional keywords) – See networkx.draw_networkx() for a description of optional
keywords.

Examples

>>> G = nx.dodecahedral_graph()
>>> nx.draw(G)
>>> nx.draw(G, pos=nx.spring_layout(G)) # use spring layout

See also:
draw_networkx(), draw_networkx_nodes(), draw_networkx_edges(),
draw_networkx_labels(), draw_networkx_edge_labels()

Notes

This function has the same name as pylab.draw and pyplot.draw so beware when using

>>> from networkx import *

since you might overwrite the pylab.draw function.


With pyplot use

>>> import matplotlib.pyplot as plt


>>> import networkx as nx
>>> G = nx.dodecahedral_graph()
>>> nx.draw(G) # networkx draw()
>>> plt.draw() # pyplot draw()

Also see the NetworkX drawing examples at https://networkx.github.io/documentation/latest/auto_examples/


index.html

10.1.3 networkx.drawing.nx_pylab.draw_networkx

draw_networkx(G, pos=None, arrows=True, with_labels=True, **kwds)


Draw the graph G using Matplotlib.

662 Chapter 10. Drawing


NetworkX Reference, Release 2.3rc1.dev20190222214247

Draw the graph with Matplotlib with options for node positions, labeling, titles, and many other drawing fea-
tures. See draw() for simple drawing without labels or axes.
Parameters
• G (graph) – A networkx graph
• pos (dictionary, optional) – A dictionary with nodes as keys and positions as values. If
not specified a spring layout positioning will be computed. See networkx.drawing.
layout for functions that compute node positions.
• arrows (bool, optional (default=True)) – For directed graphs, if True draw arrowheads.
Note: Arrows will be the same color as edges.
• arrowstyle (str, optional (default=’-|>’)) – For directed graphs, choose the style of the
arrowsheads. See :py:class: matplotlib.patches.ArrowStyle for more options.
• arrowsize (int, optional (default=10)) – For directed graphs, choose the size of the
arrow head head’s length and width. See :py:class: matplotlib.patches.
FancyArrowPatch for attribute mutation_scale for more info.
• with_labels (bool, optional (default=True)) – Set to True to draw labels on the nodes.
• ax (Matplotlib Axes object, optional) – Draw the graph in the specified Matplotlib axes.
• nodelist (list, optional (default G.nodes())) – Draw only specified nodes
• edgelist (list, optional (default=G.edges())) – Draw only specified edges
• node_size (scalar or array, optional (default=300)) – Size of nodes. If an array is specified
it must be the same length as nodelist.
• node_color (color string, or array of floats, (default=’#1f78b4’)) – Node color. Can be
a single color format string, or a sequence of colors with the same length as nodelist. If
numeric values are specified they will be mapped to colors using the cmap and vmin,vmax
parameters. See matplotlib.scatter for more details.
• node_shape (string, optional (default=’o’)) – The shape of the node. Specification is as
matplotlib.scatter marker, one of ‘so^>v<dph8’.
• alpha (float, optional (default=1.0)) – The node and edge transparency
• cmap (Matplotlib colormap, optional (default=None)) – Colormap for mapping intensities
of nodes
• vmin,vmax (float, optional (default=None)) – Minimum and maximum for node colormap
scaling
• linewidths ([None | scalar | sequence]) – Line width of symbol border (default =1.0)
• width (float, optional (default=1.0)) – Line width of edges
• edge_color (color string, or array of floats (default=’r’)) – Edge color. Can be a single color
format string, or a sequence of colors with the same length as edgelist. If numeric values are
specified they will be mapped to colors using the edge_cmap and edge_vmin,edge_vmax
parameters.
• edge_cmap (Matplotlib colormap, optional (default=None)) – Colormap for mapping in-
tensities of edges
• edge_vmin,edge_vmax (floats, optional (default=None)) – Minimum and maximum for
edge colormap scaling
• style (string, optional (default=’solid’)) – Edge line style (solid|dashed|dotted,dashdot)

10.1. Matplotlib 663


NetworkX Reference, Release 2.3rc1.dev20190222214247

• labels (dictionary, optional (default=None)) – Node labels in a dictionary keyed by node of


text labels
• font_size (int, optional (default=12)) – Font size for text labels
• font_color (string, optional (default=’k’ black)) – Font color string
• font_weight (string, optional (default=’normal’)) – Font weight
• font_family (string, optional (default=’sans-serif’)) – Font family
• label (string, optional) – Label for graph legend

Notes

For directed graphs, arrows are drawn at the head end. Arrows can be turned off with keyword arrows=False.

Examples

>>> G = nx.dodecahedral_graph()
>>> nx.draw(G)
>>> nx.draw(G, pos=nx.spring_layout(G)) # use spring layout

>>> import matplotlib.pyplot as plt


>>> limits = plt.axis('off') # turn of axis

Also see the NetworkX drawing examples at https://networkx.github.io/documentation/latest/auto_examples/


index.html
See also:
draw(), draw_networkx_nodes(), draw_networkx_edges(), draw_networkx_labels(),
draw_networkx_edge_labels()

10.1.4 networkx.drawing.nx_pylab.draw_networkx_nodes

draw_networkx_nodes(G, pos, nodelist=None, node_size=300, node_color=’#1f78b4’, node_shape=’o’,


alpha=1.0, cmap=None, vmin=None, vmax=None, ax=None, linewidths=None,
edgecolors=None, label=None, **kwds)
Draw the nodes of the graph G.
This draws only the nodes of the graph G.
Parameters
• G (graph) – A networkx graph
• pos (dictionary) – A dictionary with nodes as keys and positions as values. Positions should
be sequences of length 2.
• ax (Matplotlib Axes object, optional) – Draw the graph in the specified Matplotlib axes.
• nodelist (list, optional) – Draw only specified nodes (default G.nodes())
• node_size (scalar or array) – Size of nodes (default=300). If an array is specified it must
be the same length as nodelist.

664 Chapter 10. Drawing


NetworkX Reference, Release 2.3rc1.dev20190222214247

• node_color (color string, or array of floats) – Node color. Can be a single color format
string (default=’#1f78b4’), or a sequence of colors with the same length as nodelist. If
numeric values are specified they will be mapped to colors using the cmap and vmin,vmax
parameters. See matplotlib.scatter for more details.
• node_shape (string) – The shape of the node. Specification is as matplotlib.scatter marker,
one of ‘so^>v<dph8’ (default=’o’).
• alpha (float or array of floats) – The node transparency. This can be a single alpha value
(default=1.0), in which case it will be applied to all the nodes of color. Otherwise, if it is
an array, the elements of alpha will be applied to the colors in order (cycling through alpha
multiple times if necessary).
• cmap (Matplotlib colormap) – Colormap for mapping intensities of nodes (default=None)
• vmin,vmax (floats) – Minimum and maximum for node colormap scaling (default=None)
• linewidths ([None | scalar | sequence]) – Line width of symbol border (default =1.0)
• edgecolors ([None | scalar | sequence]) – Colors of node borders (default = node_color)
• label ([None| string]) – Label for legend
Returns PathCollection of the nodes.
Return type matplotlib.collections.PathCollection

Examples

>>> G = nx.dodecahedral_graph()
>>> nodes = nx.draw_networkx_nodes(G, pos=nx.spring_layout(G))

Also see the NetworkX drawing examples at https://networkx.github.io/documentation/latest/auto_examples/


index.html
See also:
draw(), draw_networkx(), draw_networkx_edges(), draw_networkx_labels(),
draw_networkx_edge_labels()

10.1.5 networkx.drawing.nx_pylab.draw_networkx_edges

draw_networkx_edges(G, pos, edgelist=None, width=1.0, edge_color=’k’, style=’solid’, alpha=1.0,


arrowstyle=’-|>’, arrowsize=10, edge_cmap=None, edge_vmin=None,
edge_vmax=None, ax=None, arrows=True, label=None, node_size=300,
nodelist=None, node_shape=’o’, **kwds)
Draw the edges of the graph G.
This draws only the edges of the graph G.
Parameters
• G (graph) – A networkx graph
• pos (dictionary) – A dictionary with nodes as keys and positions as values. Positions should
be sequences of length 2.
• edgelist (collection of edge tuples) – Draw only specified edges(default=G.edges())
• width (float, or array of floats) – Line width of edges (default=1.0)

10.1. Matplotlib 665


NetworkX Reference, Release 2.3rc1.dev20190222214247

• edge_color (color string, or array of floats) – Edge color. Can be a single color format string
(default=’r’), or a sequence of colors with the same length as edgelist. If numeric values are
specified they will be mapped to colors using the edge_cmap and edge_vmin,edge_vmax
parameters.
• style (string) – Edge line style (default=’solid’) (solid|dashed|dotted,dashdot)
• alpha (float) – The edge transparency (default=1.0)
• edge_ cmap (Matplotlib colormap) – Colormap for mapping intensities of edges (de-
fault=None)
• edge_vmin,edge_vmax (floats) – Minimum and maximum for edge colormap scaling (de-
fault=None)
• ax (Matplotlib Axes object, optional) – Draw the graph in the specified Matplotlib axes.
• arrows (bool, optional (default=True)) – For directed graphs, if True draw arrowheads.
Note: Arrows will be the same color as edges.
• arrowstyle (str, optional (default=’-|>’)) – For directed graphs, choose the style of the
arrow heads. See :py:class: matplotlib.patches.ArrowStyle for more options.
• arrowsize (int, optional (default=10)) – For directed graphs, choose the size of the
arrow head head’s length and width. See :py:class: matplotlib.patches.
FancyArrowPatch for attribute mutation_scale for more info.
• label ([None| string]) – Label for legend
Returns
• matplotlib.collection.LineCollection – LineCollection of the edges
• list of matplotlib.patches.FancyArrowPatch – FancyArrowPatch instances of the di-
rected edges
• Depending whether the drawing includes arrows or not.

Notes

For directed graphs, arrows are drawn at the head end. Arrows can be turned off with keyword arrows=False.
Be sure to include node_size as a keyword argument; arrows are drawn considering the size of nodes.

Examples

>>> G = nx.dodecahedral_graph()
>>> edges = nx.draw_networkx_edges(G, pos=nx.spring_layout(G))

>>> G = nx.DiGraph()
>>> G.add_edges_from([(1, 2), (1, 3), (2, 3)])
>>> arcs = nx.draw_networkx_edges(G, pos=nx.spring_layout(G))
>>> alphas = [0.3, 0.4, 0.5]
>>> for i, arc in enumerate(arcs): # change alpha values of arcs
... arc.set_alpha(alphas[i])

Also see the NetworkX drawing examples at https://networkx.github.io/documentation/latest/auto_examples/


index.html
See also:

666 Chapter 10. Drawing


NetworkX Reference, Release 2.3rc1.dev20190222214247

draw(), draw_networkx(), draw_networkx_nodes(), draw_networkx_labels(),


draw_networkx_edge_labels()

10.1.6 networkx.drawing.nx_pylab.draw_networkx_labels

draw_networkx_labels(G, pos, labels=None, font_size=12, font_color=’k’, font_family=’sans-serif’,


font_weight=’normal’, alpha=1.0, bbox=None, ax=None, **kwds)
Draw node labels on the graph G.
Parameters
• G (graph) – A networkx graph
• pos (dictionary) – A dictionary with nodes as keys and positions as values. Positions should
be sequences of length 2.
• labels (dictionary, optional (default=None)) – Node labels in a dictionary keyed by node of
text labels Node-keys in labels should appear as keys in pos. If needed use: {n:lab for
n,lab in labels.items() if n in pos}
• font_size (int) – Font size for text labels (default=12)
• font_color (string) – Font color string (default=’k’ black)
• font_family (string) – Font family (default=’sans-serif’)
• font_weight (string) – Font weight (default=’normal’)
• alpha (float) – The text transparency (default=1.0)
• ax (Matplotlib Axes object, optional) – Draw the graph in the specified Matplotlib axes.
Returns dict of labels keyed on the nodes
Return type dict

Examples

>>> G = nx.dodecahedral_graph()
>>> labels = nx.draw_networkx_labels(G, pos=nx.spring_layout(G))

Also see the NetworkX drawing examples at https://networkx.github.io/documentation/latest/auto_examples/


index.html
See also:
draw(), draw_networkx(), draw_networkx_nodes(), draw_networkx_edges(),
draw_networkx_edge_labels()

10.1.7 networkx.drawing.nx_pylab.draw_networkx_edge_labels

draw_networkx_edge_labels(G, pos, edge_labels=None, label_pos=0.5, font_size=10,


font_color=’k’, font_family=’sans-serif’, font_weight=’normal’,
alpha=1.0, bbox=None, ax=None, rotate=True, **kwds)
Draw edge labels.
Parameters
• G (graph) – A networkx graph

10.1. Matplotlib 667


NetworkX Reference, Release 2.3rc1.dev20190222214247

• pos (dictionary) – A dictionary with nodes as keys and positions as values. Positions should
be sequences of length 2.
• ax (Matplotlib Axes object, optional) – Draw the graph in the specified Matplotlib axes.
• alpha (float) – The text transparency (default=1.0)
• edge_labels (dictionary) – Edge labels in a dictionary keyed by edge two-tuple of text labels
(default=None). Only labels for the keys in the dictionary are drawn.
• label_pos (float) – Position of edge label along edge (0=head, 0.5=center, 1=tail)
• font_size (int) – Font size for text labels (default=12)
• font_color (string) – Font color string (default=’k’ black)
• font_weight (string) – Font weight (default=’normal’)
• font_family (string) – Font family (default=’sans-serif’)
• bbox (Matplotlib bbox) – Specify text box shape and colors.
• clip_on (bool) – Turn on clipping at axis boundaries (default=True)
Returns dict of labels keyed on the edges
Return type dict

Examples

>>> G = nx.dodecahedral_graph()
>>> edge_labels = nx.draw_networkx_edge_labels(G, pos=nx.spring_layout(G))

Also see the NetworkX drawing examples at https://networkx.github.io/documentation/latest/auto_examples/


index.html
See also:
draw(), draw_networkx(), draw_networkx_nodes(), draw_networkx_edges(),
draw_networkx_labels()

10.1.8 networkx.drawing.nx_pylab.draw_circular

draw_circular(G, **kwargs)
Draw the graph G with a circular layout.
Parameters
• G (graph) – A networkx graph
• kwargs (optional keywords) – See networkx.draw_networkx() for a description of optional
keywords, with the exception of the pos parameter which is not used by this function.

10.1.9 networkx.drawing.nx_pylab.draw_kamada_kawai

draw_kamada_kawai(G, **kwargs)
Draw the graph G with a Kamada-Kawai force-directed layout.
Parameters
• G (graph) – A networkx graph

668 Chapter 10. Drawing


NetworkX Reference, Release 2.3rc1.dev20190222214247

• kwargs (optional keywords) – See networkx.draw_networkx() for a description of optional


keywords, with the exception of the pos parameter which is not used by this function.

10.1.10 networkx.drawing.nx_pylab.draw_random

draw_random(G, **kwargs)
Draw the graph G with a random layout.
Parameters
• G (graph) – A networkx graph
• kwargs (optional keywords) – See networkx.draw_networkx() for a description of optional
keywords, with the exception of the pos parameter which is not used by this function.

10.1.11 networkx.drawing.nx_pylab.draw_spectral

draw_spectral(G, **kwargs)
Draw the graph G with a spectral 2D layout.
Using the unnormalized Laplacion, the layout shows possible clusters of nodes which are an approximation of
the ratio cut. The positions are the entries of the second and third eigenvectors corresponding to the ascending
eigenvalues starting from the second one.
Parameters
• G (graph) – A networkx graph
• kwargs (optional keywords) – See networkx.draw_networkx() for a description of optional
keywords, with the exception of the pos parameter which is not used by this function.

10.1.12 networkx.drawing.nx_pylab.draw_spring

draw_spring(G, **kwargs)
Draw the graph G with a spring layout.
Parameters
• G (graph) – A networkx graph
• kwargs (optional keywords) – See networkx.draw_networkx() for a description of optional
keywords, with the exception of the pos parameter which is not used by this function.

10.1.13 networkx.drawing.nx_pylab.draw_shell

draw_shell(G, **kwargs)
Draw networkx graph with shell layout.
Parameters
• G (graph) – A networkx graph
• kwargs (optional keywords) – See networkx.draw_networkx() for a description of optional
keywords, with the exception of the pos parameter which is not used by this function.

10.1. Matplotlib 669


NetworkX Reference, Release 2.3rc1.dev20190222214247

10.2 Graphviz AGraph (dot)

10.2.1 Graphviz AGraph

Interface to pygraphviz AGraph class.

Examples

>>> G = nx.complete_graph(5)
>>> A = nx.nx_agraph.to_agraph(G)
>>> H = nx.nx_agraph.from_agraph(A)

See also:
Pygraphviz http://pygraphviz.github.io/

from_agraph(A[, create_using]) Returns a NetworkX Graph or DiGraph from a Py-


Graphviz graph.
to_agraph(N) Returns a pygraphviz graph from a NetworkX graph N.
write_dot(G, path) Write NetworkX graph G to Graphviz dot format on
path.
read_dot(path) Returns a NetworkX graph from a dot file on path.
graphviz_layout(G[, prog, root, args]) Create node positions for G using Graphviz.
pygraphviz_layout(G[, prog, root, args]) Create node positions for G using Graphviz.

10.2.2 networkx.drawing.nx_agraph.from_agraph

from_agraph(A, create_using=None)
Returns a NetworkX Graph or DiGraph from a PyGraphviz graph.
Parameters
• A (PyGraphviz AGraph) – A graph created with PyGraphviz
• create_using (NetworkX graph constructor, optional (default=nx.Graph)) – Graph type to
create. If graph instance, then cleared before populated.

Examples

>>> K5 = nx.complete_graph(5)
>>> A = nx.nx_agraph.to_agraph(K5)
>>> G = nx.nx_agraph.from_agraph(A)

Notes

The Graph G will have a dictionary G.graph_attr containing the default graphviz attributes for graphs, nodes
and edges.
Default node attributes will be in the dictionary G.node_attr which is keyed by node.
Edge attributes will be returned as edge data in G. With edge_attr=False the edge data will be the Graphviz edge
weight attribute or the value 1 if no edge weight attribute is found.

670 Chapter 10. Drawing


NetworkX Reference, Release 2.3rc1.dev20190222214247

10.2.3 networkx.drawing.nx_agraph.to_agraph

to_agraph(N)
Returns a pygraphviz graph from a NetworkX graph N.
Parameters N (NetworkX graph) – A graph created with NetworkX

Examples

>>> K5 = nx.complete_graph(5)
>>> A = nx.nx_agraph.to_agraph(K5)

Notes

If N has an dict N.graph_attr an attempt will be made first to copy properties attached to the graph (see
from_agraph) and then updated with the calling arguments if any.

10.2.4 networkx.drawing.nx_agraph.write_dot

write_dot(G, path)
Write NetworkX graph G to Graphviz dot format on path.
Parameters
• G (graph) – A networkx graph
• path (filename) – Filename or file handle to write

10.2.5 networkx.drawing.nx_agraph.read_dot

read_dot(path)
Returns a NetworkX graph from a dot file on path.
Parameters path (file or string) – File name or file handle to read.

10.2.6 networkx.drawing.nx_agraph.graphviz_layout

graphviz_layout(G, prog=’neato’, root=None, args=”)


Create node positions for G using Graphviz.
Parameters
• G (NetworkX graph) – A graph created with NetworkX
• prog (string) – Name of Graphviz layout program
• root (string, optional) – Root node for twopi layout
• args (string, optional) – Extra arguments to Graphviz layout program
• Returns (dictionary) – Dictionary of x, y, positions keyed by node.

10.2. Graphviz AGraph (dot) 671


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.petersen_graph()
>>> pos = nx.nx_agraph.graphviz_layout(G)
>>> pos = nx.nx_agraph.graphviz_layout(G, prog='dot')

Notes

This is a wrapper for pygraphviz_layout.

10.2.7 networkx.drawing.nx_agraph.pygraphviz_layout

pygraphviz_layout(G, prog=’neato’, root=None, args=”)


Create node positions for G using Graphviz.
Parameters
• G (NetworkX graph) – A graph created with NetworkX
• prog (string) – Name of Graphviz layout program
• root (string, optional) – Root node for twopi layout
• args (string, optional) – Extra arguments to Graphviz layout program
• Returns (dictionary) – Dictionary of x, y, positions keyed by node.

Examples

>>> G = nx.petersen_graph()
>>> pos = nx.nx_agraph.graphviz_layout(G)
>>> pos = nx.nx_agraph.graphviz_layout(G, prog='dot')

Notes

If you use complex node objects, they may have the same string representation and GraphViz could treat them
as the same node. The layout may assign both nodes a single location. See Issue #1568 If this occurs in your
case, consider relabeling the nodes just for the layout computation using something similar to:
H = nx.convert_node_labels_to_integers(G, label_attribute=’node_label’) H_layout =
nx.nx_agraph.pygraphviz_layout(G, prog=’dot’) G_layout = {H.nodes[n][‘node_label’]: p for
n, p in H_layout.items()}

10.3 Graphviz with pydot

10.3.1 Pydot

Import and export NetworkX graphs in Graphviz dot format using pydot.
Either this module or nx_agraph can be used to interface with graphviz.
See also:

672 Chapter 10. Drawing


NetworkX Reference, Release 2.3rc1.dev20190222214247

pydot https://github.com/erocarrera/pydot
Graphviz https://www.graphviz.org
DOT

from_pydot(P) Returns a NetworkX graph from a Pydot graph.


to_pydot(N) Returns a pydot graph from a NetworkX graph N.
write_dot(G, path) Write NetworkX graph G to Graphviz dot format on
path.
read_dot(path) Returns a NetworkX MultiGraph or
MultiDiGraph from the dot file with the passed
path.
graphviz_layout(G[, prog, root]) Create node positions using Pydot and Graphviz.
pydot_layout(G[, prog, root]) Create node positions using pydot and Graphviz.

10.3.2 networkx.drawing.nx_pydot.from_pydot

from_pydot(P)
Returns a NetworkX graph from a Pydot graph.
Parameters P (Pydot graph) – A graph created with Pydot
Returns G – A MultiGraph or MultiDiGraph.
Return type NetworkX multigraph

Examples

>>> K5 = nx.complete_graph(5)
>>> A = nx.nx_pydot.to_pydot(K5)
>>> G = nx.nx_pydot.from_pydot(A) # return MultiGraph

# make a Graph instead of MultiGraph >>> G = nx.Graph(nx.nx_pydot.from_pydot(A))

10.3.3 networkx.drawing.nx_pydot.to_pydot

to_pydot(N)
Returns a pydot graph from a NetworkX graph N.
Parameters N (NetworkX graph) – A graph created with NetworkX

Examples

>>> K5 = nx.complete_graph(5)
>>> P = nx.nx_pydot.to_pydot(K5)

10.3. Graphviz with pydot 673


NetworkX Reference, Release 2.3rc1.dev20190222214247

Notes

10.3.4 networkx.drawing.nx_pydot.write_dot

write_dot(G, path)
Write NetworkX graph G to Graphviz dot format on path.
Path can be a string or a file handle.

10.3.5 networkx.drawing.nx_pydot.read_dot

read_dot(path)
Returns a NetworkX MultiGraph or MultiDiGraph from the dot file with the passed path.
If this file contains multiple graphs, only the first such graph is returned. All graphs _except_ the first are silently
ignored.
Parameters path (str or file) – Filename or file handle.
Returns G – A MultiGraph or MultiDiGraph.
Return type MultiGraph or MultiDiGraph

Notes

Use G = nx.Graph(read_dot(path)) to return a Graph instead of a MultiGraph.

10.3.6 networkx.drawing.nx_pydot.graphviz_layout

graphviz_layout(G, prog=’neato’, root=None, **kwds)


Create node positions using Pydot and Graphviz.
Returns a dictionary of positions keyed by node.

Examples

>>> G = nx.complete_graph(4)
>>> pos = nx.nx_pydot.graphviz_layout(G)
>>> pos = nx.nx_pydot.graphviz_layout(G, prog='dot')

Notes

This is a wrapper for pydot_layout.

10.3.7 networkx.drawing.nx_pydot.pydot_layout

pydot_layout(G, prog=’neato’, root=None, **kwds)


Create node positions using pydot and Graphviz.
Parameters
• G (Graph) – NetworkX graph to be laid out.

674 Chapter 10. Drawing


NetworkX Reference, Release 2.3rc1.dev20190222214247

• prog (optional[str]) – Basename of the GraphViz command with which to layout this graph.
Defaults to neato: default GraphViz command for undirected graphs.
Returns Dictionary of positions keyed by node.
Return type dict

Examples

>>> G = nx.complete_graph(4)
>>> pos = nx.nx_pydot.pydot_layout(G)
>>> pos = nx.nx_pydot.pydot_layout(G, prog='dot')

Notes

If you use complex node objects, they may have the same string representation and GraphViz could treat them
as the same node. The layout may assign both nodes a single location. See Issue #1568 If this occurs in your
case, consider relabeling the nodes just for the layout computation using something similar to:
H = nx.convert_node_labels_to_integers(G, label_attribute=’node_label’) H_layout =
nx.nx_pydot.pydot_layout(G, prog=’dot’) G_layout = {H.nodes[n][‘node_label’]: p for n, p in
H_layout.items()}

10.4 Graph Layout

10.4.1 Layout

Node positioning algorithms for graph drawing.


For random_layout() the possible resulting shape is a square of side [0, scale] (default: [0, 1]) Changing center
shifts the layout by that amount.
For the other layout routines, the extent is [center - scale, center + scale] (default: [-1, 1]).
Warning: Most layout routines have only been tested in 2-dimensions.

bipartite_layout(G, nodes[, align, scale, . . . ]) Position nodes in two straight lines.


circular_layout(G[, scale, center, dim]) Position nodes on a circle.
kamada_kawai_layout(G[, dist, pos, weight, . . . ]) Position nodes using Kamada-Kawai path-length cost-
function.
random_layout(G[, center, dim, seed]) Position nodes uniformly at random in the unit square.
rescale_layout(pos[, scale]) Returns scaled position array to (-scale, scale) in all
axes.
shell_layout(G[, nlist, scale, center, dim]) Position nodes in concentric circles.
spring_layout(G[, k, pos, fixed, . . . ]) Position nodes using Fruchterman-Reingold force-
directed algorithm.
spectral_layout(G[, weight, scale, center, dim]) Position nodes using the eigenvectors of the graph
Laplacian.

10.4. Graph Layout 675


NetworkX Reference, Release 2.3rc1.dev20190222214247

10.4.2 networkx.drawing.layout.bipartite_layout

bipartite_layout(G, nodes, align=’vertical’, scale=1, center=None, as-


pect_ratio=1.3333333333333333)
Position nodes in two straight lines.
Parameters
• G (NetworkX graph or list of nodes) – A position will be assigned to every node in G.
• nodes (list or container) – Nodes in one node set of the bipartite graph. This set will be
placed on left or top.
• align (string (default=’vertical’)) – The alignment of nodes. Vertical or horizontal.
• scale (number (default: 1)) – Scale factor for positions.
• center (array-like or None) – Coordinate pair around which to center the layout.
• aspect_ratio (number (default=4/3):) – The ratio of the width to the height of the layout.
Returns pos – A dictionary of positions keyed by node.
Return type dict

Examples

>>> G = nx.bipartite.gnmk_random_graph(3, 5, 10, seed=123)


>>> top = nx.bipartite.sets(G)[0]
>>> pos = nx.bipartite_layout(G, top)

Notes

This algorithm currently only works in two dimensions and does not try to minimize edge crossings.

10.4.3 networkx.drawing.layout.circular_layout

circular_layout(G, scale=1, center=None, dim=2)


Position nodes on a circle.
Parameters
• G (NetworkX graph or list of nodes) – A position will be assigned to every node in G.
• scale (number (default: 1)) – Scale factor for positions.
• center (array-like or None) – Coordinate pair around which to center the layout.
• dim (int) – Dimension of layout. If dim>2, the remaining dimensions are set to zero in the
returned positions. If dim<2, a ValueError is raised.
Returns pos – A dictionary of positions keyed by node
Return type dict
Raises ValueError – If dim < 2

676 Chapter 10. Drawing


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(4)
>>> pos = nx.circular_layout(G)

Notes

This algorithm currently only works in two dimensions and does not try to minimize edge crossings.

10.4.4 networkx.drawing.layout.kamada_kawai_layout

kamada_kawai_layout(G, dist=None, pos=None, weight=’weight’, scale=1, center=None, dim=2)


Position nodes using Kamada-Kawai path-length cost-function.
Parameters
• G (NetworkX graph or list of nodes) – A position will be assigned to every node in G.
• dist (float (default=None)) – A two-level dictionary of optimal distances between nodes,
indexed by source and destination node. If None, the distance is computed using short-
est_path_length().
• pos (dict or None optional (default=None)) – Initial positions for nodes as a dictionary with
node as keys and values as a coordinate list or tuple. If None, then use circular_layout() for
dim >= 2 and a linear layout for dim == 1.
• weight (string or None optional (default=’weight’)) – The edge attribute that holds the
numerical value used for the edge weight. If None, then all edge weights are 1.
• scale (number (default: 1)) – Scale factor for positions.
• center (array-like or None) – Coordinate pair around which to center the layout.
• dim (int) – Dimension of layout.
Returns pos – A dictionary of positions keyed by node
Return type dict

Examples

>>> G = nx.path_graph(4)
>>> pos = nx.kamada_kawai_layout(G)

10.4.5 networkx.drawing.layout.random_layout

random_layout(G, center=None, dim=2, seed=None)


Position nodes uniformly at random in the unit square.
For every node, a position is generated by choosing each of dim coordinates uniformly at random on the interval
[0.0, 1.0).
NumPy (http://scipy.org) is required for this function.
Parameters
• G (NetworkX graph or list of nodes) – A position will be assigned to every node in G.

10.4. Graph Layout 677


NetworkX Reference, Release 2.3rc1.dev20190222214247

• center (array-like or None) – Coordinate pair around which to center the layout.
• dim (int) – Dimension of layout.
• seed (int, RandomState instance or None optional (default=None)) – Set the random state
for deterministic node layouts. If int, seed is the seed used by the random number gen-
erator, if numpy.random.RandomState instance, seed is the random number generator, if
None, the random number generator is the RandomState instance used by numpy.random.
Returns pos – A dictionary of positions keyed by node
Return type dict

Examples

>>> G = nx.lollipop_graph(4, 3)
>>> pos = nx.random_layout(G)

10.4.6 networkx.drawing.layout.rescale_layout

rescale_layout(pos, scale=1)
Returns scaled position array to (-scale, scale) in all axes.
The function acts on NumPy arrays which hold position information. Each position is one row of the array. The
dimension of the space equals the number of columns. Each coordinate in one column.
To rescale, the mean (center) is subtracted from each axis separately. Then all values are scaled so that the
largest magnitude value from all axes equals scale (thus, the aspect ratio is preserved). The resulting NumPy
Array is returned (order of rows unchanged).
Parameters
• pos (numpy array) – positions to be scaled. Each row is a position.
• scale (number (default: 1)) – The size of the resulting extent in all directions.
Returns pos – scaled positions. Each row is a position.
Return type numpy array

10.4.7 networkx.drawing.layout.shell_layout

shell_layout(G, nlist=None, scale=1, center=None, dim=2)


Position nodes in concentric circles.
Parameters
• G (NetworkX graph or list of nodes) – A position will be assigned to every node in G.
• nlist (list of lists) – List of node lists for each shell.
• scale (number (default: 1)) – Scale factor for positions.
• center (array-like or None) – Coordinate pair around which to center the layout.
• dim (int) – Dimension of layout, currently only dim=2 is supported. Other dimension values
result in a ValueError.
Returns pos – A dictionary of positions keyed by node

678 Chapter 10. Drawing


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type dict


Raises ValueError – If dim != 2

Examples

>>> G = nx.path_graph(4)
>>> shells = [[0], [1, 2, 3]]
>>> pos = nx.shell_layout(G, shells)

Notes

This algorithm currently only works in two dimensions and does not try to minimize edge crossings.

10.4.8 networkx.drawing.layout.spring_layout

spring_layout(G, k=None, pos=None, fixed=None, iterations=50, threshold=0.0001, weight=’weight’,


scale=1, center=None, dim=2, seed=None)
Position nodes using Fruchterman-Reingold force-directed algorithm.
Parameters
• G (NetworkX graph or list of nodes) – A position will be assigned to every node in G.
• k (float (default=None)) – Optimal distance between nodes. If None the distance is set to
1/sqrt(n) where n is the number of nodes. Increase this value to move nodes farther apart.
• pos (dict or None optional (default=None)) – Initial positions for nodes as a dictionary
with node as keys and values as a coordinate list or tuple. If None, then use random initial
positions.
• fixed (list or None optional (default=None)) – Nodes to keep fixed at initial position.
• iterations (int optional (default=50)) – Maximum number of iterations taken
• threshold (float optional (default = 1e-4)) – Threshold for relative error in node position
changes. The iteration stops if the error is below this threshold.
• weight (string or None optional (default=’weight’)) – The edge attribute that holds the
numerical value used for the edge weight. If None, then all edge weights are 1.
• scale (number (default: 1)) – Scale factor for positions. Not used unless fixed is None.
• center (array-like or None) – Coordinate pair around which to center the layout. Not used
unless fixed is None.
• dim (int) – Dimension of layout.
• seed (int, RandomState instance or None optional (default=None)) – Set the random state
for deterministic node layouts. If int, seed is the seed used by the random number gen-
erator, if numpy.random.RandomState instance, seed is the random number generator, if
None, the random number generator is the RandomState instance used by numpy.random.
Returns pos – A dictionary of positions keyed by node
Return type dict

10.4. Graph Layout 679


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

>>> G = nx.path_graph(4)
>>> pos = nx.spring_layout(G)

# The same using longer but equivalent function name >>> pos = nx.fruchterman_reingold_layout(G)

10.4.9 networkx.drawing.layout.spectral_layout

spectral_layout(G, weight=’weight’, scale=1, center=None, dim=2)


Position nodes using the eigenvectors of the graph Laplacian.
Using the unnormalized Laplacion, the layout shows possible clusters of nodes which are an approximation
of the ratio cut. If dim is the number of dimensions then the positions are the entries of the dim eigenvectors
corresponding to the ascending eigenvalues starting from the second one.
Parameters
• G (NetworkX graph or list of nodes) – A position will be assigned to every node in G.
• weight (string or None optional (default=’weight’)) – The edge attribute that holds the
numerical value used for the edge weight. If None, then all edge weights are 1.
• scale (number (default: 1)) – Scale factor for positions.
• center (array-like or None) – Coordinate pair around which to center the layout.
• dim (int) – Dimension of layout.
Returns pos – A dictionary of positions keyed by node
Return type dict

Examples

>>> G = nx.path_graph(4)
>>> pos = nx.spectral_layout(G)

Notes

Directed graphs will be considered as undirected graphs when positioning the nodes.
For larger graphs (>500 nodes) this will use the SciPy sparse eigenvalue solver (ARPACK).

680 Chapter 10. Drawing


CHAPTER

ELEVEN

RANDOMNESS

Random Number Generators (RNGs) are often used when generating, drawing and computing properties or manipu-
lating networks. NetworkX provides functions which use one of two standard RNGs: NumPy’s package numpy.
random or Python’s built-in package random. They each provide the same algorithm for generating numbers
(Mersenne Twister). Their interfaces are similar (dangerously similar) and yet distinct. They each provide a global
default instance of their generator that is shared by all programs in a single session. For the most part you can use
the RNGs as NetworkX has them set up and you’ll get reasonable pseudorandom results (results that are statistically
random, but created in a deterministic manner).
Sometimes you want more control over how the numbers are generated. In particular, you need to set the seed of the
generator to make your results reproducible – either for scientific publication or for debugging. Both RNG packages
have easy functions to set the seed to any integer, thus determining the subsequent generated values. Since this package
(and many others) use both RNGs you may need to set the seed of both RNGs. Even if we strictly only used one of
the RNGs, you may find yourself using another package that uses the other. Setting the state of the two global RNGs
is as simple setting the seed of each RNG to an arbitrary integer:

>>> import random


>>> random.seed(246) # or any integer
>>> import numpy
>>> numpy.random.seed(4812)

Many users will be satisfied with this level of control.


For people who want even more control, we include an optional argument to functions that use an RNG. This argument
is called seed, but determines more than the seed of the RNG. It tells the function which RNG package to use, and
whether to use a global or local RNG.

>>> from networkx import path_graph, random_layout


>>> G = path_graph(9)
>>> pos = random_layout(G, seed=None) # use (either) global default RNG
>>> pos = random_layout(G, seed=42) # local RNG just for this call
>>> pos = random_layout(G, seed=numpy.random) # use numpy global RNG
>>> random_state = numpy.random.RandomState(42)
>>> pos = random_layout(G, seed=random_state) # use/reuse your own RNG

Each NetworkX function that uses an RNG was written with one RNG package in mind. It either uses random
or numpy.random by default. But some users want to only use a single RNG for all their code. This seed
argument provides a mechanism so that any function can use a numpy.random RNG even if the function is written
for random. It works as follows.
The default behavior (when seed=None) is to use the global RNG for the function’s preferred package. If seed is set
to an integer value, a local RNG is created with the indicated seed value and is used for the duration of that function
(including any calls to other functions) and then discarded. Alternatively, you can specify seed=numpy.random
to ensure that the global numpy RNG is used whether the function expects it or not. Finally, you can provide a numpy

681
NetworkX Reference, Release 2.3rc1.dev20190222214247

RNG to be used by the function. The RNG is then available to use in other functions or even other package like
sklearn. In this way you can use a single RNG for all random numbers in your project.
While it is possible to assign seed a random-style RNG for NetworkX functions written for the random package
API, the numpy RNG interface has too many nice features for us to ensure a random-style RNG will work in all
functions. In practice, you can do most things using only random RNGs (useful if numpy is not available). But your
experience will be richer if numpy is available.
To summarize, you can easily ignore the seed argument and use the global RNGs. You can specify to use only the
numpy global RNG with seed=numpy.random. You can use a local RNG by providing an integer seed value. And
you can provide your own numpy RNG, reusing it for all functions. It is easier to use numpy RNGs if you want a
single RNG for your computations.

682 Chapter 11. Randomness


CHAPTER

TWELVE

EXCEPTIONS

12.1 Exceptions

Base exceptions and errors for NetworkX.


class NetworkXException
Base class for exceptions in NetworkX.
class NetworkXError
Exception for a serious error in NetworkX
class NetworkXPointlessConcept
Raised when a null graph is provided as input to an algorithm that cannot use it.
The null graph is sometimes considered a pointless concept1 , thus the name of the exception.

References

class NetworkXAlgorithmError
Exception for unexpected termination of algorithms.
class NetworkXUnfeasible
Exception raised by algorithms trying to solve a problem instance that has no feasible solution.
class NetworkXNoPath
Exception for algorithms that should return a path when running on graphs where such a path does not exist.
class NetworkXNoCycle
Exception for algorithms that should return a cycle when running on graphs where such a cycle does not exist.
class NodeNotFound
Exception raised if requested node is not present in the graph
class HasACycle
Raised if a graph has a cycle when an algorithm expects that it will have no cycles.
class NetworkXUnbounded
Exception raised by algorithms trying to solve a maximization or a minimization problem instance that is un-
bounded.
class NetworkXNotImplemented
Exception raised by algorithms not implemented for a type of graph.
1 Harary, F. and Read, R. “Is the Null Graph a Pointless Concept?” In Graphs and Combinatorics Conference, George Washington University.

New York: Springer-Verlag, 1973.

683
NetworkX Reference, Release 2.3rc1.dev20190222214247

class AmbiguousSolution
Raised if more than one valid solution exists for an intermediary step of an algorithm.
In the face of ambiguity, refuse the temptation to guess. This may occur, for example, when trying to determine
the bipartite node sets in a disconnected bipartite graph when computing bipartite matchings.
class ExceededMaxIterations
Raised if a loop iterates too many times without breaking.
This may occur, for example, in an algorithm that computes progressively better approximations to a value but
exceeds an iteration bound specified by the user.
class PowerIterationFailedConvergence(num_iterations, *args, **kw)
Raised when the power iteration method fails to converge within a specified iteration limit.
num_iterations is the number of iterations that have been completed when this exception was raised.

684 Chapter 12. Exceptions


CHAPTER

THIRTEEN

UTILITIES

13.1 Helper Functions

Miscellaneous Helpers for NetworkX.


These are not imported into the base networkx namespace but can be accessed, for example, as

>>> import networkx


>>> networkx.utils.is_string_like('spam')
True

is_string_like(obj) Check if obj is string.


flatten(obj[, result]) Return flattened version of (possibly nested) iterable ob-
ject.
iterable(obj) Return True if obj is iterable with a well-defined len().
is_list_of_ints(intlist) Return True if list is a list of ints.
make_str(x) Returns the string representation of t.
generate_unique_node() Generate a unique node label.
default_opener(filename) Opens filename using system’s default program.
pairwise(iterable[, cyclic]) s -> (s0, s1), (s1, s2), (s2, s3), . . .
groups(many_to_one) Converts a many-to-one mapping into a one-to-many
mapping.
create_random_state([random_state]) Returns a numpy.random.RandomState instance de-
pending on input.

13.1.1 networkx.utils.misc.is_string_like

is_string_like(obj)
Check if obj is string.

13.1.2 networkx.utils.misc.flatten

flatten(obj, result=None)
Return flattened version of (possibly nested) iterable object.

685
NetworkX Reference, Release 2.3rc1.dev20190222214247

13.1.3 networkx.utils.misc.iterable

iterable(obj)
Return True if obj is iterable with a well-defined len().

13.1.4 networkx.utils.misc.is_list_of_ints

is_list_of_ints(intlist)
Return True if list is a list of ints.

13.1.5 networkx.utils.misc.make_str

make_str(x)
Returns the string representation of t.

13.1.6 networkx.utils.misc.generate_unique_node

generate_unique_node()
Generate a unique node label.

13.1.7 networkx.utils.misc.default_opener

default_opener(filename)
Opens filename using system’s default program.
Parameters filename (str) – The path of the file to be opened.

13.1.8 networkx.utils.misc.pairwise

pairwise(iterable, cyclic=False)
s -> (s0, s1), (s1, s2), (s2, s3), . . .

13.1.9 networkx.utils.misc.groups

groups(many_to_one)
Converts a many-to-one mapping into a one-to-many mapping.
many_to_one must be a dictionary whose keys and values are all hashable.
The return value is a dictionary mapping values from many_to_one to sets of keys from many_to_one that
have that value.
For example:

>>> from networkx.utils import groups


>>> many_to_one = {'a': 1, 'b': 1, 'c': 2, 'd': 3, 'e': 3}
>>> groups(many_to_one)
{1: {'a', 'b'}, 2: {'c'}, 3: {'d', 'e'}}

686 Chapter 13. Utilities


NetworkX Reference, Release 2.3rc1.dev20190222214247

13.1.10 networkx.utils.misc.create_random_state

create_random_state(random_state=None)
Returns a numpy.random.RandomState instance depending on input.
Parameters random_state (int or RandomState instance or None optional (default=None)) – If int,
return a numpy.random.RandomState instance set with seed=int. if numpy.random.RandomState
instance, return it. if None or numpy.random, return the global random number generator used
by numpy.random.

13.2 Data Structures and Algorithms

Union-find data structure.

UnionFind.union(*objects) Find the sets containing the objects and merge them all.

13.2.1 networkx.utils.union_find.UnionFind.union

UnionFind.union(*objects)
Find the sets containing the objects and merge them all.

13.3 Random Sequence Generators

Utilities for generating random numbers, random sequences, and random selections.

powerlaw_sequence(n[, exponent, seed]) Return sample sequence of length n from a power law
distribution.
cumulative_distribution(distribution) Returns normalized cumulative distribution from dis-
crete distribution.
discrete_sequence(n[, distribution, . . . ]) Return sample sequence of length n from a given dis-
crete distribution or discrete cumulative distribution.
zipf_rv(alpha[, xmin, seed]) Returns a random value chosen from the Zipf distribu-
tion.
random_weighted_sample(mapping, k[, seed]) Returns k items without replacement from a weighted
sample.
weighted_choice(mapping[, seed]) Returns a single element from a weighted sample.

13.3.1 networkx.utils.random_sequence.powerlaw_sequence

powerlaw_sequence(n, exponent=2.0, seed=None)


Return sample sequence of length n from a power law distribution.

13.3.2 networkx.utils.random_sequence.cumulative_distribution

cumulative_distribution(distribution)
Returns normalized cumulative distribution from discrete distribution.

13.3. Random Sequence Generators 687


NetworkX Reference, Release 2.3rc1.dev20190222214247

13.3.3 networkx.utils.random_sequence.discrete_sequence

discrete_sequence(n, distribution=None, cdistribution=None, seed=None)


Return sample sequence of length n from a given discrete distribution or discrete cumulative distribution.
One of the following must be specified.
distribution = histogram of values, will be normalized
cdistribution = normalized discrete cumulative distribution

13.3.4 networkx.utils.random_sequence.zipf_rv

zipf_rv(alpha, xmin=1, seed=None)


Returns a random value chosen from the Zipf distribution.
The return value is an integer drawn from the probability distribution

𝑥−𝛼
𝑝(𝑥) = ,
𝜁(𝛼, 𝑥min )

where 𝜁(𝛼, 𝑥min ) is the Hurwitz zeta function.


Parameters
• alpha (float) – Exponent value of the distribution
• xmin (int) – Minimum value
• seed (integer, random_state, or None (default)) – Indicator of random number generation
state. See Randomness.
Returns x – Random value from Zipf distribution
Return type int
Raises ValueError: – If xmin < 1 or If alpha <= 1

Notes

The rejection algorithm generates random values for a the power-law distribution in uniformly bounded expected
time dependent on parameters. See1 for details on its operation.

Examples

>>> nx.zipf_rv(alpha=2, xmin=3, seed=42) # doctest: +SKIP

References

13.3.5 networkx.utils.random_sequence.random_weighted_sample

random_weighted_sample(mapping, k, seed=None)
Returns k items without replacement from a weighted sample.
The input is a dictionary of items with weights as values.
1 Luc Devroye, Non-Uniform Random Variate Generation, Springer-Verlag, New York, 1986.

688 Chapter 13. Utilities


NetworkX Reference, Release 2.3rc1.dev20190222214247

13.3.6 networkx.utils.random_sequence.weighted_choice

weighted_choice(mapping, seed=None)
Returns a single element from a weighted sample.
The input is a dictionary of items with weights as values.

13.4 Decorators

open_file(path_arg[, mode]) Decorator to ensure clean opening and closing of files.


not_implemented_for(*graph_types) Decorator to mark algorithms as not implemented
nodes_or_number(which_args) Decorator to allow number of nodes or container of
nodes.
preserve_random_state(func) Decorator to preserve the numpy.random state during a
function.
random_state(random_state_index) Decorator to generate a numpy.random.RandomState
instance.

13.4.1 networkx.utils.decorators.open_file

open_file(path_arg, mode=’r’)
Decorator to ensure clean opening and closing of files.
Parameters
• path_arg (int) – Location of the path argument in args. Even if the argument is a named
positional argument (with a default value), you must specify its index as a positional argu-
ment.
• mode (str) – String for opening mode.
Returns _open_file – Function which cleanly executes the io.
Return type function

Examples

Decorate functions like this:

@open_file(0,'r')
def read_function(pathname):
pass

@open_file(1,'w')
def write_function(G,pathname):
pass

@open_file(1,'w')
def write_function(G, pathname='graph.dot')
pass

@open_file('path', 'w+')
def another_function(arg, **kwargs):
(continues on next page)

13.4. Decorators 689


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


path = kwargs['path']
pass

13.4.2 networkx.utils.decorators.not_implemented_for

not_implemented_for(*graph_types)
Decorator to mark algorithms as not implemented
Parameters graph_types (container of strings) – Entries must be one of ‘directed’,’undirected’,
‘multigraph’, ‘graph’.
Returns _require – The decorated function.
Return type function
Raises
• NetworkXNotImplemented
• If any of the packages cannot be imported

Notes

Multiple types are joined logically with “and”. For “or” use multiple @not_implemented_for() lines.

Examples

Decorate functions like this:

@not_implemnted_for('directed')
def sp_function(G):
pass

@not_implemnted_for('directed','multigraph')
def sp_np_function(G):
pass

13.4.3 networkx.utils.decorators.nodes_or_number

nodes_or_number(which_args)
Decorator to allow number of nodes or container of nodes.
Parameters which_args (int or sequence of ints) – Location of the node arguments in args. Even if
the argument is a named positional argument (with a default value), you must specify its index
as a positional argument. If more than one node argument is allowed, can be a list of locations.
Returns _nodes_or_numbers – Function which replaces int args with ranges.
Return type function

690 Chapter 13. Utilities


NetworkX Reference, Release 2.3rc1.dev20190222214247

Examples

Decorate functions like this:


@nodes_or_number(0)
def empty_graph(nodes):
pass

@nodes_or_number([0,1])
def grid_2d_graph(m1, m2, periodic=False):
pass

@nodes_or_number(1)
def full_rary_tree(r, n)
# r is a number. n can be a number of a list of nodes
pass

13.4.4 networkx.utils.decorators.preserve_random_state

preserve_random_state(func)
Decorator to preserve the numpy.random state during a function.
Parameters func (function) – function around which to preserve the random state.
Returns wrapper – Function which wraps the input function by saving the state before calling the
function and restoring the function afterward.
Return type function

Examples

Decorate functions like this:


@preserve_random_state
def do_random_stuff(x, y):
return x + y * numpy.random.random()

Notes

If numpy.random is not importable, the state is not saved or restored.

13.4.5 networkx.utils.decorators.random_state

random_state(random_state_index)
Decorator to generate a numpy.random.RandomState instance.
Argument position random_state_index is processed by create_random_state. The result is a
numpy.random.RandomState instance.
Parameters random_state_index (int) – Location of the random_state argument in args that is to
be used to generate the numpy.random.RandomState instance. Even if the argument is a named
positional argument (with a default value), you must specify its index as a positional argument.
Returns _random_state – Function whose random_state keyword argument is a RandomState in-
stance.

13.4. Decorators 691


NetworkX Reference, Release 2.3rc1.dev20190222214247

Return type function

Examples

Decorate functions like this:

@np_random_state(0)
def random_float(random_state=None):
return random_state.rand()

@np_random_state(1)
def random_array(dims, random_state=1):
return random_state.rand(*dims)

See also:
py_random_state()

13.5 Cuthill-Mckee Ordering

Cuthill-McKee ordering of graph nodes to produce sparse matrices

cuthill_mckee_ordering(G[, heuristic]) Generate an ordering (permutation) of the graph nodes


to make a sparse matrix.
reverse_cuthill_mckee_ordering(G[, Generate an ordering (permutation) of the graph nodes
heuristic]) to make a sparse matrix.

13.5.1 networkx.utils.rcm.cuthill_mckee_ordering

cuthill_mckee_ordering(G, heuristic=None)
Generate an ordering (permutation) of the graph nodes to make a sparse matrix.
Uses the Cuthill-McKee heuristic (based on breadth-first search)1 .
Parameters
• G (graph) – A NetworkX graph
• heuristic (function, optional) – Function to choose starting node for RCM algorithm. If
None a node from a pseudo-peripheral pair is used. A user-defined function can be supplied
that takes a graph object and returns a single node.
Returns nodes – Generator of nodes in Cuthill-McKee ordering.
Return type generator

Examples

>>> from networkx.utils import cuthill_mckee_ordering


>>> G = nx.path_graph(4)
(continues on next page)
1 E. Cuthill and J. McKee. Reducing the bandwidth of sparse symmetric matrices, In Proc. 24th Nat. Conf. ACM, pages 157-172, 1969.
http://doi.acm.org/10.1145/800195.805928

692 Chapter 13. Utilities


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> rcm = list(cuthill_mckee_ordering(G))
>>> A = nx.adjacency_matrix(G, nodelist=rcm) # doctest: +SKIP

Smallest degree node as heuristic function:

>>> def smallest_degree(G):


... return min(G, key=G.degree)
>>> rcm = list(cuthill_mckee_ordering(G, heuristic=smallest_degree))

See also:
reverse_cuthill_mckee_ordering()

Notes

The optimal solution the the bandwidth reduction is NP-complete2 .

References

13.5.2 networkx.utils.rcm.reverse_cuthill_mckee_ordering

reverse_cuthill_mckee_ordering(G, heuristic=None)
Generate an ordering (permutation) of the graph nodes to make a sparse matrix.
Uses the reverse Cuthill-McKee heuristic (based on breadth-first search)1 .
Parameters
• G (graph) – A NetworkX graph
• heuristic (function, optional) – Function to choose starting node for RCM algorithm. If
None a node from a pseudo-peripheral pair is used. A user-defined function can be supplied
that takes a graph object and returns a single node.
Returns nodes – Generator of nodes in reverse Cuthill-McKee ordering.
Return type generator

Examples

>>> from networkx.utils import reverse_cuthill_mckee_ordering


>>> G = nx.path_graph(4)
>>> rcm = list(reverse_cuthill_mckee_ordering(G))
>>> A = nx.adjacency_matrix(G, nodelist=rcm) # doctest: +SKIP

Smallest degree node as heuristic function:

>>> def smallest_degree(G):


... return min(G, key=G.degree)
>>> rcm = list(reverse_cuthill_mckee_ordering(G, heuristic=smallest_degree))

2 Steven S. Skiena. 1997. The Algorithm Design Manual. Springer-Verlag New York, Inc., New York, NY, USA.
1 E. Cuthill and J. McKee. Reducing the bandwidth of sparse symmetric matrices, In Proc. 24th Nat. Conf. ACM, pages 157-72, 1969.
http://doi.acm.org/10.1145/800195.805928

13.5. Cuthill-Mckee Ordering 693


NetworkX Reference, Release 2.3rc1.dev20190222214247

See also:
cuthill_mckee_ordering()

Notes

The optimal solution the the bandwidth reduction is NP-complete2 .

References

13.6 Context Managers

reversed(G) A context manager for temporarily reversing a directed


graph in place.

13.6.1 networkx.utils.contextmanagers.reversed

reversed(G)
A context manager for temporarily reversing a directed graph in place.
This is a no-op for undirected graphs.
Parameters G (graph) – A NetworkX graph.

2 Steven S. Skiena. 1997. The Algorithm Design Manual. Springer-Verlag New York, Inc., New York, NY, USA.

694 Chapter 13. Utilities


CHAPTER

FOURTEEN

GLOSSARY

dictionary A Python dictionary maps keys to values. Also known as “hashes”, or “associative arrays” in other pro-
gramming languages. See https://docs.python.org/2/tutorial/datastructures.html#dictionaries
edge Edges are either two-tuples of nodes (u, v) or three tuples of nodes with an edge attribute dictionary (u,
v, dict).
ebunch An iteratable container of edge tuples like a list, iterator, or file.
edge attribute Edges can have arbitrary Python objects assigned as attributes by using keyword/value pairs when
adding an edge assigning to the G.edges[u][v] attribute dictionary for the specified edge u-v.
hashable An object is hashable if it has a hash value which never changes during its lifetime (it needs a __hash__()
method), and can be compared to other objects (it needs an __eq__() or __cmp__() method). Hashable
objects which compare equal must have the same hash value.
Hashability makes an object usable as a dictionary key and a set member, because these data structures use the
hash value internally.
All of Python’s immutable built-in objects are hashable, while no mutable containers (such as lists or dictionar-
ies) are. Objects which are instances of user-defined classes are hashable by default; they all compare unequal,
and their hash value is their id().
Definition from https://docs.python.org/2/glossary.html
nbunch An nbunch is a single node, container of nodes or None (representing all nodes). It can be a list, set, graph,
etc.. To filter an nbunch so that only nodes actually in G appear, use G.nbunch_iter(nbunch).
node A node can be any hashable Python object except None.
node attribute Nodes can have arbitrary Python objects assigned as attributes by using keyword/value pairs when
adding a node or assigning to the G.nodes[n] attribute dictionary for the specified node n.

695
NetworkX Reference, Release 2.3rc1.dev20190222214247

696 Chapter 14. Glossary


APPENDIX

TUTORIAL

This guide can help you start working with NetworkX.

A.1 Creating a graph

Create an empty graph with no nodes and no edges.

>>> import networkx as nx


>>> G = nx.Graph()

By definition, a Graph is a collection of nodes (vertices) along with identified pairs of nodes (called edges, links,
etc). In NetworkX, nodes can be any hashable object e.g., a text string, an image, an XML object, another Graph, a
customized node object, etc.

Note: Python’s None object should not be used as a node as it determines whether optional function arguments have
been assigned in many functions.

A.2 Nodes

The graph G can be grown in several ways. NetworkX includes many graph generator functions and facilities to read
and write graphs in many formats. To get started though we’ll look at simple manipulations. You can add one node at
a time,

>>> G.add_node(1)

add a list of nodes,

>>> G.add_nodes_from([2, 3])

or add any iterable container of nodes. You can also add nodes along with node attributes if your container yields
2-tuples (node, node_attribute_dict). Node attributes are discussed further below.

>>> H = nx.path_graph(10)
>>> G.add_nodes_from(H)

Note that G now contains the nodes of H as nodes of G. In contrast, you could use the graph H as a node in G.

>>> G.add_node(H)

697
NetworkX Reference, Release 2.3rc1.dev20190222214247

The graph G now contains H as a node. This flexibility is very powerful as it allows graphs of graphs, graphs of files,
graphs of functions and much more. It is worth thinking about how to structure your application so that the nodes are
useful entities. Of course you can always use a unique identifier in G and have a separate dictionary keyed by identifier
to the node information if you prefer.

Note: You should not change the node object if the hash depends on its contents.

A.3 Edges

G can also be grown by adding one edge at a time,

>>> G.add_edge(1, 2)
>>> e = (2, 3)
>>> G.add_edge(*e) # unpack edge tuple*

by adding a list of edges,

>>> G.add_edges_from([(1, 2), (1, 3)])

or by adding any ebunch of edges. An ebunch is any iterable container of edge-tuples. An edge-tuple can be a
2-tuple of nodes or a 3-tuple with 2 nodes followed by an edge attribute dictionary, e.g., (2, 3, {'weight':
3.1415}). Edge attributes are discussed further below

>>> G.add_edges_from(H.edges)

There are no complaints when adding existing nodes or edges. For example, after removing all nodes and edges,

>>> G.clear()

we add new nodes/edges and NetworkX quietly ignores any that are already present.

>>> G.add_edges_from([(1, 2), (1, 3)])


>>> G.add_node(1)
>>> G.add_edge(1, 2)
>>> G.add_node("spam") # adds node "spam"
>>> G.add_nodes_from("spam") # adds 4 nodes: 's', 'p', 'a', 'm'
>>> G.add_edge(3, 'm')

At this stage the graph G consists of 8 nodes and 3 edges, as can be seen by:

>>> G.number_of_nodes()
8
>>> G.number_of_edges()
3

We can examine the nodes and edges. Four basic graph properties facilitate reporting: G.nodes, G.edges, G.adj
and G.degree. These are set-like views of the nodes, edges, neighbors (adjacencies), and degrees of nodes in a
graph. They offer a continually updated read-only view into the graph structure. They are also dict-like in that you
can look up node and edge data attributes via the views and iterate with data attributes using methods .items(),
.data('span'). If you want a specific container type instead of a view, you can specify one. Here we use lists,
though sets, dicts, tuples and other containers may be better in other contexts.

698 Appendix A. Tutorial


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> list(G.nodes)
[1, 2, 3, 'spam', 's', 'p', 'a', 'm']
>>> list(G.edges)
[(1, 2), (1, 3), (3, 'm')]
>>> list(G.adj[1]) # or list(G.neighbors(1))
[2, 3]
>>> G.degree[1] # the number of edges incident to 1
2

One can specify to report the edges and degree from a subset of all nodes using an nbunch. An nbunch is any of: None
(meaning all nodes), a node, or an iterable container of nodes that is not itself a node in the graph.

>>> G.edges([2, 'm'])


EdgeDataView([(2, 1), ('m', 3)])
>>> G.degree([2, 3])
DegreeView({2: 1, 3: 2})

One can remove nodes and edges from the graph in a similar fashion to adding. Use meth-
ods Graph.remove_node(), Graph.remove_nodes_from(), Graph.remove_edge() and Graph.
remove_edges_from(), e.g.

>>> G.remove_node(2)
>>> G.remove_nodes_from("spam")
>>> list(G.nodes)
[1, 3, 'spam']
>>> G.remove_edge(1, 3)

When creating a graph structure by instantiating one of the graph classes you can specify data in several formats.

>>> G.add_edge(1, 2)
>>> H = nx.DiGraph(G) # create a DiGraph using the connections from G
>>> list(H.edges())
[(1, 2), (2, 1)]
>>> edgelist = [(0, 1), (1, 2), (2, 3)]
>>> H = nx.Graph(edgelist)

A.4 What to use as nodes and edges

You might notice that nodes and edges are not specified as NetworkX objects. This leaves you free to use meaningful
items as nodes and edges. The most common choices are numbers or strings, but a node can be any hashable object
(except None), and an edge can be associated with any object x using G.add_edge(n1, n2, object=x).
As an example, n1 and n2 could be protein objects from the RCSB Protein Data Bank, and x could refer to an XML
record of publications detailing experimental observations of their interaction.
We have found this power quite useful, but its abuse can lead to unexpected surprises unless one is familiar with Python.
If in doubt, consider using convert_node_labels_to_integers() to obtain a more traditional graph with
integer labels.

A.5 Accessing edges and neighbors

In addition to the views Graph.edges(), and Graph.adj(), access to edges and neighbors is possible using
subscript notation.

A.4. What to use as nodes and edges 699


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G[1] # same as G.adj[1]


AtlasView({2: {}})
>>> G[1][2]
{}
>>> G.edges[1, 2]
{}

You can get/set the attributes of an edge using subscript notation if the edge already exists.

>>> G.add_edge(1, 3)
>>> G[1][3]['color'] = "blue"
>>> G.edges[1, 2]['color'] = "red"

Fast examination of all (node, adjacency) pairs is achieved using G.adjacency(), or G.adj.items(). Note
that for undirected graphs, adjacency iteration sees each edge twice.

>>> FG = nx.Graph()
>>> FG.add_weighted_edges_from([(1, 2, 0.125), (1, 3, 0.75), (2, 4, 1.2), (3, 4, 0.
˓→375)])

>>> for n, nbrs in FG.adj.items():


... for nbr, eattr in nbrs.items():
... wt = eattr['weight']
... if wt < 0.5: print('(%d, %d, %.3f)' % (n, nbr, wt))
(1, 2, 0.125)
(2, 1, 0.125)
(3, 4, 0.375)
(4, 3, 0.375)

Convenient access to all edges is achieved with the edges property.

>>> for (u, v, wt) in FG.edges.data('weight'):


... if wt < 0.5: print('(%d, %d, %.3f)' % (u, v, wt))
(1, 2, 0.125)
(3, 4, 0.375)

A.6 Adding attributes to graphs, nodes, and edges

Attributes such as weights, labels, colors, or whatever Python object you like, can be attached to graphs, nodes, or
edges.
Each graph, node, and edge can hold key/value attribute pairs in an associated attribute dictionary (the keys must be
hashable). By default these are empty, but attributes can be added or changed using add_edge, add_node or direct
manipulation of the attribute dictionaries named G.graph, G.nodes, and G.edges for a graph G.

A.6.1 Graph attributes

Assign graph attributes when creating a new graph

>>> G = nx.Graph(day="Friday")
>>> G.graph
{'day': 'Friday'}

Or you can modify attributes later

700 Appendix A. Tutorial


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G.graph['day'] = "Monday"


>>> G.graph
{'day': 'Monday'}

A.6.2 Node attributes

Add node attributes using add_node(), add_nodes_from(), or G.nodes

>>> G.add_node(1, time='5pm')


>>> G.add_nodes_from([3], time='2pm')
>>> G.nodes[1]
{'time': '5pm'}
>>> G.nodes[1]['room'] = 714
>>> G.nodes.data()
NodeDataView({1: {'time': '5pm', 'room': 714}, 3: {'time': '2pm'}})

Note that adding a node to G.nodes does not add it to the graph, use G.add_node() to add new nodes. Similarly
for edges.

A.6.3 Edge Attributes

Add/change edge attributes using add_edge(), add_edges_from(), or subscript notation.

>>> G.add_edge(1, 2, weight=4.7 )


>>> G.add_edges_from([(3, 4), (4, 5)], color='red')
>>> G.add_edges_from([(1, 2, {'color': 'blue'}), (2, 3, {'weight': 8})])
>>> G[1][2]['weight'] = 4.7
>>> G.edges[3, 4]['weight'] = 4.2

The special attribute weight should be numeric as it is used by algorithms requiring weighted edges.

A.7 Directed graphs

The DiGraph class provides additional properties specific to directed edges, e.g., DiGraph.out_edges(),
DiGraph.in_degree(), DiGraph.predecessors(), DiGraph.successors() etc. To allow algo-
rithms to work with both classes easily, the directed versions of neighbors() is equivalent to successors()
while degree reports the sum of in_degree and out_degree even though that may feel inconsistent at times.

>>> DG = nx.DiGraph()
>>> DG.add_weighted_edges_from([(1, 2, 0.5), (3, 1, 0.75)])
>>> DG.out_degree(1, weight='weight')
0.5
>>> DG.degree(1, weight='weight')
1.25
>>> list(DG.successors(1))
[2]
>>> list(DG.neighbors(1))
[2]

Some algorithms work only for directed graphs and others are not well defined for directed graphs. Indeed the tendency
to lump directed and undirected graphs together is dangerous. If you want to treat a directed graph as undirected for
some measurement you should probably convert it using Graph.to_undirected() or with

A.7. Directed graphs 701


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> H = nx.Graph(G) # convert G to undirected graph

A.8 Multigraphs

NetworkX provides classes for graphs which allow multiple edges between any pair of nodes. The MultiGraph
and MultiDiGraph classes allow you to add the same edge twice, possibly with different edge data. This can be
powerful for some applications, but many algorithms are not well defined on such graphs. Where results are well
defined, e.g., MultiGraph.degree() we provide the function. Otherwise you should convert to a standard graph
in a way that makes the measurement well defined.

>>> MG = nx.MultiGraph()
>>> MG.add_weighted_edges_from([(1, 2, 0.5), (1, 2, 0.75), (2, 3, 0.5)])
>>> dict(MG.degree(weight='weight'))
{1: 1.25, 2: 1.75, 3: 0.5}
>>> GG = nx.Graph()
>>> for n, nbrs in MG.adjacency():
... for nbr, edict in nbrs.items():
... minvalue = min([d['weight'] for d in edict.values()])
... GG.add_edge(n, nbr, weight = minvalue)
...
>>> nx.shortest_path(GG, 1, 3)
[1, 2, 3]

A.9 Graph generators and graph operations

In addition to constructing graphs node-by-node or edge-by-edge, they can also be generated by


1. Applying classic graph operations, such as:

subgraph(G, nbunch) - induced subgraph view of G on nodes in nbunch


union(G1,G2) - graph union
disjoint_union(G1,G2) - graph union assuming all nodes are different
cartesian_product(G1,G2) - return Cartesian product graph
compose(G1,G2) - combine graphs identifying nodes common to both
complement(G) - graph complement
create_empty_copy(G) - return an empty copy of the same graph class
to_undirected(G) - return an undirected representation of G
to_directed(G) - return a directed representation of G

2. Using a call to one of the classic small graphs, e.g.,

>>> petersen = nx.petersen_graph()


>>> tutte = nx.tutte_graph()
>>> maze = nx.sedgewick_maze_graph()
>>> tet = nx.tetrahedral_graph()

3. Using a (constructive) generator for a classic graph, e.g.,

>>> K_5 = nx.complete_graph(5)


>>> K_3_5 = nx.complete_bipartite_graph(3, 5)
>>> barbell = nx.barbell_graph(10, 10)
>>> lollipop = nx.lollipop_graph(10, 20)

702 Appendix A. Tutorial


NetworkX Reference, Release 2.3rc1.dev20190222214247

4. Using a stochastic graph generator, e.g.,

>>> er = nx.erdos_renyi_graph(100, 0.15)


>>> ws = nx.watts_strogatz_graph(30, 3, 0.1)
>>> ba = nx.barabasi_albert_graph(100, 5)
>>> red = nx.random_lobster(100, 0.9, 0.9)

5. Reading a graph stored in a file using common graph formats, such as edge lists, adjacency lists, GML,
GraphML, pickle, LEDA and others.

>>> nx.write_gml(red, "path.to.file")


>>> mygraph = nx.read_gml("path.to.file")

For details on graph formats see Reading and writing graphs and for graph generator functions see Graph generators

A.10 Analyzing graphs

The structure of G can be analyzed using various graph-theoretic functions such as:

>>> G = nx.Graph()
>>> G.add_edges_from([(1, 2), (1, 3)])
>>> G.add_node("spam") # adds node "spam"
>>> list(nx.connected_components(G))
[{1, 2, 3}, {'spam'}]
>>> sorted(d for n, d in G.degree())
[0, 1, 1, 2]
>>> nx.clustering(G)
{1: 0, 2: 0, 3: 0, 'spam': 0}

Some functions with large output iterate over (node, value) 2-tuples. These are easily stored in a dict structure if
you desire.

>>> sp = dict(nx.all_pairs_shortest_path(G))
>>> sp[3]
{3: [3], 1: [3, 1], 2: [3, 1, 2]}

See Algorithms for details on graph algorithms supported.

A.11 Drawing graphs

NetworkX is not primarily a graph drawing package but basic drawing with Matplotlib as well as an interface to use
the open source Graphviz software package are included. These are part of the networkx.drawing module and
will be imported if possible.
First import Matplotlib’s plot interface (pylab works too)

>>> import matplotlib.pyplot as plt

You may find it useful to interactively test code using ipython -pylab, which combines the power of ipython and
matplotlib and provides a convenient interactive mode.
To test if the import of networkx.drawing was successful draw G using one of

A.10. Analyzing graphs 703


NetworkX Reference, Release 2.3rc1.dev20190222214247

>>> G = nx.petersen_graph()
>>> plt.subplot(121)
<matplotlib.axes._subplots.AxesSubplot object at ...>
>>> nx.draw(G, with_labels=True, font_weight='bold')
>>> plt.subplot(122)
<matplotlib.axes._subplots.AxesSubplot object at ...>
>>> nx.draw_shell(G, nlist=[range(5, 10), range(5)], with_labels=True, font_weight=
˓→'bold')

0 1
8
4 2
6
5 6 7
1
5 0
3
9 8
9
3
7
2 4

when drawing to an interactive display. Note that you may need to issue a Matplotlib

>>> plt.show()

command if you are not using matplotlib in interactive mode (see Matplotlib FAQ ).

>>> options = {
... 'node_color': 'black',
... 'node_size': 100,
... 'width': 3,
... }
>>> plt.subplot(221)
<matplotlib.axes._subplots.AxesSubplot object at ...>
>>> nx.draw_random(G, **options)
>>> plt.subplot(222)
<matplotlib.axes._subplots.AxesSubplot object at ...>
(continues on next page)

704 Appendix A. Tutorial


NetworkX Reference, Release 2.3rc1.dev20190222214247

(continued from previous page)


>>> nx.draw_circular(G, **options)
>>> plt.subplot(223)
<matplotlib.axes._subplots.AxesSubplot object at ...>
>>> nx.draw_spectral(G, **options)
>>> plt.subplot(224)
<matplotlib.axes._subplots.AxesSubplot object at ...>
>>> nx.draw_shell(G, nlist=[range(5,10), range(5)], **options)

You can find additional options via draw_networkx() and layouts via layout. You can use multiple shells with
draw_shell().

>>> G = nx.dodecahedral_graph()
>>> shells = [[2, 3, 4, 5, 6], [8, 1, 0, 19, 18, 17, 16, 15, 14, 7], [9, 10, 11, 12,
˓→13]]

>>> nx.draw_shell(G, nlist=shells, **options)

A.11. Drawing graphs 705


NetworkX Reference, Release 2.3rc1.dev20190222214247

To save drawings to a file, use, for example

>>> nx.draw(G)
>>> plt.savefig("path.png")

writes to the file path.png in the local directory. If Graphviz and PyGraphviz or pydot, are available on your system,
you can also use nx_agraph.graphviz_layout(G) or nx_pydot.graphviz_layout(G) to get the node
positions, or write the graph in dot format for further processing.

>>> from networkx.drawing.nx_pydot import write_dot


>>> pos = nx.nx_agraph.graphviz_layout(G)
>>> nx.draw(G, pos=pos)
>>> write_dot(G, 'file.dot')

See Drawing for additional details.


• ;
• ;
• .

706 Appendix A. Tutorial


BIBLIOGRAPHY

[atlas] Ronald C. Read and Robin J. Wilson, An Atlas of Graphs. Oxford University Press, 1998.

707
NetworkX Reference, Release 2.3rc1.dev20190222214247

708 Bibliography
PYTHON MODULE INDEX

a 169
networkx.algorithms.approximation, 129 networkx.algorithms.bipartite.spectral,
networkx.algorithms.approximation.clique, 165
133 networkx.algorithms.boundary, 178
networkx.algorithms.bridges,
networkx.algorithms.approximation.clustering_coefficient, 180
135 networkx.algorithms.centrality, 301
networkx.algorithms.chains, 210
networkx.algorithms.approximation.connectivity,
129 networkx.algorithms.chordal, 211
networkx.algorithms.clique, 213
networkx.algorithms.approximation.dominating_set,
135 networkx.algorithms.cluster, 217
networkx.algorithms.coloring, 222
networkx.algorithms.approximation.independent_set,
136 networkx.algorithms.communicability_alg,
networkx.algorithms.approximation.kcomponents, 226
132 networkx.algorithms.community, 227
networkx.algorithms.community.asyn_fluid,
networkx.algorithms.approximation.matching,
137 233
networkx.algorithms.approximation.ramsey,networkx.algorithms.community.centrality,
138 235
networkx.algorithms.community.community_generators,
networkx.algorithms.approximation.steinertree,
138 228
networkx.algorithms.community.community_utils,
networkx.algorithms.approximation.treewidth,
139 237
networkx.algorithms.community.kclique,
networkx.algorithms.approximation.vertex_cover,
140 231
networkx.algorithms.assortativity, 140 networkx.algorithms.community.kernighan_lin,
networkx.algorithms.bipartite, 150 228
networkx.algorithms.bipartite.basic, 151 networkx.algorithms.community.label_propagation,
networkx.algorithms.bipartite.centrality, 232
170 networkx.algorithms.community.modularity_max,
networkx.algorithms.bipartite.cluster, 231
165 networkx.algorithms.community.quality,
networkx.algorithms.bipartite.covering, 234
178 networkx.algorithms.components, 238
networkx.algorithms.bipartite.generators,networkx.algorithms.connectivity, 252
173 networkx.algorithms.connectivity.connectivity,
networkx.algorithms.bipartite.matching, 265
155 networkx.algorithms.connectivity.cuts,
networkx.algorithms.bipartite.matrix, 273
157 networkx.algorithms.connectivity.disjoint_paths,
networkx.algorithms.bipartite.projection, 262
159 networkx.algorithms.connectivity.edge_augmentation,
networkx.algorithms.bipartite.redundancy, 252

709
NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.reciprocity, 399
networkx.algorithms.connectivity.edge_kcomponents,
255 networkx.algorithms.richclub, 400
networkx.algorithms.connectivity.kcomponents,
networkx.algorithms.shortest_paths.astar,
259 436
networkx.algorithms.connectivity.kcutsets,
networkx.algorithms.shortest_paths.dense,
260 434
networkx.algorithms.connectivity.stoerwagner,
networkx.algorithms.shortest_paths.generic,
280 401
networkx.algorithms.connectivity.utils, networkx.algorithms.shortest_paths.unweighted,
281 406
networkx.algorithms.core, 282 networkx.algorithms.shortest_paths.weighted,
networkx.algorithms.covering, 285 411
networkx.algorithms.cuts, 290 networkx.algorithms.similarity, 437
networkx.algorithms.cycles, 286 networkx.algorithms.simple_paths, 444
networkx.algorithms.dag, 294 networkx.algorithms.smallworld, 448
networkx.algorithms.distance_measures, networkx.algorithms.smetric, 451
302 networkx.algorithms.sparsifiers, 452
networkx.algorithms.distance_regular, networkx.algorithms.structuralholes, 453
305 networkx.algorithms.swap, 455
networkx.algorithms.dominance, 307 networkx.algorithms.tournament, 457
networkx.algorithms.dominating, 308 networkx.algorithms.traversal.beamsearch,
networkx.algorithms.efficiency, 310 468
networkx.algorithms.euler, 311 networkx.algorithms.traversal.breadth_first_search,
networkx.algorithms.flow, 313 465
networkx.algorithms.graphical, 340 networkx.algorithms.traversal.depth_first_search,
networkx.algorithms.hierarchy, 344 460
networkx.algorithms.hybrid, 344 networkx.algorithms.traversal.edgebfs,
networkx.algorithms.isolate, 346 470
networkx.algorithms.isomorphism, 347 networkx.algorithms.traversal.edgedfs,
networkx.algorithms.isomorphism.isomorphvf2, 469
350 networkx.algorithms.tree.branchings, 474
networkx.algorithms.tree.coding, 477
networkx.algorithms.link_analysis.hits_alg,
365 networkx.algorithms.tree.mst, 482
networkx.algorithms.tree.operations, 481
networkx.algorithms.link_analysis.pagerank_alg,
361 networkx.algorithms.tree.recognition,
networkx.algorithms.link_prediction, 367 472
networkx.algorithms.triads, 486
networkx.algorithms.lowest_common_ancestors,
373 networkx.algorithms.vitality, 487
networkx.algorithms.matching, 375 networkx.algorithms.voronoi, 488
networkx.algorithms.minors, 377 networkx.algorithms.wiener, 489
networkx.algorithms.mis, 383
networkx.algorithms.node_classification, c
384 networkx.classes.function, 491
networkx.algorithms.node_classification.hmn,
networkx.classes.graphviews, 127
384 networkx.classes.ordered, 126
networkx.algorithms.node_classification.lgc,
networkx.convert, 595
385 networkx.convert_matrix, 598
networkx.algorithms.operators.all, 389
networkx.algorithms.operators.binary, d
387 networkx.drawing.layout, 675
networkx.algorithms.operators.product, networkx.drawing.nx_agraph, 670
391 networkx.drawing.nx_pydot, 672
networkx.algorithms.operators.unary, 386 networkx.drawing.nx_pylab, 661
networkx.algorithms.planarity, 395

710 Python Module Index


NetworkX Reference, Release 2.3rc1.dev20190222214247

e networkx.readwrite.nx_yaml, 649
networkx.exception, 683 networkx.readwrite.pajek, 657
networkx.readwrite.sparse6, 654
g networkx.relabel, 613
networkx.generators.atlas, 509
networkx.generators.classic, 510
u
networkx.generators.community, 563 networkx.utils, 685
networkx.generators.degree_seq, 539 networkx.utils.contextmanagers, 694
networkx.generators.directed, 546 networkx.utils.decorators, 689
networkx.generators.duplication, 537 networkx.utils.misc, 685
networkx.generators.ego, 559 networkx.utils.random_sequence, 687
networkx.generators.expanders, 518 networkx.utils.rcm, 692
networkx.generators.geometric, 550 networkx.utils.union_find, 687
networkx.generators.intersection, 561
networkx.generators.joint_degree_seq,
574
networkx.generators.lattice, 519
networkx.generators.line, 557
networkx.generators.mycielski, 576
networkx.generators.nonisomorphic_trees,
573
networkx.generators.random_clustered,
545
networkx.generators.random_graphs, 526
networkx.generators.small, 522
networkx.generators.social, 562
networkx.generators.spectral_graph_forge,
570
networkx.generators.stochastic, 560
networkx.generators.trees, 571
networkx.generators.triads, 574

l
networkx.linalg.algebraicconnectivity,
584
networkx.linalg.attrmatrix, 587
networkx.linalg.graphmatrix, 579
networkx.linalg.laplacianmatrix, 580
networkx.linalg.modularitymatrix, 591
networkx.linalg.spectrum, 583

r
networkx.readwrite.adjlist, 617
networkx.readwrite.edgelist, 624
networkx.readwrite.gexf, 630
networkx.readwrite.gml, 633
networkx.readwrite.gpickle, 638
networkx.readwrite.graph6, 651
networkx.readwrite.graphml, 639
networkx.readwrite.json_graph, 643
networkx.readwrite.leda, 648
networkx.readwrite.multiline_adjlist,
620
networkx.readwrite.nx_shp, 659

Python Module Index 711


NetworkX Reference, Release 2.3rc1.dev20190222214247

712 Python Module Index


INDEX

Symbols add_half_edge_cw() (PlanarEmbedding method),


__contains__() (DiGraph method), 48 397
__contains__() (Graph method), 22 add_half_edge_first() (PlanarEmbedding
__contains__() (MultiDiGraph method), 109 method), 397
__contains__() (MultiGraph method), 80 add_node() (DiGraph method), 39
__getitem__() (DiGraph method), 53 add_node() (Graph method), 12
__getitem__() (Graph method), 25 add_node() (MultiDiGraph method), 98
__getitem__() (MultiDiGraph method), 115 add_node() (MultiGraph method), 69
__getitem__() (MultiGraph method), 84 add_nodes_from() (DiGraph method), 39
__init__() (DiGraph method), 38 add_nodes_from() (Graph method), 13
__init__() (DiGraphMatcher method), 354 add_nodes_from() (MultiDiGraph method), 98
__init__() (EdgeComponentAuxGraph method), 259 add_nodes_from() (MultiGraph method), 70
__init__() (Edmonds method), 477 add_path() (in module networkx.classes.function),
__init__() (Graph method), 11 494
__init__() (GraphMatcher method), 352 add_star() (in module networkx.classes.function),
__init__() (MultiDiGraph method), 97 493
__init__() (MultiGraph method), 69 add_weighted_edges_from() (DiGraph method),
__iter__() (DiGraph method), 48 42
__iter__() (Graph method), 21 add_weighted_edges_from() (Graph method),
__iter__() (MultiDiGraph method), 108 16
__iter__() (MultiGraph method), 80 add_weighted_edges_from() (MultiDiGraph
__len__() (DiGraph method), 56 method), 102
__len__() (Graph method), 27 add_weighted_edges_from() (MultiGraph
__len__() (MultiDiGraph method), 118 method), 74
__len__() (MultiGraph method), 86 adj (DiGraph attribute), 53
adj (Graph attribute), 25
A adj (MultiDiGraph attribute), 114
adj (MultiGraph attribute), 84
adamic_adar_index() (in module net-
adjacency() (DiGraph method), 55
workx.algorithms.link_prediction), 369
adjacency() (Graph method), 26
add_cycle() (in module networkx.classes.function),
adjacency() (MultiDiGraph method), 116
494
adjacency() (MultiGraph method), 85
add_edge() (DiGraph method), 41
adjacency_data() (in module net-
add_edge() (Graph method), 14
workx.readwrite.json_graph), 645
add_edge() (MultiDiGraph method), 100
adjacency_graph() (in module net-
add_edge() (MultiGraph method), 72
workx.readwrite.json_graph), 645
add_edges_from() (DiGraph method), 42
adjacency_matrix() (in module net-
add_edges_from() (Graph method), 15
workx.linalg.graphmatrix), 579
add_edges_from() (MultiDiGraph method), 101
adjacency_spectrum() (in module net-
add_edges_from() (MultiGraph method), 73
workx.linalg.spectrum), 583
add_half_edge_ccw() (PlanarEmbedding
algebraic_connectivity() (in module net-
method), 397
workx.linalg.algebraicconnectivity), 584

713
NetworkX Reference, Release 2.3rc1.dev20190222214247

all_neighbors() (in module net- 299


workx.classes.function), 498 approximate_current_flow_betweenness_centrality()
all_node_cuts() (in module net- (in module networkx.algorithms.centrality),
workx.algorithms.connectivity.kcutsets), 198
261 articulation_points() (in module net-
all_pairs_bellman_ford_path() (in module workx.algorithms.components), 250
networkx.algorithms.shortest_paths.weighted), astar_path() (in module net-
428 workx.algorithms.shortest_paths.astar),
all_pairs_bellman_ford_path_length() 436
(in module net- astar_path_length() (in module net-
workx.algorithms.shortest_paths.weighted), workx.algorithms.shortest_paths.astar),
429 437
all_pairs_dijkstra() (in module net- asyn_fluidc() (in module net-
workx.algorithms.shortest_paths.weighted), workx.algorithms.community.asyn_fluid),
421 233
all_pairs_dijkstra_path() (in module net- asyn_lpa_communities() (in module net-
workx.algorithms.shortest_paths.weighted), workx.algorithms.community.label_propagation),
422 232
all_pairs_dijkstra_path_length() attr_matrix() (in module net-
(in module net- workx.linalg.attrmatrix), 587
workx.algorithms.shortest_paths.weighted), attr_sparse_matrix() (in module net-
423 workx.linalg.attrmatrix), 589
all_pairs_lowest_common_ancestor() attracting_component_subgraphs() (in mod-
(in module net- ule networkx.algorithms.components), 247
workx.algorithms.lowest_common_ancestors), attracting_components() (in module net-
373 workx.algorithms.components), 246
all_pairs_node_connectivity() attribute_assortativity_coefficient()
(in module net- (in module networkx.algorithms.assortativity),
workx.algorithms.approximation.connectivity), 142
129 attribute_mixing_dict() (in module net-
all_pairs_node_connectivity() (in module workx.algorithms.assortativity), 149
networkx.algorithms.connectivity.connectivity), attribute_mixing_matrix() (in module net-
266 workx.algorithms.assortativity), 147
all_pairs_shortest_path() (in module net- authority_matrix() (in module net-
workx.algorithms.shortest_paths.unweighted), workx.algorithms.link_analysis.hits_alg),
409 367
all_pairs_shortest_path_length() average_clustering() (in module net-
(in module net- workx.algorithms.approximation.clustering_coefficient),
workx.algorithms.shortest_paths.unweighted), 135
409 average_clustering() (in module net-
all_shortest_paths() (in module net- workx.algorithms.bipartite.cluster), 167
workx.algorithms.shortest_paths.generic), average_clustering() (in module net-
402 workx.algorithms.cluster), 220
all_simple_paths() (in module net- average_degree_connectivity() (in module
workx.algorithms.simple_paths), 444 networkx.algorithms.assortativity), 145
all_topological_sorts() (in module net- average_neighbor_degree() (in module net-
workx.algorithms.dag), 296 workx.algorithms.assortativity), 144
alternating_havel_hakimi_graph() (in mod- average_node_connectivity() (in module
ule networkx.algorithms.bipartite.generators), networkx.algorithms.connectivity.connectivity),
176 266
AmbiguousSolution (class in networkx), 683 average_shortest_path_length() (in module
ancestors() (in module networkx.algorithms.dag), networkx.algorithms.shortest_paths.generic),
295 404
antichains() (in module networkx.algorithms.dag),

714 Index
NetworkX Reference, Release 2.3rc1.dev20190222214247

B workx.generators.random_graphs), 530
balanced_tree() (in module net- bipartite_layout() (in module net-
workx.generators.classic), 511 workx.drawing.layout), 676
barabasi_albert_graph() (in module net- boundary_expansion() (in module net-
workx.generators.random_graphs), 533 workx.algorithms.cuts), 290
barbell_graph() (in module net- boykov_kolmogorov() (in module net-
workx.generators.classic), 511 workx.algorithms.flow), 327
bellman_ford_path() (in module net- branching_weight() (in module net-
workx.algorithms.shortest_paths.weighted), workx.algorithms.tree.branchings), 475
425 bridge_components() (in module net-
bellman_ford_path_length() (in module net- workx.algorithms.connectivity.edge_kcomponents),
workx.algorithms.shortest_paths.weighted), 257
425 bridges() (in module networkx.algorithms.bridges),
bellman_ford_predecessor_and_distance() 180
(in module net- build_auxiliary_edge_connectivity() (in
workx.algorithms.shortest_paths.weighted), module networkx.algorithms.connectivity.utils),
430 281
betweenness_centrality() (in module net- build_auxiliary_node_connectivity() (in
workx.algorithms.bipartite.centrality), 172 module networkx.algorithms.connectivity.utils),
betweenness_centrality() (in module net- 281
workx.algorithms.centrality), 192 build_residual_network() (in module net-
betweenness_centrality_subset() (in mod- workx.algorithms.flow), 331
ule networkx.algorithms.centrality), 194 bull_graph() (in module net-
bfs_beam_edges() (in module net- workx.generators.small), 524
workx.algorithms.traversal.beamsearch),
468 C
bfs_edges() (in module net- candidate_pairs_iter() (DiGraphMatcher
workx.algorithms.traversal.breadth_first_search), method), 355
465 candidate_pairs_iter() (GraphMatcher
bfs_predecessors() (in module net- method), 353
workx.algorithms.traversal.breadth_first_search), capacity_scaling() (in module net-
466 workx.algorithms.flow), 338
bfs_successors() (in module net- cartesian_product() (in module net-
workx.algorithms.traversal.breadth_first_search), workx.algorithms.operators.product), 391
467 categorical_edge_match() (in module net-
bfs_tree() (in module net- workx.algorithms.isomorphism), 356
workx.algorithms.traversal.breadth_first_search), categorical_multiedge_match() (in module
466 networkx.algorithms.isomorphism), 357
biadjacency_matrix() (in module net- categorical_node_match() (in module net-
workx.algorithms.bipartite.matrix), 157 workx.algorithms.isomorphism), 356
biconnected_component_edges() (in module caveman_graph() (in module net-
networkx.algorithms.components), 249 workx.generators.community), 564
biconnected_component_subgraphs() (in center() (in module net-
module networkx.algorithms.components), 250 workx.algorithms.distance_measures), 302
biconnected_components() (in module net- chain_decomposition() (in module net-
workx.algorithms.components), 248 workx.algorithms.chains), 210
bidirectional_dijkstra() (in module net- check_planarity() (in module net-
workx.algorithms.shortest_paths.weighted), workx.algorithms.planarity), 395
424 check_structure() (PlanarEmbedding method),
bidirectional_shortest_path() 398
(in module net- chordal_cycle_graph() (in module net-
workx.algorithms.shortest_paths.unweighted), workx.generators.expanders), 519
408 chordal_graph_cliques() (in module net-
binomial_graph() (in module net- workx.algorithms.chordal), 211

Index 715
NetworkX Reference, Release 2.3rc1.dev20190222214247

chordal_graph_treewidth() (in module net- networkx.generators.classic), 512


workx.algorithms.chordal), 212 compose() (in module net-
chvatal_graph() (in module net- workx.algorithms.operators.binary), 387
workx.generators.small), 524 compose_all() (in module net-
circulant_graph() (in module net- workx.algorithms.operators.all), 389
workx.generators.classic), 513 condensation() (in module net-
circular_ladder_graph() (in module net- workx.algorithms.components), 243
workx.generators.classic), 513 conductance() (in module net-
circular_layout() (in module net- workx.algorithms.cuts), 291
workx.drawing.layout), 676 configuration_model() (in module net-
clear() (DiGraph method), 45 workx.algorithms.bipartite.generators), 174
clear() (Graph method), 19 configuration_model() (in module net-
clear() (MultiDiGraph method), 106 workx.generators.degree_seq), 539
clear() (MultiGraph method), 77 connect_components() (PlanarEmbedding
clique_removal() (in module net- method), 398
workx.algorithms.approximation.clique), connected_caveman_graph() (in module net-
134 workx.generators.community), 564
cliques_containing_node() (in module net- connected_component_subgraphs() (in mod-
workx.algorithms.clique), 217 ule networkx.algorithms.components), 239
closeness_centrality() (in module net- connected_components() (in module net-
workx.algorithms.bipartite.centrality), 171 workx.algorithms.components), 239
closeness_centrality() (in module net- connected_double_edge_swap() (in module net-
workx.algorithms.centrality), 189 workx.algorithms.swap), 456
closeness_vitality() (in module net- connected_watts_strogatz_graph() (in mod-
workx.algorithms.vitality), 487 ule networkx.generators.random_graphs), 532
clustering() (in module net- constraint() (in module net-
workx.algorithms.bipartite.cluster), 166 workx.algorithms.structuralholes), 453
clustering() (in module net- contracted_edge() (in module net-
workx.algorithms.cluster), 219 workx.algorithms.minors), 377
cn_soundarajan_hopcroft() (in module net- contracted_nodes() (in module net-
workx.algorithms.link_prediction), 370 workx.algorithms.minors), 378
collaboration_weighted_projected_graph()convert_node_labels_to_integers() (in
(in module net- module networkx.relabel), 613
workx.algorithms.bipartite.projection), 161 copy() (DiGraph method), 60
color() (in module net- copy() (Graph method), 30
workx.algorithms.bipartite.basic), 153 copy() (MultiDiGraph method), 122
common_neighbors() (in module net- copy() (MultiGraph method), 89
workx.classes.function), 499 core_number() (in module net-
communicability() (in module net- workx.algorithms.core), 282
workx.algorithms.communicability_alg), cost_of_flow() (in module net-
226 workx.algorithms.flow), 336
communicability_betweenness_centrality()could_be_isomorphic() (in module net-
(in module networkx.algorithms.centrality), workx.algorithms.isomorphism), 349
201 coverage() (in module net-
communicability_exp() (in module net- workx.algorithms.community.quality), 234
workx.algorithms.communicability_alg), create_empty_copy() (in module net-
227 workx.classes.function), 492
complement() (in module net- create_random_state() (in module net-
workx.algorithms.operators.unary), 386 workx.utils.misc), 687
complete_bipartite_graph() (in module net- cubical_graph() (in module net-
workx.algorithms.bipartite.generators), 174 workx.generators.small), 524
complete_graph() (in module net- cumulative_distribution() (in module net-
workx.generators.classic), 512 workx.utils.random_sequence), 687
complete_multipartite_graph() (in module current_flow_betweenness_centrality()

716 Index
NetworkX Reference, Release 2.3rc1.dev20190222214247

(in module networkx.algorithms.centrality), degrees() (in module net-


196 workx.algorithms.bipartite.basic), 154
current_flow_betweenness_centrality_subset() dense_gnm_random_graph() (in module net-
(in module networkx.algorithms.centrality), workx.generators.random_graphs), 528
199 density() (in module net-
current_flow_closeness_centrality() (in workx.algorithms.bipartite.basic), 153
module networkx.algorithms.centrality), 191 density() (in module networkx.classes.function), 492
cut_size() (in module networkx.algorithms.cuts), desargues_graph() (in module net-
291 workx.generators.small), 524
cuthill_mckee_ordering() (in module net- descendants() (in module net-
workx.utils.rcm), 692 workx.algorithms.dag), 295
cycle_basis() (in module net- dfs_edges() (in module net-
workx.algorithms.cycles), 287 workx.algorithms.traversal.depth_first_search),
cycle_graph() (in module net- 460
workx.generators.classic), 514 dfs_labeled_edges() (in module net-
cytoscape_data() (in module net- workx.algorithms.traversal.depth_first_search),
workx.readwrite.json_graph), 646 464
cytoscape_graph() (in module net- dfs_postorder_nodes() (in module net-
workx.readwrite.json_graph), 646 workx.algorithms.traversal.depth_first_search),
463
D dfs_predecessors() (in module net-
dag_longest_path() (in module net- workx.algorithms.traversal.depth_first_search),
workx.algorithms.dag), 299 461
dag_longest_path_length() (in module net- dfs_preorder_nodes() (in module net-
workx.algorithms.dag), 300 workx.algorithms.traversal.depth_first_search),
dag_to_branching() (in module net- 462
workx.algorithms.dag), 300 dfs_successors() (in module net-
davis_southern_women_graph() (in module net- workx.algorithms.traversal.depth_first_search),
workx.generators.social), 563 462
default_opener() (in module networkx.utils.misc), dfs_tree() (in module net-
686 workx.algorithms.traversal.depth_first_search),
degree (DiGraph attribute), 57 461
degree (Graph attribute), 28 diameter() (in module net-
degree (MultiDiGraph attribute), 118 workx.algorithms.distance_measures), 303
degree (MultiGraph attribute), 87 diamond_graph() (in module net-
degree() (in module networkx.classes.function), 491 workx.generators.small), 524
degree_assortativity_coefficient() (in dictionary, 695
module networkx.algorithms.assortativity), 141 difference() (in module net-
degree_centrality() (in module net- workx.algorithms.operators.binary), 389
workx.algorithms.bipartite.centrality), 171 DiGraph (class in networkx), 34
degree_centrality() (in module net- dijkstra_path() (in module net-
workx.algorithms.centrality), 182 workx.algorithms.shortest_paths.weighted),
degree_histogram() (in module net- 413
workx.classes.function), 492 dijkstra_path_length() (in module net-
degree_mixing_dict() (in module net- workx.algorithms.shortest_paths.weighted),
workx.algorithms.assortativity), 149 414
degree_mixing_matrix() (in module net- dijkstra_predecessor_and_distance()
workx.algorithms.assortativity), 148 (in module net-
degree_pearson_correlation_coefficient() workx.algorithms.shortest_paths.weighted),
(in module networkx.algorithms.assortativity), 412
143 dinitz() (in module networkx.algorithms.flow), 326
degree_sequence_tree() (in module net- directed_configuration_model() (in module
workx.generators.degree_seq), 544 networkx.generators.degree_seq), 541
directed_havel_hakimi_graph() (in module

Index 717
NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.generators.degree_seq), 543 eccentricity() (in module net-


directed_laplacian_matrix() (in module net- workx.algorithms.distance_measures), 303
workx.linalg.laplacianmatrix), 582 edge, 695
directed_modularity_matrix() (in module net- edge attribute, 695
workx.linalg.modularitymatrix), 592 edge_betweenness_centrality() (in module
discrete_sequence() (in module net- networkx.algorithms.centrality), 193
workx.utils.random_sequence), 688 edge_betweenness_centrality_subset() (in
disjoint_union() (in module net- module networkx.algorithms.centrality), 195
workx.algorithms.operators.binary), 388 edge_bfs() (in module net-
disjoint_union_all() (in module net- workx.algorithms.traversal.edgebfs), 470
workx.algorithms.operators.all), 390 edge_boundary() (in module net-
dispersion() (in module net- workx.algorithms.boundary), 179
workx.algorithms.centrality), 302 edge_connectivity() (in module net-
dodecahedral_graph() (in module net- workx.algorithms.connectivity.connectivity),
workx.generators.small), 524 267
dominance_frontiers() (in module net- edge_current_flow_betweenness_centrality()
workx.algorithms.dominance), 308 (in module networkx.algorithms.centrality),
dominating_set() (in module net- 197
workx.algorithms.dominating), 309 edge_current_flow_betweenness_centrality_subset()
dorogovtsev_goltsev_mendes_graph() (in (in module networkx.algorithms.centrality),
module networkx.generators.classic), 514 200
double_edge_swap() (in module net- edge_dfs() (in module net-
workx.algorithms.swap), 455 workx.algorithms.traversal.edgedfs), 469
draw() (in module networkx.drawing.nx_pylab), 662 edge_disjoint_paths() (in module net-
draw_circular() (in module net- workx.algorithms.connectivity.disjoint_paths),
workx.drawing.nx_pylab), 668 262
draw_kamada_kawai() (in module net- edge_expansion() (in module net-
workx.drawing.nx_pylab), 668 workx.algorithms.cuts), 292
draw_networkx() (in module net- edge_load_centrality() (in module net-
workx.drawing.nx_pylab), 662 workx.algorithms.centrality), 202
draw_networkx_edge_labels() (in module net- edge_subgraph() (DiGraph method), 64
workx.drawing.nx_pylab), 667 edge_subgraph() (Graph method), 33
draw_networkx_edges() (in module net- edge_subgraph() (in module net-
workx.drawing.nx_pylab), 665 workx.classes.function), 497
draw_networkx_labels() (in module net- edge_subgraph() (MultiDiGraph method), 125
workx.drawing.nx_pylab), 667 edge_subgraph() (MultiGraph method), 92
draw_networkx_nodes() (in module net- EdgeComponentAuxGraph (class in net-
workx.drawing.nx_pylab), 664 workx.algorithms.connectivity.edge_kcomponents),
draw_random() (in module net- 258
workx.drawing.nx_pylab), 669 edges (DiGraph attribute), 49
draw_shell() (in module net- edges (Graph attribute), 22
workx.drawing.nx_pylab), 669 edges (MultiDiGraph attribute), 109
draw_spectral() (in module net- edges (MultiGraph attribute), 81
workx.drawing.nx_pylab), 669 edges() (in module networkx.classes.function), 499
draw_spring() (in module net- Edmonds (class in net-
workx.drawing.nx_pylab), 669 workx.algorithms.tree.branchings), 477
dual_barabasi_albert_graph() (in module net- edmonds_karp() (in module net-
workx.generators.random_graphs), 533 workx.algorithms.flow), 320
duplication_divergence_graph() (in module effective_size() (in module net-
networkx.generators.duplication), 538 workx.algorithms.structuralholes), 453
efficiency() (in module net-
E workx.algorithms.efficiency), 310
ebunch, 695 ego_graph() (in module networkx.generators.ego),
560

718 Index
NetworkX Reference, Release 2.3rc1.dev20190222214247

eigenvector_centrality() (in module net- 434


workx.algorithms.centrality), 184 freeze() (in module networkx.classes.function), 506
eigenvector_centrality_numpy() (in module from_agraph() (in module net-
networkx.algorithms.centrality), 185 workx.drawing.nx_agraph), 670
empty_graph() (in module net- from_biadjacency_matrix() (in module net-
workx.generators.classic), 515 workx.algorithms.bipartite.matrix), 158
enumerate_all_cliques() (in module net- from_dict_of_dicts() (in module net-
workx.algorithms.clique), 214 workx.convert), 596
eppstein_matching() (in module net- from_dict_of_lists() (in module net-
workx.algorithms.bipartite.matching), 155 workx.convert), 597
erdos_renyi_graph() (in module net- from_edgelist() (in module networkx.convert), 598
workx.generators.random_graphs), 529 from_graph6_bytes() (in module net-
estrada_index() (in module net- workx.readwrite.graph6), 651
workx.algorithms.centrality), 204 from_nested_tuple() (in module net-
eulerian_circuit() (in module net- workx.algorithms.tree.coding), 477
workx.algorithms.euler), 312 from_numpy_array() (in module net-
eulerize() (in module networkx.algorithms.euler), workx.convert_matrix), 604
313 from_numpy_matrix() (in module net-
ExceededMaxIterations (class in networkx), 684 workx.convert_matrix), 602
expected_degree_graph() (in module net- from_pandas_adjacency() (in module net-
workx.generators.degree_seq), 542 workx.convert_matrix), 609
extended_barabasi_albert_graph() (in mod- from_pandas_edgelist() (in module net-
ule networkx.generators.random_graphs), 534 workx.convert_matrix), 610
extrema_bounding() (in module net- from_prufer_sequence() (in module net-
workx.algorithms.distance_measures), 303 workx.algorithms.tree.coding), 479
from_pydot() (in module net-
F workx.drawing.nx_pydot), 673
fast_could_be_isomorphic() (in module net- from_scipy_sparse_matrix() (in module net-
workx.algorithms.isomorphism), 349 workx.convert_matrix), 606
fast_gnp_random_graph() (in module net- from_sparse6_bytes() (in module net-
workx.generators.random_graphs), 527 workx.readwrite.sparse6), 654
faster_could_be_isomorphic() (in module net- frucht_graph() (in module net-
workx.algorithms.isomorphism), 349 workx.generators.small), 525
fiedler_vector() (in module net- full_rary_tree() (in module net-
workx.linalg.algebraicconnectivity), 585 workx.generators.classic), 516
find_cliques() (in module net-
workx.algorithms.clique), 214 G
find_cycle() (in module net- gaussian_random_partition_graph() (in
workx.algorithms.cycles), 288 module networkx.generators.community), 567
find_induced_nodes() (in module net- general_random_intersection_graph() (in
workx.algorithms.chordal), 212 module networkx.generators.intersection), 562
flatten() (in module networkx.utils.misc), 685 generalized_degree() (in module net-
florentine_families_graph() (in module net- workx.algorithms.cluster), 221
workx.generators.social), 563 generate_adjlist() (in module net-
flow_hierarchy() (in module net- workx.readwrite.adjlist), 620
workx.algorithms.hierarchy), 344 generate_edgelist() (in module net-
floyd_warshall() (in module net- workx.readwrite.edgelist), 628
workx.algorithms.shortest_paths.dense), generate_gexf() (in module net-
434 workx.readwrite.gexf ), 632
floyd_warshall_numpy() (in module net- generate_gml() (in module net-
workx.algorithms.shortest_paths.dense), 435 workx.readwrite.gml), 636
floyd_warshall_predecessor_and_distance() generate_graphml() (in module net-
(in module net- workx.readwrite.graphml), 641
workx.algorithms.shortest_paths.dense),

Index 719
NetworkX Reference, Release 2.3rc1.dev20190222214247

generate_multiline_adjlist() (in module net- google_matrix() (in module net-


workx.readwrite.multiline_adjlist), 623 workx.algorithms.link_analysis.pagerank_alg),
generate_pajek() (in module net- 364
workx.readwrite.pajek), 659 Graph (class in networkx), 7
generate_unique_node() (in module net- graph_atlas() (in module net-
workx.utils.misc), 686 workx.generators.atlas), 509
generic_edge_match() (in module net- graph_atlas_g() (in module net-
workx.algorithms.isomorphism), 359 workx.generators.atlas), 509
generic_graph_view() (in module net- graph_clique_number() (in module net-
workx.classes.graphviews), 127 workx.algorithms.clique), 216
generic_multiedge_match() (in module net- graph_edit_distance() (in module net-
workx.algorithms.isomorphism), 360 workx.algorithms.similarity), 438
generic_node_match() (in module net- graph_number_of_cliques() (in module net-
workx.algorithms.isomorphism), 359 workx.algorithms.clique), 217
generic_weighted_projected_graph() graphviz_layout() (in module net-
(in module net- workx.drawing.nx_agraph), 671
workx.algorithms.bipartite.projection), 163 graphviz_layout() (in module net-
geographical_threshold_graph() (in module workx.drawing.nx_pydot), 674
networkx.generators.geometric), 552 greedy_branching() (in module net-
get_data() (PlanarEmbedding method), 398 workx.algorithms.tree.branchings), 475
get_edge_attributes() (in module net- greedy_color() (in module net-
workx.classes.function), 506 workx.algorithms.coloring), 222
get_edge_data() (DiGraph method), 52 greedy_modularity_communities()
get_edge_data() (Graph method), 24 (in module net-
get_edge_data() (MultiDiGraph method), 113 workx.algorithms.community.modularity_max),
get_edge_data() (MultiGraph method), 83 232
get_node_attributes() (in module net- grid_2d_graph() (in module net-
workx.classes.function), 504 workx.generators.lattice), 520
girvan_newman() (in module net- grid_graph() (in module net-
workx.algorithms.community.centrality), workx.generators.lattice), 520
235 groups() (in module networkx.utils.misc), 686
global_efficiency() (in module net-
workx.algorithms.efficiency), 311 H
global_parameters() (in module net- hamiltonian_path() (in module net-
workx.algorithms.distance_regular), 307 workx.algorithms.tournament), 457
global_reaching_centrality() (in module net- harmonic_centrality() (in module net-
workx.algorithms.centrality), 207 workx.algorithms.centrality), 205
gn_graph() (in module networkx.generators.directed), harmonic_function() (in module net-
547 workx.algorithms.node_classification.hmn),
gnc_graph() (in module net- 384
workx.generators.directed), 548 has_bridges() (in module net-
gnm_random_graph() (in module net- workx.algorithms.bridges), 181
workx.generators.random_graphs), 529 has_edge() (DiGraph method), 51
gnmk_random_graph() (in module net- has_edge() (Graph method), 23
workx.algorithms.bipartite.generators), 177 has_edge() (MultiDiGraph method), 112
gnp_random_graph() (in module net- has_edge() (MultiGraph method), 82
workx.generators.random_graphs), 528 has_node() (DiGraph method), 48
gnr_graph() (in module net- has_node() (Graph method), 21
workx.generators.directed), 547 has_node() (MultiDiGraph method), 109
goldberg_radzik() (in module net- has_node() (MultiGraph method), 80
workx.algorithms.shortest_paths.weighted), has_path() (in module net-
432 workx.algorithms.shortest_paths.generic),
gomory_hu_tree() (in module net- 405
workx.algorithms.flow), 329 HasACycle (class in networkx), 683

720 Index
NetworkX Reference, Release 2.3rc1.dev20190222214247

hashable, 695 intersection() (in module net-


havel_hakimi_graph() (in module net- workx.algorithms.operators.binary), 388
workx.algorithms.bipartite.generators), 175 intersection_all() (in module net-
havel_hakimi_graph() (in module net- workx.algorithms.operators.all), 391
workx.generators.degree_seq), 543 intersection_array() (in module net-
heawood_graph() (in module net- workx.algorithms.distance_regular), 306
workx.generators.small), 525 inverse_line_graph() (in module net-
hexagonal_lattice_graph() (in module net- workx.generators.line), 559
workx.generators.lattice), 521 is_aperiodic() (in module net-
hits() (in module net- workx.algorithms.dag), 298
workx.algorithms.link_analysis.hits_alg), is_arborescence() (in module net-
365 workx.algorithms.tree.recognition), 474
hits_numpy() (in module net- is_attracting_component() (in module net-
workx.algorithms.link_analysis.hits_alg), workx.algorithms.components), 246
366 is_biconnected() (in module net-
hits_scipy() (in module net- workx.algorithms.components), 247
workx.algorithms.link_analysis.hits_alg), is_bipartite() (in module net-
366 workx.algorithms.bipartite.basic), 151
hoffman_singleton_graph() (in module net- is_bipartite_node_set() (in module net-
workx.generators.small), 525 workx.algorithms.bipartite.basic), 152
hopcroft_karp_matching() (in module net- is_branching() (in module net-
workx.algorithms.bipartite.matching), 156 workx.algorithms.tree.recognition), 474
house_graph() (in module net- is_chordal() (in module net-
workx.generators.small), 525 workx.algorithms.chordal), 211
house_x_graph() (in module net- is_connected() (in module net-
workx.generators.small), 525 workx.algorithms.components), 238
hub_matrix() (in module net- is_digraphical() (in module net-
workx.algorithms.link_analysis.hits_alg), workx.algorithms.graphical), 341
367 is_directed() (in module net-
hypercube_graph() (in module net- workx.classes.function), 493
workx.generators.lattice), 521 is_directed() (PlanarEmbedding method), 398
is_directed_acyclic_graph() (in module net-
I workx.algorithms.dag), 298
icosahedral_graph() (in module net- is_distance_regular() (in module net-
workx.generators.small), 525 workx.algorithms.distance_regular), 305
identified_nodes() (in module net- is_dominating_set() (in module net-
workx.algorithms.minors), 379 workx.algorithms.dominating), 309
immediate_dominators() (in module net- is_edge_cover() (in module net-
workx.algorithms.dominance), 307 workx.algorithms.covering), 286
in_degree (DiGraph attribute), 57 is_empty() (in module networkx.classes.function),
in_degree (MultiDiGraph attribute), 119 493
in_degree_centrality() (in module net- is_eulerian() (in module net-
workx.algorithms.centrality), 183 workx.algorithms.euler), 312
in_edges (DiGraph attribute), 51 is_forest() (in module net-
in_edges (MultiDiGraph attribute), 112 workx.algorithms.tree.recognition), 473
incidence_matrix() (in module net- is_frozen() (in module networkx.classes.function),
workx.linalg.graphmatrix), 580 507
induced_subgraph() (in module net- is_graphical() (in module net-
workx.classes.function), 495 workx.algorithms.graphical), 341
info() (in module networkx.classes.function), 492 is_isolate() (in module net-
information_centrality() (in module net- workx.algorithms.isolate), 346
workx.algorithms.centrality), 191 is_isomorphic() (DiGraphMatcher method), 354
initialize() (DiGraphMatcher method), 354 is_isomorphic() (GraphMatcher method), 352
initialize() (GraphMatcher method), 352 is_isomorphic() (in module net-

Index 721
NetworkX Reference, Release 2.3rc1.dev20190222214247

workx.algorithms.isomorphism), 347 workx.algorithms.components), 244


is_k_edge_connected() (in module net- is_weighted() (in module net-
workx.algorithms.connectivity.edge_augmentation), workx.classes.function), 502
254 isolates() (in module networkx.algorithms.isolate),
is_kl_connected() (in module net- 346
workx.algorithms.hybrid), 345 isomorphisms_iter() (DiGraphMatcher method),
is_list_of_ints() (in module net- 355
workx.utils.misc), 686 isomorphisms_iter() (GraphMatcher method),
is_locally_k_edge_connected() 353
(in module net- iterable() (in module networkx.utils.misc), 686
workx.algorithms.connectivity.edge_augmentation),
254 J
is_matching() (in module net- jaccard_coefficient() (in module net-
workx.algorithms.matching), 375 workx.algorithms.link_prediction), 368
is_maximal_matching() (in module net- jit_data() (in module net-
workx.algorithms.matching), 375 workx.readwrite.json_graph), 648
is_multigraphical() (in module net- jit_graph() (in module net-
workx.algorithms.graphical), 342 workx.readwrite.json_graph), 648
is_negatively_weighted() (in module net- johnson() (in module net-
workx.classes.function), 502 workx.algorithms.shortest_paths.weighted),
is_partition() (in module net- 433
workx.algorithms.community.community_utils), join() (in module net-
237 workx.algorithms.tree.operations), 481
is_perfect_matching() (in module net- joint_degree_graph() (in module net-
workx.algorithms.matching), 376 workx.generators.joint_degree_seq), 575
is_pseudographical() (in module net-
workx.algorithms.graphical), 342 K
is_reachable() (in module net- k_clique_communities() (in module net-
workx.algorithms.tournament), 458 workx.algorithms.community.kclique), 231
is_semiconnected() (in module net- k_components() (in module net-
workx.algorithms.components), 251 workx.algorithms.approximation.kcomponents),
is_simple_path() (in module net- 132
workx.algorithms.simple_paths), 447 k_components() (in module net-
is_string_like() (in module networkx.utils.misc), workx.algorithms.connectivity.kcomponents),
685 259
is_strongly_connected() (in module net- k_core() (in module networkx.algorithms.core), 283
workx.algorithms.components), 240 k_corona() (in module networkx.algorithms.core),
is_strongly_connected() (in module net- 285
workx.algorithms.tournament), 458 k_crust() (in module networkx.algorithms.core), 284
is_strongly_regular() (in module net- k_edge_augmentation() (in module net-
workx.algorithms.distance_regular), 305 workx.algorithms.connectivity.edge_augmentation),
is_tournament() (in module net- 252
workx.algorithms.tournament), 459 k_edge_components() (in module net-
is_tree() (in module net- workx.algorithms.connectivity.edge_kcomponents),
workx.algorithms.tree.recognition), 473 255
is_valid_degree_sequence_erdos_gallai() k_edge_subgraphs() (in module net-
(in module networkx.algorithms.graphical), workx.algorithms.connectivity.edge_kcomponents),
343 256
is_valid_degree_sequence_havel_hakimi() k_nearest_neighbors() (in module net-
(in module networkx.algorithms.graphical), workx.algorithms.assortativity), 146
342 k_random_intersection_graph() (in module
is_valid_joint_degree() (in module net- networkx.generators.intersection), 561
workx.generators.joint_degree_seq), 574 k_shell() (in module networkx.algorithms.core), 283
is_weakly_connected() (in module net-

722 Index
NetworkX Reference, Release 2.3rc1.dev20190222214247

kamada_kawai_layout() (in module net- load_centrality() (in module net-


workx.drawing.layout), 677 workx.algorithms.centrality), 202
karate_club_graph() (in module net- local_and_global_consistency() (in module
workx.generators.social), 562 networkx.algorithms.node_classification.lgc),
katz_centrality() (in module net- 385
workx.algorithms.centrality), 186 local_bridges() (in module net-
katz_centrality_numpy() (in module net- workx.algorithms.bridges), 182
workx.algorithms.centrality), 188 local_constraint() (in module net-
kernighan_lin_bisection() (in module net- workx.algorithms.structuralholes), 454
workx.algorithms.community.kernighan_lin), local_edge_connectivity() (in module net-
228 workx.algorithms.connectivity.connectivity),
kl_connected_subgraph() (in module net- 268
workx.algorithms.hybrid), 345 local_efficiency() (in module net-
kosaraju_strongly_connected_components() workx.algorithms.efficiency), 310
(in module networkx.algorithms.components), local_node_connectivity() (in module net-
242 workx.algorithms.approximation.connectivity),
krackhardt_kite_graph() (in module net- 130
workx.generators.small), 525 local_node_connectivity() (in module net-
workx.algorithms.connectivity.connectivity),
L 270
label_propagation_communities() local_reaching_centrality() (in module net-
(in module net- workx.algorithms.centrality), 206
workx.algorithms.community.label_propagation), lollipop_graph() (in module net-
233 workx.generators.classic), 516
ladder_graph() (in module net- lowest_common_ancestor() (in module net-
workx.generators.classic), 516 workx.algorithms.lowest_common_ancestors),
laplacian_matrix() (in module net- 374
workx.linalg.laplacianmatrix), 581
laplacian_spectrum() (in module net- M
workx.linalg.spectrum), 583 make_clique_bipartite() (in module net-
large_clique_size() (in module net- workx.algorithms.clique), 216
workx.algorithms.approximation.clique), make_max_clique_graph() (in module net-
134 workx.algorithms.clique), 215
latapy_clustering() (in module net- make_small_graph() (in module net-
workx.algorithms.bipartite.cluster), 168 workx.generators.small), 523
lattice_reference() (in module net- make_str() (in module networkx.utils.misc), 686
workx.algorithms.smallworld), 449 margulis_gabber_galil_graph() (in module
LCF_graph() (in module networkx.generators.small), networkx.generators.expanders), 519
523 match() (DiGraphMatcher method), 355
les_miserables_graph() (in module net- match() (GraphMatcher method), 353
workx.generators.social), 563 max_clique() (in module net-
lexicographic_product() (in module net- workx.algorithms.approximation.clique),
workx.algorithms.operators.product), 392 133
lexicographical_topological_sort() (in max_flow_min_cost() (in module net-
module networkx.algorithms.dag), 297 workx.algorithms.flow), 337
LFR_benchmark_graph() (in module net- max_weight_matching() (in module net-
workx.algorithms.community.community_generators), workx.algorithms.matching), 376
229 maximal_independent_set() (in module net-
line_graph() (in module networkx.generators.line), workx.algorithms.mis), 383
558 maximal_matching() (in module net-
literal_destringizer() (in module net- workx.algorithms.matching), 376
workx.readwrite.gml), 637 maximum_branching() (in module net-
literal_stringizer() (in module net- workx.algorithms.tree.branchings), 475
workx.readwrite.gml), 637

Index 723
NetworkX Reference, Release 2.3rc1.dev20190222214247

maximum_flow() (in module net- workx.algorithms.tree.mst), 482


workx.algorithms.flow), 314 minimum_st_edge_cut() (in module net-
maximum_flow_value() (in module net- workx.algorithms.connectivity.cuts), 276
workx.algorithms.flow), 315 minimum_st_node_cut() (in module net-
maximum_independent_set() (in module net- workx.algorithms.connectivity.cuts), 278
workx.algorithms.approximation.independent_set),mixing_dict() (in module net-
137 workx.algorithms.assortativity), 150
maximum_spanning_arborescence() (in mod- mixing_expansion() (in module net-
ule networkx.algorithms.tree.branchings), 476 workx.algorithms.cuts), 292
maximum_spanning_edges() (in module net- modularity_matrix() (in module net-
workx.algorithms.tree.mst), 485 workx.linalg.modularitymatrix), 591
maximum_spanning_tree() (in module net- modularity_spectrum() (in module net-
workx.algorithms.tree.mst), 483 workx.linalg.spectrum), 584
metric_closure() (in module net- moebius_kantor_graph() (in module net-
workx.algorithms.approximation.steinertree), workx.generators.small), 525
138 multi_source_dijkstra() (in module net-
min_cost_flow() (in module net- workx.algorithms.shortest_paths.weighted),
workx.algorithms.flow), 335 418
min_cost_flow_cost() (in module net- multi_source_dijkstra_path() (in module
workx.algorithms.flow), 334 networkx.algorithms.shortest_paths.weighted),
min_edge_cover() (in module net- 419
workx.algorithms.bipartite.covering), 178 multi_source_dijkstra_path_length()
min_edge_cover() (in module net- (in module net-
workx.algorithms.covering), 285 workx.algorithms.shortest_paths.weighted),
min_edge_dominating_set() (in module net- 420
workx.algorithms.approximation.dominating_set),MultiDiGraph (class in networkx), 93
136 MultiGraph (class in networkx), 65
min_maximal_matching() (in module net- mycielski_graph() (in module net-
workx.algorithms.approximation.matching), workx.generators.mycielski), 577
137 mycielskian() (in module net-
min_weighted_dominating_set() workx.generators.mycielski), 576
(in module net-
workx.algorithms.approximation.dominating_set),N
135 navigable_small_world_graph() (in module
min_weighted_vertex_cover() (in module net- networkx.generators.geometric), 555
workx.algorithms.approximation.vertex_cover), nbunch, 695
140 nbunch_iter() (DiGraph method), 55
minimum_branching() (in module net- nbunch_iter() (Graph method), 26
workx.algorithms.tree.branchings), 476 nbunch_iter() (MultiDiGraph method), 116
minimum_cut() (in module net- nbunch_iter() (MultiGraph method), 85
workx.algorithms.flow), 317 negative_edge_cycle() (in module net-
minimum_cut_value() (in module net- workx.algorithms.shortest_paths.weighted),
workx.algorithms.flow), 319 431
minimum_cycle_basis() (in module net- neighbors() (DiGraph method), 52
workx.algorithms.cycles), 289 neighbors() (Graph method), 24
minimum_edge_cut() (in module net- neighbors() (in module networkx.classes.function),
workx.algorithms.connectivity.cuts), 274 498
minimum_node_cut() (in module net- neighbors() (MultiDiGraph method), 114
workx.algorithms.connectivity.cuts), 275 neighbors() (MultiGraph method), 83
minimum_spanning_arborescence() (in mod- neighbors_cw_order() (PlanarEmbedding
ule networkx.algorithms.tree.branchings), 476 method), 398
minimum_spanning_edges() (in module net- network_simplex() (in module net-
workx.algorithms.tree.mst), 483 workx.algorithms.flow), 332
minimum_spanning_tree() (in module net-

724 Index
NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.approximation (mod- networkx.algorithms.clique (module), 213


ule), 129 networkx.algorithms.cluster (module), 217
networkx.algorithms.approximation.cliquenetworkx.algorithms.coloring (module), 222
(module), 133 networkx.algorithms.communicability_alg
(module), 226
networkx.algorithms.approximation.clustering_coefficient
(module), 135 networkx.algorithms.community (module),
networkx.algorithms.approximation.connectivity 227
(module), 129 networkx.algorithms.community.asyn_fluid
networkx.algorithms.approximation.dominating_set (module), 233
(module), 135 networkx.algorithms.community.centrality
networkx.algorithms.approximation.independent_set (module), 235
(module), 136 networkx.algorithms.community.community_generators
networkx.algorithms.approximation.kcomponents (module), 228
(module), 132 networkx.algorithms.community.community_utils
networkx.algorithms.approximation.matching (module), 237
(module), 137 networkx.algorithms.community.kclique
networkx.algorithms.approximation.ramsey (module), 231
(module), 138 networkx.algorithms.community.kernighan_lin
networkx.algorithms.approximation.steinertree (module), 228
(module), 138 networkx.algorithms.community.label_propagation
networkx.algorithms.approximation.treewidth (module), 232
(module), 139 networkx.algorithms.community.modularity_max
networkx.algorithms.approximation.vertex_cover (module), 231
(module), 140 networkx.algorithms.community.quality
networkx.algorithms.assortativity (mod- (module), 234
ule), 140 networkx.algorithms.components (module),
networkx.algorithms.bipartite (module), 238
150 networkx.algorithms.connectivity (mod-
networkx.algorithms.bipartite.basic ule), 252
(module), 151 networkx.algorithms.connectivity.connectivity
networkx.algorithms.bipartite.centrality (module), 265
(module), 170 networkx.algorithms.connectivity.cuts
networkx.algorithms.bipartite.cluster (module), 273
(module), 165 networkx.algorithms.connectivity.disjoint_paths
networkx.algorithms.bipartite.covering (module), 262
(module), 178 networkx.algorithms.connectivity.edge_augmentation
networkx.algorithms.bipartite.generators (module), 252
(module), 173 networkx.algorithms.connectivity.edge_kcomponents
networkx.algorithms.bipartite.matching (module), 255
(module), 155 networkx.algorithms.connectivity.kcomponents
networkx.algorithms.bipartite.matrix (module), 259
(module), 157 networkx.algorithms.connectivity.kcutsets
networkx.algorithms.bipartite.projection (module), 260
(module), 159 networkx.algorithms.connectivity.stoerwagner
networkx.algorithms.bipartite.redundancy (module), 280
(module), 169 networkx.algorithms.connectivity.utils
networkx.algorithms.bipartite.spectral (module), 281
(module), 165 networkx.algorithms.core (module), 282
networkx.algorithms.boundary (module), 178 networkx.algorithms.covering (module), 285
networkx.algorithms.bridges (module), 180 networkx.algorithms.cuts (module), 290
networkx.algorithms.centrality (module), networkx.algorithms.cycles (module), 286
182, 301 networkx.algorithms.dag (module), 294
networkx.algorithms.chains (module), 210 networkx.algorithms.distance_measures
networkx.algorithms.chordal (module), 211 (module), 302

Index 725
NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.algorithms.distance_regular networkx.algorithms.shortest_paths.generic
(module), 305 (module), 401
networkx.algorithms.dominance (module), networkx.algorithms.shortest_paths.unweighted
307 (module), 406
networkx.algorithms.dominating (module), networkx.algorithms.shortest_paths.weighted
308 (module), 411
networkx.algorithms.efficiency (module), networkx.algorithms.similarity (module),
310 437
networkx.algorithms.euler (module), 311 networkx.algorithms.simple_paths (mod-
networkx.algorithms.flow (module), 313 ule), 444
networkx.algorithms.graphical (module), networkx.algorithms.smallworld (module),
340 448
networkx.algorithms.hierarchy (module), networkx.algorithms.smetric (module), 451
344 networkx.algorithms.sparsifiers (module),
networkx.algorithms.hybrid (module), 344 452
networkx.algorithms.isolate (module), 346 networkx.algorithms.structuralholes
networkx.algorithms.isomorphism (module), (module), 453
347 networkx.algorithms.swap (module), 455
networkx.algorithms.isomorphism.isomorphvf2 networkx.algorithms.tournament (module),
(module), 350 457
networkx.algorithms.link_analysis.hits_alg networkx.algorithms.traversal.beamsearch
(module), 365 (module), 468
networkx.algorithms.link_analysis.pagerank_algnetworkx.algorithms.traversal.breadth_first_search
(module), 361 (module), 465
networkx.algorithms.link_prediction networkx.algorithms.traversal.depth_first_search
(module), 367 (module), 460
networkx.algorithms.lowest_common_ancestors networkx.algorithms.traversal.edgebfs
(module), 373 (module), 470
networkx.algorithms.matching (module), 375 networkx.algorithms.traversal.edgedfs
networkx.algorithms.minors (module), 377 (module), 469
networkx.algorithms.mis (module), 383 networkx.algorithms.tree.branchings
networkx.algorithms.node_classification (module), 474
(module), 384 networkx.algorithms.tree.coding (module),
networkx.algorithms.node_classification.hmn 477
(module), 384 networkx.algorithms.tree.mst (module), 482
networkx.algorithms.node_classification.lgc networkx.algorithms.tree.operations
(module), 385 (module), 481
networkx.algorithms.operators.all (mod- networkx.algorithms.tree.recognition
ule), 389 (module), 472
networkx.algorithms.operators.binary networkx.algorithms.triads (module), 486
(module), 387 networkx.algorithms.vitality (module), 487
networkx.algorithms.operators.product networkx.algorithms.voronoi (module), 488
(module), 391 networkx.algorithms.wiener (module), 489
networkx.algorithms.operators.unary networkx.classes.function (module), 491
(module), 386 networkx.classes.graphviews (module), 127
networkx.algorithms.planarity (module), networkx.classes.ordered (module), 126
395 networkx.convert (module), 595
networkx.algorithms.reciprocity (module), networkx.convert_matrix (module), 598
399 networkx.drawing.layout (module), 675
networkx.algorithms.richclub (module), 400 networkx.drawing.nx_agraph (module), 670
networkx.algorithms.shortest_paths.astarnetworkx.drawing.nx_pydot (module), 672
(module), 436 networkx.drawing.nx_pylab (module), 661
networkx.algorithms.shortest_paths.densenetworkx.exception (module), 683
(module), 434 networkx.generators.atlas (module), 509

726 Index
NetworkX Reference, Release 2.3rc1.dev20190222214247

networkx.generators.classic (module), 510 networkx.readwrite.multiline_adjlist


networkx.generators.community (module), (module), 620
563 networkx.readwrite.nx_shp (module), 659
networkx.generators.degree_seq (module), networkx.readwrite.nx_yaml (module), 649
539 networkx.readwrite.pajek (module), 657
networkx.generators.directed (module), 546 networkx.readwrite.sparse6 (module), 654
networkx.generators.duplication (module), networkx.relabel (module), 613
537 networkx.utils (module), 685
networkx.generators.ego (module), 559 networkx.utils.contextmanagers (module),
networkx.generators.expanders (module), 694
518 networkx.utils.decorators (module), 689
networkx.generators.geometric (module), networkx.utils.misc (module), 685
550 networkx.utils.random_sequence (module),
networkx.generators.intersection (mod- 687
ule), 561 networkx.utils.rcm (module), 692
networkx.generators.joint_degree_seq networkx.utils.union_find (module), 687
(module), 574 NetworkXAlgorithmError (class in networkx), 683
networkx.generators.lattice (module), 519 NetworkXError (class in networkx), 683
networkx.generators.line (module), 557 NetworkXException (class in networkx), 683
networkx.generators.mycielski (module), NetworkXNoCycle (class in networkx), 683
576 NetworkXNoPath (class in networkx), 683
networkx.generators.nonisomorphic_trees NetworkXNotImplemented (class in networkx), 683
(module), 573 NetworkXPointlessConcept (class in networkx),
networkx.generators.random_clustered 683
(module), 545 NetworkXUnbounded (class in networkx), 683
networkx.generators.random_graphs (mod- NetworkXUnfeasible (class in networkx), 683
ule), 526 new_edge_key() (MultiDiGraph method), 103
networkx.generators.small (module), 522 new_edge_key() (MultiGraph method), 74
networkx.generators.social (module), 562 newman_watts_strogatz_graph() (in module
networkx.generators.spectral_graph_forge networkx.generators.random_graphs), 530
(module), 570 next_face_half_edge() (PlanarEmbedding
networkx.generators.stochastic (module), method), 398
560 node, 695
networkx.generators.trees (module), 571 node attribute, 695
networkx.generators.triads (module), 574 node_boundary() (in module net-
networkx.linalg.algebraicconnectivity workx.algorithms.boundary), 179
(module), 584 node_clique_number() (in module net-
networkx.linalg.attrmatrix (module), 587 workx.algorithms.clique), 217
networkx.linalg.graphmatrix (module), 579 node_connected_component() (in module net-
networkx.linalg.laplacianmatrix (module), workx.algorithms.components), 239
580 node_connectivity() (in module net-
networkx.linalg.modularitymatrix (mod- workx.algorithms.approximation.connectivity),
ule), 591 131
networkx.linalg.spectrum (module), 583 node_connectivity() (in module net-
networkx.readwrite.adjlist (module), 617 workx.algorithms.connectivity.connectivity),
networkx.readwrite.edgelist (module), 624 272
networkx.readwrite.gexf (module), 630 node_disjoint_paths() (in module net-
networkx.readwrite.gml (module), 633 workx.algorithms.connectivity.disjoint_paths),
networkx.readwrite.gpickle (module), 638 264
networkx.readwrite.graph6 (module), 651 node_expansion() (in module net-
networkx.readwrite.graphml (module), 639 workx.algorithms.cuts), 293
networkx.readwrite.json_graph (module), node_link_data() (in module net-
643 workx.readwrite.json_graph), 643
networkx.readwrite.leda (module), 648

Index 727
NetworkX Reference, Release 2.3rc1.dev20190222214247

node_link_graph() (in module net- number_of_selfloops() (in module net-


workx.readwrite.json_graph), 644 workx.classes.function), 500
node_redundancy() (in module net- number_strongly_connected_components()
workx.algorithms.bipartite.redundancy), (in module networkx.algorithms.components),
169 241
NodeNotFound (class in networkx), 683 number_weakly_connected_components() (in
nodes (DiGraph attribute), 46 module networkx.algorithms.components), 244
nodes (Graph attribute), 20 numeric_assortativity_coefficient() (in
nodes (MultiDiGraph attribute), 107 module networkx.algorithms.assortativity), 142
nodes (MultiGraph attribute), 78 numeric_mixing_matrix() (in module net-
nodes() (in module networkx.classes.function), 498 workx.algorithms.assortativity), 148
nodes_or_number() (in module net- numerical_edge_match() (in module net-
workx.utils.decorators), 690 workx.algorithms.isomorphism), 358
nodes_with_selfloops() (in module net- numerical_multiedge_match() (in module net-
workx.classes.function), 501 workx.algorithms.isomorphism), 358
non_edges() (in module networkx.classes.function), numerical_node_match() (in module net-
499 workx.algorithms.isomorphism), 357
non_neighbors() (in module net-
workx.classes.function), 498 O
nonisomorphic_trees() (in module net- octahedral_graph() (in module net-
workx.generators.nonisomorphic_trees), workx.generators.small), 526
573 omega() (in module networkx.algorithms.smallworld),
normalized_cut_size() (in module net- 451
workx.algorithms.cuts), 293 open_file() (in module networkx.utils.decorators),
normalized_laplacian_matrix() (in module 689
networkx.linalg.laplacianmatrix), 581 optimal_edit_paths() (in module net-
not_implemented_for() (in module net- workx.algorithms.similarity), 439
workx.utils.decorators), 690 optimize_edit_paths() (in module net-
NotATree, 486 workx.algorithms.similarity), 443
null_graph() (in module net- optimize_graph_edit_distance() (in module
workx.generators.classic), 517 networkx.algorithms.similarity), 441
number_attracting_components() (in module order() (DiGraph method), 56
networkx.algorithms.components), 246 order() (Graph method), 27
number_connected_components() (in module order() (MultiDiGraph method), 117
networkx.algorithms.components), 238 order() (MultiGraph method), 86
number_of_cliques() (in module net- OrderedDiGraph (class in networkx), 126
workx.algorithms.clique), 217 OrderedGraph (class in networkx), 126
number_of_edges() (DiGraph method), 59 OrderedMultiDiGraph (class in networkx), 126
number_of_edges() (Graph method), 29 OrderedMultiGraph (class in networkx), 126
number_of_edges() (in module net- out_degree (DiGraph attribute), 58
workx.classes.function), 499 out_degree (MultiDiGraph attribute), 119
number_of_edges() (MultiDiGraph method), 121 out_degree_centrality() (in module net-
number_of_edges() (MultiGraph method), 88 workx.algorithms.centrality), 183
number_of_isolates() (in module net- out_edges (DiGraph attribute), 50
workx.algorithms.isolate), 347 out_edges (MultiDiGraph attribute), 111
number_of_nodes() (DiGraph method), 56 overall_reciprocity() (in module net-
number_of_nodes() (Graph method), 27 workx.algorithms.reciprocity), 400
number_of_nodes() (in module net- overlap_weighted_projected_graph()
workx.classes.function), 498 (in module net-
number_of_nodes() (MultiDiGraph method), 117 workx.algorithms.bipartite.projection), 162
number_of_nodes() (MultiGraph method), 86
number_of_nonisomorphic_trees() (in mod- P
ule networkx.generators.nonisomorphic_trees), pagerank() (in module net-
573 workx.algorithms.link_analysis.pagerank_alg),

728 Index
NetworkX Reference, Release 2.3rc1.dev20190222214247

361 preferential_attachment() (in module net-


pagerank_numpy() (in module net- workx.algorithms.link_prediction), 370
workx.algorithms.link_analysis.pagerank_alg), preferential_attachment_graph() (in mod-
362 ule networkx.algorithms.bipartite.generators),
pagerank_scipy() (in module net- 176
workx.algorithms.link_analysis.pagerank_alg), prefix_tree() (in module net-
363 workx.generators.trees), 572
pairwise() (in module networkx.utils.misc), 686 preflow_push() (in module net-
pappus_graph() (in module net- workx.algorithms.flow), 324
workx.generators.small), 526 preserve_random_state() (in module net-
parse_adjlist() (in module net- workx.utils.decorators), 691
workx.readwrite.adjlist), 619 projected_graph() (in module net-
parse_edgelist() (in module net- workx.algorithms.bipartite.projection), 159
workx.readwrite.edgelist), 629 pydot_layout() (in module net-
parse_gml() (in module networkx.readwrite.gml), workx.drawing.nx_pydot), 674
635 pygraphviz_layout() (in module net-
parse_graphml() (in module net- workx.drawing.nx_agraph), 672
workx.readwrite.graphml), 642
parse_leda() (in module networkx.readwrite.leda), Q
649 quotient_graph() (in module net-
parse_multiline_adjlist() (in module net- workx.algorithms.minors), 380
workx.readwrite.multiline_adjlist), 623
parse_pajek() (in module net- R
workx.readwrite.pajek), 658 ra_index_soundarajan_hopcroft() (in mod-
partial_duplication_graph() (in module net- ule networkx.algorithms.link_prediction), 371
workx.generators.duplication), 538 radius() (in module net-
path_graph() (in module net- workx.algorithms.distance_measures), 304
workx.generators.classic), 517 ramsey_R2() (in module net-
percolation_centrality() (in module net- workx.algorithms.approximation.ramsey),
workx.algorithms.centrality), 208 138
performance() (in module net- random_clustered_graph() (in module net-
workx.algorithms.community.quality), 235 workx.generators.random_clustered), 545
periphery() (in module net- random_degree_sequence_graph() (in module
workx.algorithms.distance_measures), 304 networkx.generators.degree_seq), 544
petersen_graph() (in module net- random_geometric_graph() (in module net-
workx.generators.small), 526 workx.generators.geometric), 550
PlanarEmbedding (class in net- random_graph() (in module net-
workx.algorithms.planarity), 396 workx.algorithms.bipartite.generators), 176
planted_partition_graph() (in module net- random_k_out_graph() (in module net-
workx.generators.community), 566 workx.generators.directed), 548
power() (in module net- random_kernel_graph() (in module net-
workx.algorithms.operators.product), 394 workx.generators.random_graphs), 535
PowerIterationFailedConvergence (class in random_layout() (in module net-
networkx), 684 workx.drawing.layout), 677
powerlaw_cluster_graph() (in module net- random_lobster() (in module net-
workx.generators.random_graphs), 534 workx.generators.random_graphs), 536
powerlaw_sequence() (in module net- random_partition_graph() (in module net-
workx.utils.random_sequence), 687 workx.generators.community), 565
pred (DiGraph attribute), 54 random_powerlaw_tree() (in module net-
predecessor() (in module net- workx.generators.random_graphs), 537
workx.algorithms.shortest_paths.unweighted), random_powerlaw_tree_sequence() (in mod-
410 ule networkx.generators.random_graphs), 537
predecessors() (DiGraph method), 54 random_reference() (in module net-
predecessors() (MultiDiGraph method), 116 workx.algorithms.smallworld), 449

Index 729
NetworkX Reference, Release 2.3rc1.dev20190222214247

random_regular_graph() (in module net- remove_edge() (Graph method), 16


workx.generators.random_graphs), 532 remove_edge() (MultiDiGraph method), 103
random_shell_graph() (in module net- remove_edge() (MultiGraph method), 75
workx.generators.random_graphs), 536 remove_edges_from() (DiGraph method), 43
random_state() (in module net- remove_edges_from() (Graph method), 17
workx.utils.decorators), 691 remove_edges_from() (MultiDiGraph method),
random_tournament() (in module net- 104
workx.algorithms.tournament), 459 remove_edges_from() (MultiGraph method), 75
random_tree() (in module net- remove_node() (DiGraph method), 40
workx.generators.trees), 571 remove_node() (Graph method), 13
random_weighted_sample() (in module net- remove_node() (MultiDiGraph method), 99
workx.utils.random_sequence), 688 remove_node() (MultiGraph method), 71
read_adjlist() (in module net- remove_nodes_from() (DiGraph method), 40
workx.readwrite.adjlist), 617 remove_nodes_from() (Graph method), 14
read_dot() (in module networkx.drawing.nx_agraph), remove_nodes_from() (MultiDiGraph method),
671 100
read_dot() (in module networkx.drawing.nx_pydot), remove_nodes_from() (MultiGraph method), 71
674 rescale_layout() (in module net-
read_edgelist() (in module net- workx.drawing.layout), 678
workx.readwrite.edgelist), 625 resource_allocation_index() (in module net-
read_gexf() (in module networkx.readwrite.gexf ), workx.algorithms.link_prediction), 368
631 restricted_view() (in module net-
read_gml() (in module networkx.readwrite.gml), 634 workx.classes.function), 496
read_gpickle() (in module net- reverse() (DiGraph method), 64
workx.readwrite.gpickle), 638 reverse() (in module net-
read_graph6() (in module net- workx.algorithms.operators.unary), 386
workx.readwrite.graph6), 652 reverse() (MultiDiGraph method), 126
read_graphml() (in module net- reverse_cuthill_mckee_ordering() (in mod-
workx.readwrite.graphml), 640 ule networkx.utils.rcm), 693
read_leda() (in module networkx.readwrite.leda), reverse_havel_hakimi_graph() (in module net-
649 workx.algorithms.bipartite.generators), 175
read_multiline_adjlist() (in module net- reverse_view() (in module net-
workx.readwrite.multiline_adjlist), 621 workx.classes.function), 497
read_pajek() (in module networkx.readwrite.pajek), reverse_view() (in module net-
657 workx.classes.graphviews), 127
read_shp() (in module networkx.readwrite.nx_shp), reversed() (in module net-
659 workx.utils.contextmanagers), 694
read_sparse6() (in module net- rich_club_coefficient() (in module net-
workx.readwrite.sparse6), 655 workx.algorithms.richclub), 400
read_weighted_edgelist() (in module net- ring_of_cliques() (in module net-
workx.readwrite.edgelist), 627 workx.generators.community), 568
read_yaml() (in module net- robins_alexander_clustering() (in module
workx.readwrite.nx_yaml), 650 networkx.algorithms.bipartite.cluster), 169
reciprocity() (in module net- rooted_product() (in module net-
workx.algorithms.reciprocity), 399 workx.algorithms.operators.product), 393
reconstruct_path() (in module net-
workx.algorithms.shortest_paths.dense), S
436 s_metric() (in module networkx.algorithms.smetric),
relabel_gexf_graph() (in module net- 451
workx.readwrite.gexf ), 633 scale_free_graph() (in module net-
relabel_nodes() (in module networkx.relabel), 613 workx.generators.directed), 549
relaxed_caveman_graph() (in module net- score_sequence() (in module net-
workx.generators.community), 565 workx.algorithms.tournament), 459
remove_edge() (DiGraph method), 43

730 Index
NetworkX Reference, Release 2.3rc1.dev20190222214247

second_order_centrality() (in module net- (in module net-


workx.algorithms.centrality), 209 workx.algorithms.shortest_paths.unweighted),
sedgewick_maze_graph() (in module net- 406
workx.generators.small), 526 single_source_shortest_path_length()
selfloop_edges() (in module net- (in module net-
workx.classes.function), 500 workx.algorithms.shortest_paths.unweighted),
semantic_feasibility() (DiGraphMatcher 407
method), 355 single_target_shortest_path()
semantic_feasibility() (GraphMatcher (in module net-
method), 353 workx.algorithms.shortest_paths.unweighted),
set_edge_attributes() (in module net- 407
workx.classes.function), 504 single_target_shortest_path_length()
set_node_attributes() (in module net- (in module net-
workx.classes.function), 503 workx.algorithms.shortest_paths.unweighted),
sets() (in module net- 408
workx.algorithms.bipartite.basic), 152 size() (DiGraph method), 59
shell_layout() (in module net- size() (Graph method), 28
workx.drawing.layout), 678 size() (MultiDiGraph method), 120
shortest_augmenting_path() (in module net- size() (MultiGraph method), 87
workx.algorithms.flow), 322 soft_random_geometric_graph() (in module
shortest_path() (in module net- networkx.generators.geometric), 551
workx.algorithms.shortest_paths.generic), spanner() (in module net-
401 workx.algorithms.sparsifiers), 452
shortest_path_length() (in module net- spectral_bipartivity() (in module net-
workx.algorithms.shortest_paths.generic), workx.algorithms.bipartite.spectral), 165
403 spectral_graph_forge() (in module net-
shortest_simple_paths() (in module net- workx.generators.spectral_graph_forge),
workx.algorithms.simple_paths), 447 570
sigma() (in module networkx.algorithms.smallworld), spectral_layout() (in module net-
450 workx.drawing.layout), 680
simple_cycles() (in module net- spectral_ordering() (in module net-
workx.algorithms.cycles), 287 workx.linalg.algebraicconnectivity), 586
single_source_bellman_ford() (in module spring_layout() (in module net-
networkx.algorithms.shortest_paths.weighted), workx.drawing.layout), 679
426 square_clustering() (in module net-
single_source_bellman_ford_path() workx.algorithms.cluster), 221
(in module net- star_graph() (in module net-
workx.algorithms.shortest_paths.weighted), workx.generators.classic), 517
427 steiner_tree() (in module net-
single_source_bellman_ford_path_length() workx.algorithms.approximation.steinertree),
(in module net- 138
workx.algorithms.shortest_paths.weighted), stochastic_block_model() (in module net-
428 workx.generators.community), 568
single_source_dijkstra() (in module net- stochastic_graph() (in module net-
workx.algorithms.shortest_paths.weighted), workx.generators.stochastic), 560
414 stoer_wagner() (in module net-
single_source_dijkstra_path() (in module workx.algorithms.connectivity.stoerwagner),
networkx.algorithms.shortest_paths.weighted), 280
416 strategy_connected_sequential() (in mod-
single_source_dijkstra_path_length() ule networkx.algorithms.coloring), 224
(in module net- strategy_connected_sequential_bfs() (in
workx.algorithms.shortest_paths.weighted), module networkx.algorithms.coloring), 224
417 strategy_connected_sequential_dfs() (in
single_source_shortest_path() module networkx.algorithms.coloring), 224

Index 731
NetworkX Reference, Release 2.3rc1.dev20190222214247

strategy_independent_set() (in module net- tetrahedral_graph() (in module net-


workx.algorithms.coloring), 225 workx.generators.small), 526
strategy_largest_first() (in module net- thresholded_random_geometric_graph() (in
workx.algorithms.coloring), 225 module networkx.generators.geometric), 556
strategy_random_sequential() (in module net- to_agraph() (in module net-
workx.algorithms.coloring), 225 workx.drawing.nx_agraph), 671
strategy_saturation_largest_first() (in to_dict_of_dicts() (in module networkx.convert),
module networkx.algorithms.coloring), 225 596
strategy_smallest_last() (in module net- to_dict_of_lists() (in module networkx.convert),
workx.algorithms.coloring), 225 597
strong_product() (in module net- to_directed() (DiGraph method), 62
workx.algorithms.operators.product), 393 to_directed() (Graph method), 32
strongly_connected_component_subgraphs()to_directed() (in module net-
(in module networkx.algorithms.components), workx.classes.function), 493
242 to_directed() (MultiDiGraph method), 124
strongly_connected_components() (in mod- to_directed() (MultiGraph method), 91
ule networkx.algorithms.components), 241 to_edgelist() (in module networkx.convert), 597
strongly_connected_components_recursive() to_graph6_bytes() (in module net-
(in module networkx.algorithms.components), workx.readwrite.graph6), 653
242 to_nested_tuple() (in module net-
subgraph() (DiGraph method), 63 workx.algorithms.tree.coding), 478
subgraph() (Graph method), 32 to_networkx_graph() (in module net-
subgraph() (in module networkx.classes.function), workx.convert), 595
495 to_numpy_array() (in module net-
subgraph() (MultiDiGraph method), 124 workx.convert_matrix), 600
subgraph() (MultiGraph method), 91 to_numpy_matrix() (in module net-
subgraph_centrality() (in module net- workx.convert_matrix), 599
workx.algorithms.centrality), 203 to_numpy_recarray() (in module net-
subgraph_centrality_exp() (in module net- workx.convert_matrix), 601
workx.algorithms.centrality), 204 to_pandas_adjacency() (in module net-
subgraph_is_isomorphic() (DiGraphMatcher workx.convert_matrix), 608
method), 355 to_pandas_edgelist() (in module net-
subgraph_is_isomorphic() (GraphMatcher workx.convert_matrix), 610
method), 353 to_prufer_sequence() (in module net-
subgraph_isomorphisms_iter() (DiGraph- workx.algorithms.tree.coding), 480
Matcher method), 355 to_pydot() (in module networkx.drawing.nx_pydot),
subgraph_isomorphisms_iter() (Graph- 673
Matcher method), 353 to_scipy_sparse_matrix() (in module net-
subgraph_view() (in module net- workx.convert_matrix), 605
workx.classes.graphviews), 127 to_sparse6_bytes() (in module net-
succ (DiGraph attribute), 54 workx.readwrite.sparse6), 655
succ (MultiDiGraph attribute), 115 to_undirected() (DiGraph method), 61
successors() (DiGraph method), 53 to_undirected() (Graph method), 31
successors() (MultiDiGraph method), 115 to_undirected() (in module net-
symmetric_difference() (in module net- workx.classes.function), 493
workx.algorithms.operators.binary), 389 to_undirected() (MultiDiGraph method), 123
syntactic_feasibility() (DiGraphMatcher to_undirected() (MultiGraph method), 90
method), 355 to_vertex_cover() (in module net-
syntactic_feasibility() (GraphMatcher workx.algorithms.bipartite.matching), 156
method), 353 topological_sort() (in module net-
workx.algorithms.dag), 295
T transitive_closure() (in module net-
tensor_product() (in module net- workx.algorithms.dag), 298
workx.algorithms.operators.product), 394 transitive_reduction() (in module net-

732 Index
NetworkX Reference, Release 2.3rc1.dev20190222214247

workx.algorithms.dag), 299 W
transitivity() (in module net- watts_strogatz_graph() (in module net-
workx.algorithms.cluster), 218 workx.generators.random_graphs), 531
traverse_face() (PlanarEmbedding method), 399 waxman_graph() (in module net-
tree_all_pairs_lowest_common_ancestor() workx.generators.geometric), 554
(in module net- weakly_connected_component_subgraphs()
workx.algorithms.lowest_common_ancestors), (in module networkx.algorithms.components),
374 245
tree_data() (in module net- weakly_connected_components() (in module
workx.readwrite.json_graph), 646 networkx.algorithms.components), 245
tree_graph() (in module net- weighted_choice() (in module net-
workx.readwrite.json_graph), 647 workx.utils.random_sequence), 689
treewidth_min_degree() (in module net- weighted_projected_graph() (in module net-
workx.algorithms.approximation.treewidth), workx.algorithms.bipartite.projection), 160
139 wheel_graph() (in module net-
treewidth_min_fill_in() (in module net- workx.generators.classic), 518
workx.algorithms.approximation.treewidth), wiener_index() (in module net-
139 workx.algorithms.wiener), 489
triad_graph() (in module net- windmill_graph() (in module net-
workx.generators.triads), 574 workx.generators.community), 569
triadic_census() (in module net- within_inter_cluster() (in module net-
workx.algorithms.triads), 486 workx.algorithms.link_prediction), 372
triangles() (in module net- write_adjlist() (in module net-
workx.algorithms.cluster), 218 workx.readwrite.adjlist), 618
triangular_lattice_graph() (in module net- write_dot() (in module net-
workx.generators.lattice), 522 workx.drawing.nx_agraph), 671
trivial_graph() (in module net- write_dot() (in module networkx.drawing.nx_pydot),
workx.generators.classic), 518 674
truncated_cube_graph() (in module net- write_edgelist() (in module net-
workx.generators.small), 526 workx.readwrite.edgelist), 626
truncated_tetrahedron_graph() (in module write_gexf() (in module networkx.readwrite.gexf ),
networkx.generators.small), 526 631
turan_graph() (in module net- write_gml() (in module networkx.readwrite.gml),
workx.generators.classic), 518 634
tutte_graph() (in module net- write_gpickle() (in module net-
workx.generators.small), 526 workx.readwrite.gpickle), 639
write_graph6() (in module net-
U workx.readwrite.graph6), 653
uniform_random_intersection_graph() (in write_graphml() (in module net-
module networkx.generators.intersection), 561 workx.readwrite.graphml), 641
union() (in module net- write_multiline_adjlist() (in module net-
workx.algorithms.operators.binary), 387 workx.readwrite.multiline_adjlist), 622
union() (UnionFind method), 687 write_pajek() (in module net-
union_all() (in module net- workx.readwrite.pajek), 658
workx.algorithms.operators.all), 390 write_shp() (in module networkx.readwrite.nx_shp),
update() (DiGraph method), 44 660
update() (Graph method), 17 write_sparse6() (in module net-
update() (MultiDiGraph method), 105 workx.readwrite.sparse6), 656
update() (MultiGraph method), 76 write_weighted_edgelist() (in module net-
workx.readwrite.edgelist), 627
V write_yaml() (in module net-
volume() (in module networkx.algorithms.cuts), 294 workx.readwrite.nx_yaml), 650
voronoi_cells() (in module net-
workx.algorithms.voronoi), 488

Index 733
NetworkX Reference, Release 2.3rc1.dev20190222214247

Z
zipf_rv() (in module net-
workx.utils.random_sequence), 688

734 Index

You might also like