ASSOSA UNIVERSITY

College of Computing and Informatics

DEPARTMENT OF COMPUTER SCIENCE

Fundamental Database System

Chapter 6: Relational Algebra and Calculus

DBMS- Fundamental Database System 1

 Relational Query Languages
 Query languages: Allow manipulation and retrieval of data from a
database.
 Relational model supports simple, powerful QLs:
 Strong formal foundation based on logic.
 Allows for much optimization.
 Query Languages != programming languages!
 QLs not expected to be “Turing complete”.
 QLs not intended to be used for complex calculations.
 QLs support easy, efficient access to large data sets.
06/22/2024 2
 Formal Query Languages
 What is the basis of Query Languages?
 Two formal Query Languages form the basis of “real” query
languages (e.g., SQL):
 Relational Algebra: Operational, it provides a recipe for
evaluating the query. Useful for representing execution
plans.
 Relational Calculus: Lets users describe what they want,
rather than how to compute it. (Nonoperational,
declarative.)

 Understanding Algebra & Calculus is key to understanding

SQL, query processing!

06/22/2024 3
 Preliminaries
 A query is applied to relation instances, and the result of a query is
also a relation instance.
 Schemas of input relations for a query are fixed (but query will run
regardless of instance!)
 The schema for the result of a given query is also fixed! Determined by
definition of query language constructs.

 Positional vs. named-field notation:

 Positional notation easier for formal definitions, named-field notation more
readable.
 Both used in Relational Algebra and SQL.
06/22/2024 4
 Example Instances
 “Sailors” and “Reserves”
relations for our examples.

 We’ll use “named field

notation”, which assumes that
names of fields in query results
are “inherited” from names of
fields in query input relations.

5 06/22/2024
 Relational Algebra
 Basic operations:
 
Selection ( ) Selects a subset of rows from relation.
 Projection ( ) Deletes unwanted columns from relation.


Cross-product ( ) Allows us to combine two relations.
 Set-difference 
( ) Tuples in reln. 1, but not in reln. 2.
 Union (  ) Tuples in reln. 1 and in reln. 2.
 Additional operations:
 Intersection, join, division, renaming: Not essential, but (very!)
useful.
 Since each operation returns a relation, operations can be
composed! (Algebra is “closed”.)

06/22/2024 6
 What is an “Algebra”

 Set of operands and operations that they are “closed” under all
compositions.
 Examples
 Boolean algebra - operands are the logical values True and
False, and operations include AND(), OR(), NOT(), etc.
 Integer algebra - operands are the set of integers, operands
include ADD(), SUB(), MUL(), NEG(), etc. many of which have
special in-fix operator symbols (+,-,*,-)
 In our case operands are relations, what are the operators?
06/22/2024 7
 Projection
 Deletes attributes that are not
sname rating
in projection list.
 Schema of result contains yuppy 9
exactly the fields in the lubber 8
projection list, with the same guppy 5
names that they had in the rusty 10
(only) input relation.  sname,rating(S2)
 Projection operator has to
eliminate duplicates! (Why??) age
 Note: real systems typically 35.0
don’t do duplicate elimination 55.5
unless the user explicitly asks
for it. (Why not?)  age(S2)

8 06/22/2024
 Selection
 Selects rows that satisfy
selection condition. sid sname rating age
 No duplicates in result! 28 yuppy 9 35.0
(Why?) 58 rusty 10 35.0
 Schema of result identical to  rating  8(S2)
schema of (only) input sname rating
relation. yuppy 9
 Result relation can be the rusty 10
input for another relational
algebra operation! (Operator  sname,rating( rating 8(S2))
composition.)
9 06/22/2024
 Union, Intersection, Set-Difference
 All of these operations take sid sname rating age
two input relations, which 22 dustin 7 45.0
must be union-compatible: 31 lubber 8 55.5
 Same number of fields. 58 rusty 10 35.0
 `Corresponding’ fields have 44 guppy 5 35.0
the same type. 28 yuppy 9 35.0
 What is the schema of result?
S1 S2
sid sname rating age
sid sname rating age
22 dustin 7 45.0
31 lubber 8 55.5
S1 S2 58 rusty 10 35.0

10 06/22/2024
 Cross-Product
 Each row of S1 is paired with each row of R1.
 Result schema has one field per field of S1 and R1, with
field names `inherited’ if possible.
 Conflict: Both S1 and R1 have a field called Sid.
(sid) sname rating age (sid) bid day
22 dustin 7 45.0 22 101 10/10/96
22 dustin 7 45.0 58 103 11/12/96
31 lubber 8 55.5 22 101 10/10/96
31 lubber 8 55.5 58 103 11/12/96
58 rusty 10 35.0 22 101 10/10/96
58 rusty 10 35.0 58 103 11/12/96

 Renaming operator:  (C(1 sid1, 5  sid2), S1 R1)

06/22/2024 11
 Joins
sname( Res rves) Sailors)
 bid103
Condition Join: R C S = C (R  S)

(sid) sname rating age (sid) bid day

22 dustin 7 45.0 58 103 11/12/96
31 lubber 8 55.5 58 103 11/12/96

S1 sname(bid103Res rves) SailoS1.sid

rs) < R1.sid R1

 Result schema same as that of cross-product.

 Fewer tuples than cross-product, might be able to compute more
efficiently
 Sometimes called a theta-join.

06/22/2024 12
Joins
 Equi-Join: A special case of condition join where the
condition c contains only equalities and ^.
sid sname rating age bid day
22 dustin 7 45.0 101 10/10/96
58 rusty 10 35.0 103 11/12/96

sname( Res rves) Sailors)

 S1 bid103 sid R1
 Result schema similar to cross-product, but only one copy of
fields for which equality is specified.
 Natural Join: Equijoin on all common fields.

06/22/2024 13
 Find names of sailors who’ve reserved boat #103

 Solution 1:  sname (( bid 103 Re serves)  Sailors)

 Solution 2:  (Temp1,  Re serves)

bid  103
 ( Temp2, Temp1  Sailors)
 sname (Temp2)

 Solution 3:  sname ( (Re serves  Sailors))

bid 103

06/22/2024 14
Find names of sailors who’ve reserved a
red boat
 Information about boat color only available in Boats; So, need
an extra join:

 sname (( Boats)  Re serves  Sailors)

color ' red '

 A more efficient solution:

 sname ( ((  Boats)  Re s)  Sailors)
sid bid color ' red '
 A query optimizer can find this given the first solution!

06/22/2024 15
Find sailors who’ve reserved a red or a
green boat
 Can identify all red or green boats, then find sailors
who’ve reserved one of these boats:
 (Tempboats, ( Boats))
color ' red '  color ' green '
 sname(Tempboats  Re serves  Sailors)

 Can also define Tempboats using union! (How?)

 What happens if is replaced by in this query?

06/22/2024 16
Find sailors who’ve reserved a red and a green boat
 Previous approach won’t work! Must identify sailors who’ve
reserved red boats, sailors who’ve reserved green boats, then find
the intersection (note that sid is a key for Sailors):

 (Tempred,  (( Boats)  Re serves))

sid color ' red '

 (Tempgreen,  (( Boats)  Re serves))

sid color ' green'

 sname((Tempred  Tempgreen)  Sailors)

06/22/2024 17
 Relational Calculus
 Comes in two flavors: Tuple relational calculus (TRC) and Domain
relational calculus (DRC).
 Calculus has variables, constants, comparison ops, logical connectives,
and quantifiers.
 TRC: Variables range over (i.e., get bound to) tuples.
 DRC: Variables range over domain elements (= field values).
 Both TRC and DRC are simple subsets of first-order logic.
 Expressions in the calculus are called formulas. An answer tuple is
essentially an assignment of constants to variables that make the formula
evaluate to true.
06/22/2024 18
Tuple Relational Calculus
 Query has the form: { T | p(T)}.
 Answer includes all tuples T that make the formula p(T) be
true.
 Formula is recursively defined, starting with simple atomic
formulas (getting tuples from relations or making comparisons
of values), and building bigger and better formulas using the
logical connectives.

06/22/2024 19
TRC Formulas
 Atomic formula:
 R  Rel, or R.a op S.b, or R.a op constant
 op is one of , , , , , 
 Formula:
 an atomic formula, or
  p, p  q, p  q
, where p and q are formulas, or
 X ( p( X )) , where variable X is free in p(X), or
  X ( p( X )) , where variable X is free in p(X)

06/22/2024 20
Free and Bound Variables
 The use of quantifiers X and X in a formula is said
to bind X.
 A variable that is not bound is free.

 Let us revisit the definition of a query: {T|p(T)}

 There is an important restriction: the variable T that appears to
the left of ‘|’ must be the only free variable in the formula p(...).

06/22/2024 21
Find all sailors with a rating above 7
 {S | S  Sailors ^ S.rating > 7}
 Query is evaluated on an instance of Sailors
 Tuple variable S is instantiated to each tuple of this
instance in turn, and the condition “S.rating > 7” is
applied to each such tuple.
 Answer contains all instances of S (which are tuples of
Sailors) satisfying the condition.

06/22/2024 22
Find sailors rated > 7 who’ve reserved boat #103
 {S | (S  Sailors) ^ (S.rating > 7) ^ ( R  Reserves
(R.sid = S.sid ^ R.bid = 103))}
 Note the use of  to find a tuple in Reserves that `joins
with’ the Sailors tuple under consideration.
 R is bound, S is not.

06/22/2024 23
Unsafe Queries, Expressive Power
 It is possible to write syntactically correct calculus
queries that have an infinite number of answers! Such
queries are called unsafe.
-E.g S |  S  Sailors
   
   
   
   
 

 It is known that every query that can be expressed in

relational algebra can be expressed as a safe query in DRC /
TRC; the converse is also true.
 Relational Completeness: Query language (e.g., SQL) can
express every query that is expressible in relational
algebra/calculus.

06/22/2024 24
 Summary
 The relational model has rigorously defined query languages
that are simple and powerful.
 Relational algebra is more operational; useful as internal
representation for query evaluation plans.
 Relational calculus is non-operational, and users define
queries in terms of what they want, not in terms of how to
compute it. (Declarativeness.)
 Several ways of expressing a given query; a query optimizer
should choose the most efficient version.
 Algebra and safe calculus have same expressive power,
leading to the notion of relational completeness.

06/22/2024 25
 Nested Relations
 Attributes can be scalar (as before) or relation-valued
 Definition is recursive
 Example:
create table Book (title: string, author: string,

copies: (publ: string,

pages: integer,
date: integer))
 “copies” is a relation-valued field

06/22/2024 26
Nested Relational Algebra
 A spectrum of algebras can be defined
 At one end:
 Relational algebra plus nest () and unnest ():
 If B =

06/22/2024 27
Nesting and Unnesting

 Nulls must be supported in algebra

  ( (B, copies), copies (publ, pages, date)) = B
 ,  inverses iff S  N
 S is set of scalar fields
 N is set of non-scalar fields
 This is called PNF (partitioned normal form)
06/22/2024 28
Extending Relational Operators
 At another end of spectrum:
 Selection allows set comparators and constants and use of
select, project inside the formula
 Projection allows arbitrary NF2 algebra expression in addition
to simple field names
 Union, difference: recursive definitions
 Cross product: usually just relational.
 Example: retrieve title, number of pages of all books by
Melville:
 [title, [pages](copies)]([author=‘Melville’](B))

06/22/2024 29
Nested Relations Summary
 An early step on the way to OODBMS
 No products, only prototypes, but:
 Many ideas from NF2 relations have survived
 Collection types in SQL3 (nesting, unnesting)
 Algebra ideas useful for Object Database QP
 Can provide a more natural model of data
 Are a straightforward extension of relations:
 many solutions are thus also straightforward
 formal foundation of relational model remains

06/22/2024 30
.

Thank You!

06/22/2024 31