Lecture Notes on Discrete Mathematics

A K Lal

S Pati

July 9, 2016

General Instructions
Throughout this book, an item with a label 2.1.3 means the 3-rd item of the 1-st section of
Chapter 2. While defining a new term or a new notation we shall use bold face letters. The
symbol := is used at places we are defining a term using equality symbol. A symbol !! at the
end of a statement reminds the reader to verify the statement by writing a proof, if necessary.
We assume that the reader is familiar with the very basic of counting. A reader who is not,
may avoid the counting items in the initial parts till we start to discuss counting.
We also assume that the reader is familiar with some very basic definitions involving sets.
This book is written with the primary purpose of making the reader understand the discussion.
We do not intend to write elaborate proofs for the reader to read, as there is no end to elaboration.
We request the reader to take each statement in the book with the best possible natural meaning.
Here are a few collected quotes, mainly intended to inspire the authors.
Albert Einstein

The value of a college education is not the learning of many facts but the training of

the mind to think.

Imagination is more important than knowledge. For knowledge is limited, whereas

imagination embraces the entire world, stimulating progress, giving birth to evolution.
It is, strictly speaking, a real factor in scientific research.

Everything should be made as simple as possible, but no simpler.

Do not worry about your difficulties in Mathematics. I can assure you mine are still

greater.

Contents
1 Basic Set Theory

1.1

Common notations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.2

preliminaries . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

1.3

More on principle of mathematical induction . . . . . . . . . . . . . . . . . . . .

12

1.4

Integers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

16

2 Advanced topics in set theory

29

2.1

Families of Sets . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

29

2.2

More on Relations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

30

2.3

More on functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

35

2.4

Supplying bijections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

38

3 Countability, cardinal numbers* and partial order

43

3.1

Countable-uncountable . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

43

3.2

Partial orders . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

46

4 Logic

57

4.1

Propositional logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

57

4.2

Predicate logic . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

71

5 Lattices and Boolean Algebra

77

5.1

Lattices . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

77

5.2

Boolean Algebras . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

83

6 Counting
6.1

89

Permutations and combinations . . . . . . . . . . . . . . . . . . . . . . . . . . . .

89

6.1.1

Multinomial theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

96

6.2

Circular permutations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

97

6.3

Solutions in nonnegative integers . . . . . . . . . . . . . . . . . . . . . . . . . . . 102

6.4

Set partitions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105

6.5

Lattice paths and Catalan numbers . . . . . . . . . . . . . . . . . . . . . . . . . . 110

6.6

Some Generalizations . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 112

3

CONTENTS

7 Advanced Counting Principles

115

7.1

Pigeonhole Principle . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 115

7.2

Principle of Inclusion and Exclusion . . . . . . . . . . . . . . . . . . . . . . . . . 119

7.3

Generating functions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 122

7.4

Recurrence relation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 132

7.5

Generating function from recurrence relation . . . . . . . . . . . . . . . . . . . . 135

8 Graphs

145

8.1

Basic Concepts . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 145

8.2

Connectedness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 151

8.3

Isomorphism in graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 155

8.4

Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157

8.5

Connectivity

8.6

Eulerian Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 166

8.7

Hamiltonian Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 169

8.8

Bipartite Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 172

8.9

Matching . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173

. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164

8.10 Ramsey Numbers . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 176

8.11 Degree Sequence . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
8.12 Planar Graphs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 178
8.13 Vertex Coloring . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
8.14 Adjacency Matrix . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 182
8.15 More Exercises . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 183

Chapter 1

Basic Set Theory

1.1

Common notations

The following are some notations we shall follow throughout this document.
N
: the set of natural numbers
N0

the set of integers

the set of rational numbers

the set of real numbers

[n]

Ac

the set {1, 2, . . . , n}

P(A)

the power set of A

the cartesian product of A and B

the empty set

the integer p divides the integer a

AB
p|a

1.2

the set N {0}, called the set of whole numbers

the complement of a set A in some set that will be clear from the context

preliminaries

We expect the readers to have familiarity with the following definitions.

Definition 1.2.1. Let A and B be two sets.
1. [Subset of a set] If C is a set such that each element of C is also an element of A, then
C is said to be a subset of the set A, denoted C A.
2. [Equality of sets] The sets A and B are said to be equal if A B and B A, denoted
A = B.

3. [Cartesian product of sets] The cartesian product of A and B, denoted A B, is the

set of all ordered pairs (a, b), where a A and b B. Specifically, A B = {(a, b) | a
A, b B}.

4. [Set complement] Let C A. The complement of C in A, denoted C c , is a set that

contains every element of A that is not an element of C. Specifically, C c = {x A | x 6 C}.

5

CHAPTER 1. BASIC SET THEORY

5. [Set union] The union of A and B, denoted A B, is the set that exactly contains all the
elements of A and all the elements of B. Specifically, A B = {x | x A or x B}.

6. [Set intersection] The intersection of A and B, denoted A B, is the set that only

contains the common elements of A and B. Specifically, A B = {x | x A and x B}.

The set A and B are said to be disjoint if A B = .

7. [Set difference] The set difference of A and B, denoted A \ B, is a set that contains all
those elements of A which are not in B. Specifically, A \ B = {x A | x 6 B}.

8. [Symmetric difference] The symmetric difference of A and B, denoted AB, equals

(A \ B) (B \ A).
Example 1.2.2. Let A = {{b, c}, {{b}, {c}}, b} and B = {a, b, c}. Then
1. A B = {b},
2. A B = {a, b, c, {b, c}, {{b}, {c}} },
3. A \ B = {{b, c}, {{b}, {c}}},
4. B \ A = {a, c}, and
5. AB = {{b, c}, {{b}, {c}}, a, c}.
The following are a few well known facts. The readers are supposed to verify them for clarity.
Fact 1.2.3.

1. For any set A, we have A A and A.

2. If A B then it is not necessary that B A.

3. For any set A and B, the sets A \ B, A B and B \ A are pairwise disjoint. Thus, A B
is their disjoint union. That is,

A B = (A \ B) (A B) (B \ A).

(1.1)

4. For any set A and B, the sets A \ B and A B are disjoint. Thus, A is their disjoint union.
That is,

A = (A \ B) (A B)

(1.2)

5. If B A, then A = (A \ B) B (follows from (1.2)).

6. AB = (A B) \ (A B) (follows from (1.1) and Item 3).
Definition 1.2.4. [Power set] Let A be a set and B A. Then, the set that contains all

subsets of B is called the power set of B, denoted P(A).

Example 1.2.5.

1. Let A = . Then, P() = {, A} = {}.

2. Let A = {}. Then, P(A) = {, A} = {, {}}.

3. Let A = {a, b, c}. Then, P(A) = {, {a}, {b}, {c}, {a, b}, {a, c}, {b, c}, {a, b, c}}.

1.2. PRELIMINARIES

4. Let A = {{b, c}, {{b}, {c}}}. Then, P(A) = {, {{b, c}}, {{{b}, {c}}}, {{b, c}, {{b}, {c}}} }.
Definition 1.2.6. [Relation, domain set and codomain set] Let A and B be two sets. A
relation f from A to B, denoted f : A B, is a subset of A B. The set A is called the

domain set and the set B is called the codomain set. Thus, for any sets A and B the sets
and A B are always relations from A to B.

Example 1.2.7. Let A = [3], B = {a, b, c} and f = {(1, a), (1, b), (2, c)}. Then, f : A B is a

relation. We can draw a picture for f .1

c
f

Definition 1.2.8.

1. [Domain, range and inverse relation] Let f : A B be a relation.

Then, by the domain of f 2 we mean the set dom f := {a | (a, y) f } and by the range of

f we mean the set rng f := {b | (x, b) f }. The inverse of f is f 1 := {(y, x) | (x, y) f }.

Notice that f 1 is a relation from B to A. For example, the relation f in Example 1.2.7
has dom f = {1, 2}, rng f = {a, b, c} and f 1 = {(a, 1), (b, 1), (c, 2)}.

2. [Pre-image and image] Let f : A B be a relation and (x, y) f . We call x a pre-image

of y and y an image of x. Also, for any set X, we define f (X) := {y | (x, y) f, x X}.

Thus, f (X) = if X A = . We write f (x) to mean f ({x}). We write f (x) = y to mean

that f (x) = {y}. For example, consider the relation f in Example 1.2.7. Then,
(a) f (1) = {a, b}, f (2) = c and f (3) = .

(b) f 1 (c) = 2, f 1 (b) = 1, f 1 (1) = {}, f 1 (4) = {}.

(c) for X = {1, 4, c}, one has f (X) = {a, b} and f 1 (X) = {2}.

Definition 1.2.9. [Single valued relation and function] A relation f : A B is single

valued if f (x) is a singleton, for each a dom f . A function f from A to B is a single valued
relation such that dom f = A. Henceforth, for any function f , we assume that dom f 6= .
For example, the relation f in Example 1.2.7 is not single valued. However, if we delete (1, a)

from f , then it is single valued. Moreover, to make f a function from A to B, we need to

add either (3, a), or (3, b), or (3, c). In particular, the relations g1 = {(1, b), (2, c), (3, b)} and
g2 = {(1, b), (2, c), (3, a)} are indeed functions.

The following is an immediate consequence of the definition.

1
2

We use pictures to help our understanding and they are not parts of proof.
The domain set is the set from which we define our relations but dom f is the domain of the particular relation

f . They are different.

CHAPTER 1. BASIC SET THEORY

Proposition 1.2.10. Let f : A B be a relation and S be any set. Then,

1. f (S) 6= dom(f ) S 6= .
2. f 1 (S) 6= rng(f ) S 6= .
Proof. We will prove only one way implication. The other way is left for the reader.
Part 1: Since f (S) 6= , one can find a S A and b B such that (a, b) f . This, in turn,

implies that a dom(f ). As a S, a dom(f ) S.

Part 2: Since rng(f ) S 6= , one can find b rng(f ) S and a A such that (a, b) f . This,

in turn, implies that a f 1 (b) f 1 (S) as b S.

Definition 1.2.11. [One-one/Injection] A function f : A B is called one-one (also called

an injection), if |f 1 (b)| 1, for each b B. Equivalently, f : A B is one-one if f (x) 6= f (y)

is true, for each pair x 6= y in A. Equivalently, f is one-one if x = y is true, for each pair x, y A

whenever f (x) = f (y).

Convention:

Let p(x) be a polynomial in x with integer coefficients. Then, by writing f : Z Z is

a function defined by f (x) = p(x), we mean the function f = {(a, p(a)) | a Z}. For

example, the function f (x) = x2 stands for the set {(a, a2 ) | a Z}.
Example 1.2.12.

1. In Definition 1.2.9, the function g2 is one-one whereas g1 is not one-one.

2. The function f : Z Z defined by f (x) = x2 is not one-one.

3. The function f : N N0 defined by f (x) = x2 is one-one.
4. The function f : [3] {a, b, c, d} defined by f (1) = c, f (2) = b and f (3) = a, is one-one.
Verify that there are 24 one-one functions f : [3] {a, b, c, d}.

5. Let =
6 A ( B. Then, f (x) = x is a one-one map from A to B.
6. There is no one-one function from the set [3] to its proper subset [2].
7. There are one-one functions f from the set N to its proper subset {2, 3, . . .}. One of them
is given by f (1) = 3, f (2) = 2 and f (n) = n + 1, for n 3.

Definition 1.2.13. [Restriction function] Let f : X Y be a function and A X, A 6= .

Then, by fA we mean the restriction of f to A and denote it by fA . That is, fA = {(x, y) |

(x, y) f, x A}.

Example 1.2.14. Define f : R R as f (x) = 1 if x is irrational and f (x) = 0 if x is rational.

Then, fQ : Q R is the constant 0 function. That is, fQ (x) = 0, for all x = dom f = Q.

Proposition 1.2.15. Let f : A B be a one-one function and C be a nonempty subset of A.

Then, fC is also one-one.

Proof. Let if possible, fC (x) = fC (y), for some x, y C. Then, by definition of fC , we have

f (x) = f (y). As f is one-one, we get x = y. Thus, fC is one-one.

1.2. PRELIMINARIES

Definition 1.2.16. [Onto/Surjection] A function f : A B is called onto (also called a

surjection), if f 1 (b) 6= , for each b B. Equivalently, f : A B is onto if each z B has

some pre-image in A.
Example 1.2.17.

1. In Definition 1.2.9, the function g2 is onto whereas g1 is not onto.

2. There are 6 onto functions from [3] to [2]. For example, f (1) = 1, f (2) = 2 and f (3) = 2
is one such function.
3. Let =
6 A ( B. Choose a A. Then, g(y) =

B to A.

y
a

if y A,

if y B \ A.

is an onto map from

4. There is no onto function from the set [2] to its proper superset [3].
5. There are onto functions f from the set {2, 3, . . .} to its proper superset N. One of them
is f (x) = x 1.

Definition 1.2.18. [Bijection and equivalent set] Let A and B be two sets. A function
f : A B is said to be a bijection if f is one-one as well as onto. The sets A and B are said
to be equivalent if there exists a bijection f : A B.
Example 1.2.19.

1. In Definition 1.2.9, the function g2 is a bijection.

2. The function f : [3] {a, b, c} defined by f (1) = c, f (2) = b and f (3) = a, is a bijection.
Thus, the set {a, b, c} is equivalent to [3].

3. Let =
6 A A. Then, f (x) = x is a bijection. Thus, the set A is equivalent to itself.
4. If f : A B is a bijection then f 1 : B A is a bijection. Thus, if A is equivalent to B
then B is equivalent to A.

5. The set N is equivalent to {2, 3, . . .}. Indeed the function f : N {2, 3, . . .} defined by
f (1) = 3, f (2) = 2 and f (n) = n + 1, for n 3 is a bijection.

The following is known as the principle of mathematical induction in weak form.

Axiom 1.2.20. [Principle of mathematical induction (PMI): Weak form] Let S N be a
set which satisfies
1. 1 S and
2. k + 1 S whenever k S.
Then, S = N.1
Fact 1.2.21.

1. Let A, B and C be sets and let f : A B and g : B C be bijections.

Then, h : A C defined by h(x) = g(f (x)) is a bijection.

1

PMI is actually a part of Peano axioms that defines N as: a) 1 N. b) For each n N, the successor

s(n) N. c) 1 is not a successor of any natural number. d) If s(m) = s(n) happens for natural numbers m and
n, then m = n. e) Let S N such that 1 S and s(k) S, for each k S. Then, S = N.

10

CHAPTER 1. BASIC SET THEORY

Proof. h is one-one: Let if possible h(x) = h(y), for some x, y A. Then, by definition,
g(f (x)) = g(f (y)), for some f (x), f (y) B. As g is one-one, we get f (x) = f (y). Now,

using f is one-one, we get x = y and hence h is one-one.

h is onto: Let c C. Then, the condition that g is onto implies that there exists b B

such that g(b) = c. Also, for b B, the condition that f is onto implies that there exists

a A such that f (a) = b. Thus, we see that h(a) = g(f (a)) = g(b) = c and hence the

required result follows.

2. Let A and B be two disjoint sets and let f : A [n] and g : (

B [m] be two bijections.
f (x)
if x A
is a
Then, the function h : A B [m + n] defined by h(x) =
g(x) + n if x B
bijection.
3. Fix n 2 and let f : A [n] be a bijection such that for(a fixed element a A, one has
f (x)
if f (x) k 1
is
f (a) = k. Then, g : A \ {a} [n 1] defined by g(x) =
f (x) 1 if f (x) k + 1
a bijection.
4. For any positive integers n and k, there is no bijection from [n] to [n + k].
Proof. Let us fix k and prove the result by induction on n. The result is clearly true for
n = 1 as k + 1 2. So, let the result be true for n. We need to prove it for n + 1. On
the contrary, assume that there exists a bijection f : [n + 1] [n + 1 + k]. Then, by

Fact 1.2.21.3, we get a bijection g : [n] [n + k], where a = n + 1. Thus, we arrive at a

contradiction to the induction assumption.

Definition 1.2.22. [Number of elements in a set] A set A is said to be finite if either A is

empty or A is equivalent to [n], for some natural number n. A set which is not finite is called
infinite. We say A has n elements or the number of elements in A is n to mean that A
is equivalent to [n]. We write |A| = n to mean that A has n elements. Conventionally, the

number of elements in an empty set is zero. If f : [n] A is a bijection, then A can be listed
as {a1 = f (1), . . . , an = f (n)}.
Fact 1.2.23.
m + n.

1. Let A and B be two disjoint sets with |A| = m and |B| = n. Then, |AB| =

Proof. Use Fact 1.2.21.2.

2. Any subset of [n] is finite.
Proof. We use PMI to prove this. It is true for n = 1. Let the result be true for [n 1].

Now, let S [n]. If n 6 S, then S [n 1] and hence using PMI the result follows. If

n S, let T = S \ {n}. Then, by PMI, T is finite and hence by Fact 1.2.23.1, S is finite
as S is disjoint union of T and {n}.

3. Any subset of a finite set is finite.

Proof. Let |S| = n, for some n N. Then, there is a bijection f : S [n]. Let T S.

If T is empty then there is nothing to prove. Else, consider the map fT : T f (T ).

1.2. PRELIMINARIES

11

This map is a bijection. By Fact 1.2.23.2, f (T ) [n] is finite and hence T is finite (use
Fact 1.2.21.1).

4. Let A and B be two finite sets, then |A B| = |A| + |B| |A B|.

Proof. Using (1.1), A B = (A \ B) (A B) (B \ A). As the sets A \ B, A B and
B \ A are finite and pairwise disjoint, the result follows from Fact 1.2.23.1.

5. Let A be a nonempty finite set. Then, for any set B, |A| = |A \ B| + |A B|.
Proof. As A is finite, A \ B and A B are also finite. Now, use Fact 1.2.23.1.
6. Let A be a nonempty finite set and B A, then |B| |A|. In particular, if B ( A then
|B| < |A|.

Proof. Since B = A B, the result follows from Fact 1.2.23.5.

7. Let n and k be two fixed positive integers. Then, there is no one-one function from [n + k]
to [n].
Proof. Suppose there exists a one-one function f : [n + k] [n], for some n and k. Put

B = f ([n + k]) [n]. Then, notice that f : [n + k] B is a bijection and hence the sets B

and [n + k] are equivalent. Thus, by definition and Fact 1.2.23.6, n + k = |B| n < n + 1.

Or equivalently, k < 1 contradicting the assumption that k 1.

8. The set N is infinite.

Proof. Assume that the set N is finite and |N| = n, for some natural number n. Let

f : N [n] be a bijection. Then, f[n+1] is the restriction of f on [n + 1]. Thus, by

Proposition 1.2.15, f[n+1] is also one-one, contradicting Fact 1.2.23.7.

9. Let A be a finite nonempty set and x be a fixed symbol. Now, consider the set B =
{(x, a) | a A}. Then, |A| = |B|.
Proof. Define the function f : A B by f (a) = (x, a), for all a A. Then, f is a

bijection.

10. Let A be an infinite set and B A. Then, B is infinite.

Proof. If B is finite then by Fact 1.2.23.3, A is finite. A contradiction to A being an
infinite set.
11. Let A be an infinite set and B be a finite set. Then, A \ B is also infinite. In particular,
if a A, then A \ {a} is also infinite.

Proof. If A \ B is finite, then by Fact 1.2.23.1, the set (A \ B) B is also finite. But

A (A \ B) B and hence by Fact 1.2.23.3, A is finite as well. A contradiction to A being

an infinite set.

12. A set A is infinite if and only if there is a one-one function f : N A.

Proof. Let A be infinite. So, A 6= . Let a1 A. Put f (1) = a1 and A1 = A \ {a1 }. By
Fact 1.2.23.11, A1 is infinite. Assume that we have defined f (1), . . . , f (k) and obtained

Ak = Ak1 \ {ak }. As Ak1 was infinite, by Fact 1.2.23.11, Ak is infinite. Hence, Ak 6= .

12

CHAPTER 1. BASIC SET THEORY

Let ak+1 Ak . Define f (k + 1) = ak+1 and Ak+1 = Ak \ {ak+1 }. By applying induction, f
gets defined on N. Notice that by construction ak+1
/ {a1 , . . . , ak }. Hence, f is one-one.

Conversely, let f : N A be one-one. Then, f : N f (N) is a bijection. If f (N) is finite

then N is finite as well, contradicting Fact 1.2.23.8. Hence, f (N) is infinite. As f (N) A,
using Fact 1.2.23.10, A is infinite as well.

13. A set is infinite if and only if it is equivalent to a proper subset of itself.

Proof. Let S be an infinite set. Then, by Fact 1.2.23.12, there
is a one-one function
(
x,
if x 6 f (N)
f : N S. Now define a map g : S S \ {f (1)} by g(x) =
f (k + 1), if x = f (k).
Then, g is indeed a bijection. Thus, S is equivalent to its proper subset S \ {f (1)}.
Conversely, let S be a set and T a proper subset of S such that there is a bijection
f : S T . Suppose that S is finite and let |S| = n. Then, by Fact 1.2.23.6, T is also

finite and |T | = m < n. On the other hand, by the assumption that S is finite and there

is a bijection from S to T , we have m = n, a contradiction.

Exercise 1.2.24. [Optional]

1. Do there exist unique sets X and Y such that X \ Y = {1, 3, 5, 7} and Y \ X = {2, 4, 8}?
2. In a class of 60 students, all the students play either football or cricket. If 20 students play
both football and cricket, determine the number of players for each game if the number of
students who play football is
(a) 14 more than the number of students who play cricket.
(b) exactly 5 times the number of students who play only cricket.
(c) a multiple of 2 and 3 and leaves a remainder 3 when divided by 5.
(d) is a factor of 90 and the number of students who play cricket is a factor of 70.

1.3

More on principle of mathematical induction

The following is known as the principle of mathematical induction in strong form.

Theorem 1.3.1. [Principle of mathematical induction (PMI): Strong form] Let S N be a

set which satisfies

1. 1 S and

2. k + 1 S whenever [k] S holds.

Then, S = N.
Proof. Define T = {k S | [k] S}. Then, 1 T as 1 S and [1] S. Now, suppose

k T . Then, by definition [k] S. Therefore, the hypothesis implies that k + 1 S and hence
[k + 1] = [k] {k + 1} S. Thus, k + 1 T . Hence, by using the weak form of PMI on T , we

conclude that T = N, which in turn implies that S = N.

1.3. MORE ON PRINCIPLE OF MATHEMATICAL INDUCTION

13

Theorem 1.3.2. [Another form of PMI] Let S Z be a set which satisfies

1. k0 S and
2. k + 1 S whenever {k0 , k0 + 1, . . . , k} S.
Then, {k0 , k0 + 1, . . .} S.
Proof. Consider T = {x (k0 1) | x S, x k0 }. Then, 1 T as k0 S and 1 =

k0 (k0 1). Now, let [k] T . Then, {k0 , k0 + 1, . . . , k0 + k 1} S. Hence, by the hypothesis,

(k0 + k 1) + 1 = k0 + k S. Therefore, by definition of T , we have k + 1 T and hence using

the strong form of PMI, T = N. Thus, the required result follows.

The next result is commonly known as the Well-Ordering Principle which states that every
nonempty subset of natural numbers contains its least element.
Theorem 1.3.3. [Application of PMI in strong form: A nonempty subset of N contains its
minimum] Let =
6 A N. Then, the least element of A is a member of A.

Proof. For each fixed positive integer k, let P (k) mean the statement each nonempty subset A
of N that contains k, also contains its minimum.
Notice that P (1) is true. Now, assume that P (1), . . . , P (k) are true. We need to show that
P (k + 1) is true as well. Hence, consider a set A such that k + 1 A. If {1, . . . , k} A = ,

then k + 1 = min A, we are done. If r {1, . . . , k} A, then r k and hence by induction

hypothesis, P (r) is true. So, P (k + 1) is true. Hence, by the strong form of PMI the required
result follows.
By using Theorem 1.3.2, we can also prove the following generalization of Theorem 1.3.3. The

proof is similar to the proof of Theorem 1.3.3 and is left to the reader.
Theorem 1.3.4. [Well-ordering principle] Fix k Z. Let A be a nonempty subset of {k, k +

1, . . .}. Then, A contains its minimum element.

Theorem 1.3.5. [Archimedean property for positive integers] Let x, y N. Then, there
exists n N such that nx y.

Proof. On the contrary assume that such an n N does not exist. That is, nx < y for every

n N. That is, y nx N, for all n N. Now, consider the set S = {y nx | n N0 }.

Then, y S and hence S is a nonempty subset of N0 . Therefore, by the well-ordering principle

(Theorem 1.3.4), S contains its least element, say y mx. Then, by assumption, the integer

y (m + 1)x 0, y (m + 1)x S and y (m + 1)x < y mx. A contradicts to the minimality

of y mx. Thus, our assumption is invalid and hence the required result follows.

The next result gives the equivalence of the weak form of PMI with the strong form of PMI.

Theorem 1.3.6. [Equivalence of PMI in weak form and PMI in strong form] Fix a natural
number k0 and let P (n) be a statement about a natural number n. Suppose that P means the
statement P (n) is true, for each n N, n k0 . Then, P can be proved using the weak form
of PMI if and only if P can be proved using the strong form of PMI.

14

CHAPTER 1. BASIC SET THEORY

Proof. Let us assume that the statement P has been proved using the weak form of PMI. Hence,
P (k0 ) is true. Further, whenever P (n) is true, we are able to establish that P (n + 1) is true.
Therefore, we can establish that P (n + 1) is true if P (k0 ), . . . , P (n) are true. Hence, P can be
proved using the strong form of PMI.
So, now let us assume that the statement P has been proved using the strong form of PMI.
Now, define Q(n) to mean P () holds for = k0 , k0 + 1, . . . , n. Notice that Q(k0 ) is true.
Suppose that Q(n) is true (this means that P () is true for = k0 , k0 + 1, . . . , n). By hypothesis,
we know that P has been proved using the strong form of PMI. That is, P (n+1) is true whenever
P () is true for = k0 , k0 + 1, . . . , n. This, in turn, means that Q(n + 1) is true. Hence, by the
weak form of PMI, Q(n) is true for all n k0 . Thus, we are able to prove P using the weak

form of PMI.

Theorem 1.3.7. [Optional: Application of PMI in weak form: AM-GM inequality] Fix a
positive integer n and let a1 , a2 , . . . , an be non-negative real numbers. Then
Arithmetic Mean (AM) :=

a1 + + an
n a1 an =: (GM) Geometric Mean.
n

Proof. The inequality clearly holds for n = 1 and 2. Assume that it holds for every choice of n
non-negative real numbers. Now, let a1 , . . . , an , an+1 be a set of n + 1 non-negative real numbers
with a1 = max{a1 , . . . , an+1 } and an+1 = min{a1 , . . . , an+1 }. Define A =

a1 +a2 ++an+1
.
n+1

Then,

note that a1 A an+1 . Hence, (a1 A)(A an+1 ) 0, i.e., A(a1 + an+1 A) a1 an+1 .

Now, apply induction hypothesis on the n non-negative real numbers a2 , . . . , an , a1 + an+1 A

to get

p
n

a2 + + an + (a1 + an+1 A)
= A.
n
(a2 a3 an (a1 + an+1 A))A (a2 a3 an ) a1 an+1 . Therefore,

a2 an (a1 + an+1 A)

So, we have An+1

by PMI, the inequality holds, for each n N.

In the next example, we illustrate the use of PMI to establish some given identities (properties,

statements) involving natural numbers.

n(n + 1)
.
2
Ans: The result is clearly true for n = 1. So, let us assume that 1 + 2 + + n =

Example 1.3.8. Prove that 1 + 2 + + n =

Then, using the induction hypothesis, we have
1 + 2 + + n + (n + 1) =

n(n + 1)
n+1
+ (n + 1) =
(n + 2) .
2
2

Thus, the result holds for n + 1 and hence by the weak form of PMI, the result follows.
Exercise 1.3.9. [Optional] Prove using PMI.
n(n + 1)(2n + 1)
.
1. 12 + 22 + + n2 =
6
2. 1 + 3 + + (2n 1) = n2 , for all n N.
3. n(n + 1) is even, for all n N.
4. 3 divides n3 n, for all n N.

n(n+1)
.
2

1.3. MORE ON PRINCIPLE OF MATHEMATICAL INDUCTION

15

5. 5 divides n5 n, for all n N.

Practice 1.3.10. [Wrong use of PMI: Can you find the error?] The following is an incorrect
proof of if a set of n balls contains a green ball then all the balls in the set are green. Find the
error.
Proof. The statement holds trivially for n = 1. Assume that the statement is true for n k.

Take a collection Bk+1 of k + 1 balls that contains at least one green ball. From Bk+1 , pick a

collection Bk of k balls that contains at least one green ball. Then, by the induction hypothesis,
each ball in Bk is green. Now, remove one ball from Bk and put the ball which was left out in
the beginning. Call it Bk . Again by induction hypothesis, each ball in Bk is green. Thus, each
ball in Bk+1 is green. Hence, by PMI, our proof is complete.
Exercise 1.3.11. [Optional]
1. Let x R with x 6= 1. Then, prove that 1 + x + x2 + + xn =

n
P

k=0

xk =

xn+1 1
.
x1

2. Let a, a + d, a + 2d, . . . , a + (n 1)d be the first n terms of an arithmetic progression. Then,

S=

n1
X
i=0

(a + id) = a + (a + d) + + (a + (n 1)d) =

n
(2a + (n 1)d) .
2

3. Let a, ar, ar 2 , . . . , ar n1 be the first n terms of a geometric progression, with r 6= 1. Then,

n1
P i
rn 1
.
S = a + ar + + ar n1 =
ar = a
r1
i=0
4. Prove that
(a) 6 divides n3 n, for all n N.
(b) 7 divides n7 n, for all n N.
(c) 3 divides 22n 1, for all n N.
(d) 9 divides 22n 3n 1, for all n N.
(e) 10 divides n9 n, for all n N.
(f ) 12 divides 22n+2 3n4 + 3n2 4, for all n N.


n(n + 1) 2
3
3
3
(g) 1 + 2 + + n =
.
2
5. Determine a formula for 1 2 + 2 3 + 3 4 + + (n 1) n and prove it.
6. Determine a formula for 1 2 3 + 2 3 4 + 3 4 5 + + (n 1) n (n + 1) and prove it.
7. Determine a formula for 1 3 5 + 2 4 6 + + n (n + 2) (n + 4) and prove it.
8. [Informative] For all n 32, there exist nonnegative integers x and y such that n =
5x + 9y. [Hint: First prove it for the starting 5 numbers, 32, 33, 34, 35, 36.]

9. [Informative] Prove that, for all n 40, there exist nonnegative integers x and y such
that n = 5x + 11y.

10. For every positive integer n 3 prove that 2n > n2 > 2n + 1.

16

CHAPTER 1. BASIC SET THEORY

11. Let r R with r > 1. Then, prove that

(1 + r)n 1 + rn for all n N.

(1.3)

12. [Informative] Prove that for > 0,



p
Y
p(p + 1)
1 p2 (p + 1)2 p(p + 1)(2p + 1)
(1 + l) 1 +
+

2 .
2
2
4
6
l=1

13. [Informative] By an L-shaped piece, we mean a piece of the type shown in the picture.
Consider a 2n 2n square with one unit square cut. See picture.

4 4 and 8 8 squares with a unit square cut

L-shaped piece

Show that a 2n 2n square with one unit square cut, can be covered with L-shaped pieces.
14. [Informative] Verify that (k+1)5 k5 = 5k4 +10k3 +10k2 +5k+1. Now, put k = 1, 2, . . . , n
n
n
n
n
n
P
P
P
P
P
1. Now, use
k+
k2 + 5
k3 + 10
k4 + 10
and add to get (n + 1)5 1 = 5
the formulas for

n
P

k=1

k3 ,

n
P

k=1

k2 ,

k=1
n
P

k=1

k and

k=1

n
P

k=1

k=1

1 to get a expression for

k=1

n
P

k4 .

k=1

k=1

15. [Informative: A general result than AM-GM]

(a) Let a1 , . . . , a9 be nonnegative real numbers such that the sum a1 + + a9 = 5.

Assume that a1 6= a2 . Consider

2
a1 +a2
a3 a9 .
2

a1 +a2 a1 +a2
2 ,
2 , a 3 , . . . , a9 .

Argue that a1 a9

(b) Let a1 , . . . , an be any nonnegative real numbers such that the sum a1 + + an = r0 .
Argue that the highest value of a1 an is obtained when a1 = = an = r0 /n.

(c) Let a1 , . . . , an be fixed nonnegative real numbers such that the sum a1 + + an = r0 .
Conclude from the previous item that (r0 /n)n a1 an , the AM-GM inequality.

1.4

Integers

In this section, we study some properties of integers. We start with the division algorithm.
Lemma 1.4.1. [Division algorithm] Let a and b be two integers with b > 0. Then, there exist
unique integers q, r such that a = qb + r, where 0 r < b. The integer q is called the quotient

and r, the remainder.

Proof. Existence: Take S = {a + bx | x Z} N0 . Then, a + |a|b S. Hence, S is a nonempty

subset of N0 . Therefore, by the Well-Ordering Principle, S contains its minimum, say s0 . So,
s0 = a + bx0 , for some x0 Z. Notice that s0 0. We claim that s0 < b.

1.4. INTEGERS

17

If s0 b then s0 b 0 and hence s0 b = a + b(x0 1) S, a contradiction to s0 being the

minimum element of S. Now, put q = x0 and r = s0 . Thus, we have obtained q and r such

that a = qb + r, with 0 r < b.

Uniqueness: Assume that there exist integers q1 , q2 , r1 and r2 satisfying a = q1 b+r1 , 0 r1 < b

and a = q2 b+r2 , 0 r2 < b. Without loss of generality, we assume r1 r2 . Then, 0 r2 r1 < b.

Notice that r2 r1 = (q1 q2 )b. So, 0 (q1 q2 )b < b. But the only integer multiple of b which

lies in [0, b) is 0. Hence, q1 q2 = 0. Thus, r1 = r2 as well. This completes the proof.

Definition 1.4.2. [Divisibility]

1. [Divisor] Let a, b Z with b 6= 0. If a = bc, for some c Z then b is said to divide (be a
divisor of) a and is denoted b | a.

Discussion: If a is a nonzero integer then the set of positive divisors of a is always nonempty
(as 1 | a) and finite (as a positive divisor of a is less than or equal to |a|).
2. [Greatest common divisor] Let a and b be two nonzero integers. Then, the set S of
their common positive divisors is nonempty and finite. Thus, S contains its greatest
element. This element is called the greatest common divisor of a and b and is denoted
gcd(a, b). On similar lines one can define the greatest common divisor of non-zero integers
a1 , a2 , . . . , an as the largest positive integer that divides each of a1 , a2 , . . . , an , denoted
gcd(a1 , . . . , an ).
3. [Relatively prime/Co-prime integers] An integer a is said to be relatively prime to an
integer b if gcd(a, b) = 1. Or, two integers a and b are said to be co-prime if gcd(a, b) = 1.
The next remark follows directly from the definition and the division algorithm.
Remark 1.4.3. Let a, b Z \ {0} and d = gcd(a, b). Then, for any positive common divisor c

of a and b, one has c | d.

The next result is often stated as the gcd(a, b) is an integer linear combination of a and b.
Theorem 1.4.4. [Bezouts identity]

Let a and b be two nonzero integers. Then, there exist

integers x0 , y0 such that d = ax0 + by0 , where d = gcd(a, b).

Proof. Consider the set S = {ax + by | x, y Z} N. Then, either a S or a S. Thus, S

is a nonempty subset of N. Hence, by the Well-ordering principle, S contains its least element,
say d. As d S, we have d = ax0 + by0 , for some x0 , y0 Z. We claim that d = gcd(a, b).

Note that d is positive. Let c be any positive common divisor of a and b. Then, c | ax0 +by0 = d

as x0 , y0 Z. We now show that d | a and d | b.

By division algorithm, there exist integers q and r such that a = dq + r, with 0 r < d. Thus,

we need to show that r = 0.

On the contrary, assume that 0 < r < d. Then

r = a dq = a q(ax0 + by0 ) = a(1 qx0 ) + b(qy0 ) {ax + by | x, y Z}.

18

CHAPTER 1. BASIC SET THEORY

Hence, r is a positive integer in S which is strictly less than d. This contradicts the fact that d
is the least element of S. Thus, r = 0 and hence d|a. Similarly, d|b.
The division algorithm gives us an idea to algorithmically compute the greatest common divisor
of two integers, commonly known as the Euclids algorithm.
Discussion 1.4.5.

1. Note that d, the number obtained by the application of the Well-

ordering principle in the proof of Theorem 1.4.4 has the property that d divides ax + by,
for all x, y Z. So, for every choice of integers x, y, gcd(a, b) divides ax + by.
2. Let a, b Z \ {0}. By division algorithm, a = |b|q + r, for some integers q, r Z with
0 r < |b|. Then,

gcd(a, b) = gcd(a, |b|) = gcd(|b|, r).

To show the second equality, note that r = a |b|q and hence gcd(a, |b|) | r. Thus,

gcd(a, |b|) | gcd(|b|, r). Similarly, gcd(|b|, r) | gcd(a, |b|) as a = |b|q + r.

3. We can now apply the above idea repeatedly to find the greatest common divisor of two
given nonzero integers. This is called the Euclids algorithm. For example, to find
gcd(155, 275), we proceed as follows
275 = (2) 155 + 35
155 = 4 35 + 15

(so, gcd(275, 155) = gcd(155, 35))

(so, gcd(155, 35) = gcd(35, 15))

35 = 2 15 + 5

(so, gcd(35, 15) = gcd(15, 5))

15 = 3 5

(so, gcd(15, 5) = 5).

To write 5 = gcd(155, 275) in the form 155x0 + (275)y0 , notice that

5 = 35215 = 352(155435) = 9352155 = 9(275+2155)2155 = 9(275)+16155.
Also, note that 275 = 555 and 155 = 531 and thus, 5 = (9+31x)(275)+(16+55x)155,
for all x Z. Therefore, we see that there are infinite number of choices for the pair

(x, y) Z2 , for which d = ax + by.

4. [Euclids algorithm] In general, given two nonzero integers a and b, the algorithm proceeds
as follows:
a = bq0 + r0 with 0 r0 < b,

b = r0 q1 + r1 with 0 r1 < r0 ,

r0 = r1 q2 + r2 with 0 r2 < r1 ,
.
..
. = ..

r1 = r2 q3 + r3 with 0 r3 < r2 ,

r1 = r q+1 + r+1 with 0 r+1 < r ,

r = r+1 q+2 .

The process will take at most b 1 steps as 0 r0 < b. Also, note that gcd(a, b) = r+1

and r+1 can be recursively obtained, using backtracking. That is,

r+1 = r1 r q+1 = r1 q+1 (r2 r1 q ) = r1 (1 + q+1 q ) q+1 r2 = .

1.4. INTEGERS
Exercise 1.4.6.

19
1. Let a, b, c N. Then, prove the following:

(a) If gcd(a, b) = d, then gcd( ad , db ) = 1.

(b) gcd(a, bc) = 1 if and only if gcd(a, b) = 1 and gcd(a, c) = 1.
2. Prove that the system 15x + 12y = b has a solution for x, y Z if and only if 3 divides b.
3. [Diophantine Equation] Let a, b, c Z \ {0}. Then, the linear system ax + by = c, in the
unknowns x, y Z has a solution if and only if gcd(a, b) divides c. Furthermore, determine

all pairs (x, y) Z Z such that ax + by is indeed c.

4. Let a1 , a2 , . . . , an N. Then, prove that gcd(a1 , a2 , . . . , an ) = gcd(gcd(a1 , a2 ), a3 , . . . , an ).

5. Euclids algorithm can sometimes be applied to check whether two numbers which are functions of an unknown integer n, are relatively prime or not? For example, we can use the
algorithm to prove that gcd(2n + 3, 5n + 7) = 1 for every n Z.
6. [Informative] Suppose a milkman has only 3 cans of sizes 7, 9 and 16 liters. If the milkman
has 16 litres of milk then using the 3 cans, specified as above, what is the minimum number
of operations required to deliver 1 liter of milk to a customer? Explain.
To proceed further, we need the following definitions.
Definition 1.4.7. [Prime/Composite numbers]
1. [unity] The positive integer 1 is called the unity (or the unit element) of Z.
2. [prime] A positive integer p is said to be a prime, if p is not a unit and p has exactly
two positive divisors, namely, 1 and p.
3. [composite] A positive integer r is called composite if r is neither a unit nor a prime.
We are now ready to prove an important result that helps us in proving the fundamental
theorem of arithmetic.
Lemma 1.4.8. [Euclids lemma] Let p be a prime and let a, b Z. If p | ab then either p | a

or p | b.

Proof. If p | a, we are done. So, assume that p a. As p is a prime, gcd(p, a) = 1. Thus, we can

find integers x, y such that 1 = ax + py. As p | ab, we have

p | abx + pby = b(ax + py) = b 1 = b.

Thus, if p|ab then either p|a or p|b.
As a repeated application of Lemma 1.4.8, we have the following result and hence the proof is
omitted.
Corollary 1.4.9. Let n be an integer such that n | ab and gcd(n, a) = 1. Then, n | b.

20

CHAPTER 1. BASIC SET THEORY

Now, we are ready to prove the fundamental theorem of arithmetic that states that every
positive integer greater than 1 is either a prime or is a product of primes. This product is
unique, except for the order in which the prime factors appear.
Theorem 1.4.10. [Fundamental theorem of arithmetic]

Let n N with n 2. Then,

there exist prime numbers p1 > p2 > > pk and positive integers s1 , s2 , . . . , sk such that
n = ps11 ps22 pskk , for some k 1. Moreover, if n also equals q1t1 q2t2 qt , for distinct primes

q1 > q2 > > q and positive integers t1 , t2 , . . . , t then k = and for each i, 1 i k, pi = qi

and si = ti .

Proof. We prove the result using the strong form of the principle of mathematical induction.
The result is clearly true for n = 2. So, let the result be true for all m, 2 m n 1. If n is a

prime, then we have nothing to prove. Else, n has a prime divisor p. Then, apply induction on
n
p

to get the required result.

Theorem 1.4.11. [Euclid: Infinitude of primes] The number of primes is infinite.

Proof. On the contrary assume that the number of primes is finite, say p1 = 2, p2 = 3, . . . , pk .
Now, consider the positive integer N = p1 p2 pk + 1. Then, we see that none of the primes

p1 , p2 , . . . , pk divides N which contradicts Theorem 1.4.10. Thus, the result follows.

Proposition 1.4.12. [Primality testing] Let n N with n 2. Suppose that for every prime

p n, p does not divide n, then n is prime.

Proof. Suppose n = xy, for 2 x, y < n. Then, either x n or y n. Without loss of

generality, assume x n. If x is a prime, we are done. Else, take a prime divisor of x to get
a contradiction.

Exercise 1.4.13. [Informative] Prove that there are infinitely many primes of the form 4n1.
Definition 1.4.14. [Least common multiple] Let a, b Z. Then, the least common mul-

tiple of a and b, denoted lcm(a, b), is the smallest positive integer that is a multiple of both a
and b.
Theorem 1.4.15. Let a, b N. Then, gcd(a, b) lcm(a, b) = ab. Thus, lcm(a, b) = ab if and

only if gcd(a, b) = 1.

Proof. Let d = gcd(a, b). Then, d = as + bt, for some s, t Z, a = a1 d, b = b2 d, for some

a1 , b1 N. We need to show that lcm(a, b) = a1 b1 d = ab1 = a1 b, which is clearly a multiple of

both a and b. Let c N be any common multiple of a and b. To show, a1 b1 d divides c. Note
that

as

c c
a, b

c
cd
c(as + bt)
c
c
=
=
= s+ t Z
a1 b1 d
(a1 d) (b1 d)
ab
b
a

Z and s, t Z. Thus, a1 b1 d = lcm(a, b) divides c and hence lcm(a, b) is indeed the

smallest. Thus, the required result follows.

Exercise 1.4.16.

1. If gcd(b, c) = 1, then gcd(a, bc) = gcd(a, b) gcd(a, c).

1.4. INTEGERS

21

2. If gcd(a, b) = d, then gcd(an , bn ) = dn for all n N.

Definition 1.4.17. [Modular Arithmetic] Fix a positive integer n. Then, an integer a is said
to be congruent to an integer b modulo n, denoted a b (mod n), if n divides a b.
Example 1.4.18.

1. The numbers 10 and 22 are equivalent modulo 4 as 4 | 12 = 22 10

and 4 | 32 = 22 (10).

2. Let n N be a perfect square. That is, there exists an integer m such that n = m2 . Then,
n 0, 1 (mod 4) as any integer m 0, 1, 2 (mod 4) and hence m2 0, 1 (mod 4).

3. Let S = {15, 115, 215, . . .}. Then, S doesnt contain any perfect square as for each s S,
s 3 (mod 4).

4. It can be easily verified that any two even (odd) integers are equivalent modulo 2 as
2 | 2(l m) = 2l 2m (2 | 2(l m) = ((2l + 1) (2m + 1))).
5. Let n be a fixed positive integer and let S = {0, 1, 2, . . . , n 1}.
(a) Then, by division algorithm, for any a Z there exists a unique b S such that b a

(mod n). The number a (mod n) (in short, b) is called the residue of a modulo n.
n1
S
(b) Thus, the set of integers, Z =
{a + kn : k Z}. That is, every integer is congruent
a=0

to an element of S. The set S is taken as the standard representative for the set
of residue classes modulo n.

Theorem 1.4.19. Let n be a positive integer. Then, the following results hold.
1. Let a b (mod n) and b c (mod n), for some a, b, c Z. Then, a c (mod n).
2. Let a b (mod n), for some a, b Z. Then, a + c b + c (mod n), a c b c (mod n)
and ac bc (mod n), for all c Z.

3. Let a b (mod n) and c d (mod n), for some a, b, c, d Z. Then, ac bd (mod n)

and ac bd (mod n). In particular, am bm (mod n), for all m N.

4. Let ac bc (mod n), for some non-zero a, b, c Z. Then, a b (mod n), whenever
n
gcd(c, n) = 1. In general, a b (mod
).
gcd(c, n)
Proof. We will only prove two parts. The readers should supply the proof of other parts.
Part 3: Note that ac bd ac bc + bc bd c(a b) + b(c d). Thus, n | ac bd, whenever

n | a b and n | c d.

In particular, taking c = a and d = b and repeatedly applying the above result, one has

am bm (mod n), for all m N.

Part 4: Let gcd(c, n) = d. Then, there exist non-zero c1 , n1 Z and c = c1 d, n = n1 d. Thus,

n | ac bc means that n1 d | c1 d(a b). This, in turn implies that n1 | c1 (a b). Hence, by
n
Corollary 1.4.9, we get
= n1 | a b.
gcd(c, n)
As an application, we have the following result, popularly known as the Fermats little theorem.
Theorem 1.4.20. [Fermats Little Theorem] Let p be a prime and n N. Then, np
n (mod p).

22

CHAPTER 1. BASIC SET THEORY

Proof. Note that if p|n, then obviously, np n (mod p). So, let us assume that gcd(p, n) = 1.
Then, we need to show that np1 1 (mod p).

To do this, we consider the set S = {n (mod p), 2n (mod p), . . . , (p 1)n (mod p)}. Since,

gcd(p, n) = 1, by second and fourth parts of Theorem 1.4.19, an bn (mod p) if and only if

a b (mod p). Thus, S {1, 2, . . . , p 1}. Hence,

np1 (p 1)! = n 2n (p 1)n 1 2 (p 1).

Thus, the condition gcd((p 1)!, p) = 1 implies that np1 1 (mod p).
Before coming to the next result, we look at the following examples.

Example 1.4.21.

1. As gcd(251, 13) = 1, we see that 25112 1 (mod 13). Hence, 25112

leaves the remainder 1, when divided by 13.

2. As 255 2 (mod 23), gcd(255, 23) = 1. Hence,

25527 (25522 ) (2555 ) (mod 23) 25

(mod 23) 32

(mod 2)3 9

(mod 2)3.

3. Note that 3 9 + 13 (2) 1 (mod 13). So, the system 9x 4 (mod 13) has the solution
x x 1 x (3 9 + 13 (2)) 3 9x 3 4 12

(mod 13).

4. Verify that 9 (5) + 23 (2) = 1. Hence, the system 9x 1 (mod 23) has the solution
x x 1 x (9 (5) + 23 (2)) (5) (9x) 5 18

(mod 23).

5. The system 3x 15 (mod 30) has solutions x = 5, 15, 25, whereas the system 7x = 15 has
only the solution x = 15. Also, verify that the system 3x 5 (mod 30) has no solution.

Theorem 1.4.22. [Linear Congruence] Let n be a positive integer and let a and b be non-zero
integers. Then, the system ax b (mod n) has at least one solution if and only if gcd(a, n) | b.
Moreover, if d = gcd(a, n) then ax b (mod n) has exactly d solutions in {0, 1, 2, . . . , n 1}.

Proof. Let x0 be a solution of ax b (mod n). Then, by definition, ax0 b = nq, for some

q Z. Thus, b = ax0 nq. But, gcd(a, n) | a, n and hence gcd(a, n) | ax0 nq = b.

Suppose d = gcd(a, n) | b. Then, b = b1 d, for some b1 Z. Also, by Euclidean algorithm,

there exists x0 , y0 Z such that ax0 + ny0 = d. Hence,

a(x0 b1 ) b1 (ax0 ) b1 (ax0 + ny0 ) b1 d b (mod n).

This completes the proof of the first part.
To proceed further, assume that x1 , x2 are two solutions. Then, ax1 ax2 (mod n) and
n
n
hence, by Theorem 1.4.19.4, x1 x2 (mod ). Thus, we can find x2 {0, 1, . . . , } such that
d
d
n
x = x2 + k is a solution of ax b (mod n), for 0 k d 1. Verify that these xs are distinct
d
and lie between 0 and n 1. Hence, the required result follows.
Exercise 1.4.23.

1. Prove Theorem 1.4.19.

1.4. INTEGERS

23

2. Prove that the numbers 19, 119, 219, . . . cannot be perfect squares.
3. Prove that the numbers 10, 110, 210, . . . cannot be perfect squares.
4. Prove that the system 3x 4 (mod 28) is equivalent to the system x 20 (mod 28).
5. Determine the solutions of the system 3x 5 (mod 65).
6. Determine the solutions of the system 15x 295 (mod 100).
7. Prove that the pair of systems 3x 4 (mod 28) and 4x 2 (mod 27) is equivalent to

the pair x 20 (mod 28) and x 14 (mod 27). Hence, prove that the above system is
equivalent to solving either 20 + 28k 14 (mod 27) or 14 + 27k 20 (mod 28) for the

unknown quantity k. Thus, verify that k = 21 is the solution for the first case and k = 22
for the other. Hence, x = 20 + 28 21 = 608 = 14 + 22 27 is a solution of the above pair.


p!
8. Let p be a prime. Then, prove that p | kp =
, for 1 k p 1.
k!(p k)!
9. [Informative] Let p be a prime. Then, the set
(a) Zp = {0, 1, 2, . . . , p 1} has the following properties:
i. for every a, b Zp , a + b (mod p) Zp .
ii. for every a, b Zp , a + b = b + a (mod p).

iii. for every a, b, c Zp , a + (b + c) (a + b) + c (mod p).

iv. for every a Zp , a + 0 a (mod p).

v. for every a Zp , a + (p a) 0 (mod p).

(b) Zp = {1, 2, . . . , p 1} has the following properties:

i. for every a, b Zp , a b (mod p) Zp .
ii. for every a, b Zp , a b = b a (mod p).

iii. for every a, b, c Zp , a (b c) (a b) c (mod p).

iv. for every a Zp , a 1 a (mod p).

v. for every a Zp , a b 1 (mod p). To see this, note that gcd(a, p) = 1. Hence,
by Euclids algorithm, there exists x, y Z such that ax + py = 1. Define b x
(mod p). Then,

abaxax+py 1

(mod p).

In algebra, any set, say F, in which addition and multiplication can be defined
in such a way that the above properties are satisfied then F is called a field. So,
Zp = {0, 1, 2, . . . , p 1} is an example of a field. In general, the well known examples
of fields are:
i. Q, the set of rational numbers.
ii. R, the set of real numbers.
iii. C, the set of complex numbers.
(c) From now on let p be an odd prime.
i. Then, the equation x2 1 (mod p) has exactly two solutions, namely x = 1 and
x = p 1 in Zp . This is true as p is a prime dividing x2 1 = (x 1)(x + 1)

24

CHAPTER 1. BASIC SET THEORY

implies that either p|x 1 or p|x + 1. Also, 0 is the only number in Zp that is
divisible by p.

ii. Then, for a {2, 3, . . . , p 2}, we can find a number b {2, 3, . . . , p 2} Zp

with a b 1 (mod p) and b 6= a.

iii. Thus, for 1 i

p1
2 ,

we have pairs {ai , bi } that are pairwise disjoint and satisfy

ai bi 1 (mod p). Moreover,

p1
2

i=1

{ai , bi } = {2, 3, . . . , p 2}.

iv. Hence, 2 3 (p 2) 1 (mod p).

v. We thus have the following famous theorem called the Wilsons Theorem: Let
p be a prime. Then, (p 1)! 1 (mod p).

Proof. Note that from the previous step, we have

(p 1)! 1 (p 1) 2 3 (p 2) 1 1 1 (mod p).
vi. [Primality Testing] Let n be a positive integer. Then, (n 1)! 1 (mod n) if
and only if n is a prime.

Theorem 1.4.24. [Chinese remainder theorem] Fix a positive integer m and let n1 , n, . . . , nm
be pairwise co-prime positive integers. Then, the linear system
x a1

(mod n1 )

x a2
..
.

(mod n2 )

x am

(mod nm )

has a unique solution modulo N = n1 n2 nm .

M
Proof. For 1 k m, define Mk =
. Then, gcd(Mk , nk ) = 1 and hence there exist integers
nk
xk , yk such that Mk xk + nk yk = 1 for 1 k m. Then,

Define x0 =
relations.

m
P

Mk xk 1

(mod nk ) and Mk xk 0 (mod n ) for 6= k.

Mk xk ak . Then, it can be easily verified that x0 satisfies the required congruence

k=1

Example 1.4.25.

1. Let us come back to Exercise 1.4.23.7. In this case, note that a1 =

20, a2 = 14, n1 = 28 and n2 = 27. So, M = 28 27 = 756, M1 = 27 and M2 = 28. As,

27 (1) + 28 1 = 1, we see that x1 = 1 and x2 = 1. Thus,

x0 = 27 1 20 + 28 1 14 540 + 392 148 608

(mod 756).

Alternate: Note that 27 (1) + 28 1 = 1. So, 20 27 (1) 20 (mod 28) and

14 28 1 14 (mod 27). Thus,

27 (1) 20 + 28 1 14 27 (1) 20
28 1 14

(mod 28) 20

(mod 28)

(mod 27) 14 (mod 27).

But, 148 608 (mod 756) and hence x0 = 608 is the required answer.

1.4. INTEGERS

25

2. Let x be a number which when divided by 8, 15 and 17 gives remainders 5, 6 and 8,

respectively. Then, what will be the remainder when x is divided by 2040?
Ans: Note that the question reduces to finding x N such that
x 5 (mod 8),

x 6 (mod 15),

x 8 (mod 17).

Also, M = 2040, M1 = 255, M2 = 136 and M3 = 120 with

8 32 + (1) 255 = 1, 1 136 + (9)15 = 1 and 1 120 + (7) 17 = 1.
Hence, the required remainder is
501 = 255 (1) 5 + 136 1 6 + 120 1 8 M1 x1 a1 + M2 x2 a2 + M3 x3 a3
5 (255)

(mod 8) 5 (mod 8)

6 136

(mod 15) 6

(mod 15)

8 120

(mod 17) 8

(mod 17).

Exercise 1.4.26.

(mod 2040)

1. Find the smallest positive integer which when divided by 4 leaves a

remained 1 and when divided by 9 leaves a remainder 2.

2. Find the smallest positive integer which when divided by 8 leaves a remained 4 and when
divided by 15 leaves a remainder 10.
3. Does there exist a positive integer n such that
n4

(mod 14), n 6

(mod 18)?

Give reasons for your answer. What if we replace 6 or 4 with an odd number?
4. [Informative] Let n be a positive integer. Then, the set
(a) Zn = {0, 1, 2, . . . , n 1} has the following properties:
i. for every a, b Zn , a + b (mod n) Zn .
ii. for every a, b Zn , a + b = b + a (mod n).

iii. for every a, b, c Zn , a + (b + c) (a + b) + c (mod n).

iv. for every a Zn , a + 0 a (mod n).

v. for every a Zn , a + (p a) 0 (mod n).

vi. for every a, b Zn , a b (mod n) Zn .

vii. for every a, b Zn , a b = b a (mod n).

viii. for every a, b, c Zn , a (b c) (a b) c (mod n).

ix. for every a Zn , a 1 a (mod n).

In algebra, any set, say R, in which addition and multiplication can be defined in

such a way that the above properties are satisfied then R is called a commutative

ring with unity. So, Zn = {0, 1, 2, . . . , n 1} is an example of a commutative ring

with unity. In general, the well known examples of commutative ring with unity are:
i. Z, the set of integers.

26

CHAPTER 1. BASIC SET THEORY

ii. Q, the set of rational numbers.
iii. R, the set of real numbers.
iv. C, the set of complex numbers.
(b) Now, let m and n be two co-prime positive integers. Then, by the above, the sets
Zm , Zn , and Zmn are commutative rings with unity. In the following, we show that
there is a one-to-one correspondence (ring isomorphism) between Zm Zn and Zmn .

To do so, define

f : Zmn Zm Zn by f (x) = (x

(mod m), x

(mod n)) for all x Zmn .

Then, defining addition and multiplication in Zm Zn componentwise and using

Theorem 1.4.19, we have the following:

i. f (x + y) = f (x) + f (y), for all x, y Zmn .
ii. f (x y) = f (x) f (y), for all x, y Zmn .

iii. for every (a, b) Zm Zn , by CRT, there exists a unique x Zmn such that
xa

(mod m) and x b

(mod n).

iv. also, | Zm Zn |=| Zmn |= mn..

Hence, we have obtained the required one-one correspondence, commonly known as
the ring isomorphism. That is, the two rings Zm Zn and Zmn are isomorphic.
5. [Arithmetic Function] A function f : N C is called an arithmetic function.
[Multiplicative Function] An arithmetic function f is called a multiplicative function if

f (mn) = f (m)f (n) for all m, n N with gcd(m, n) = 1.

We now give a list of arithmetic functions which are quite popular in number theory. The
reader should determine the functions that are multiplicative.
(a) Consider the function Id(n) = n, for all n N.
(b) Consider the function U(n) = 1, for all n N.

(c) Fix a positive integer m and for all n N, consider the function m (n) =

if n = m

0 otherwise.
(d) Recall that a number n N is called squarefree, if for any prime p, p divides n but
p2 doesnt divide n. Consider the function : N C, defined by

if n is not squarefree,

0
(n) =
1
if n is squarefree and has even number of prime factors,

1 if n is squarefree and has odd number of prime factors,

This function is commonly known as the M

obius function. For example, (1) =
1, (2) = 1, (3) = 1, (4) = 0, . . . , (10) = 1, . . ..

(e) Consider the function : N C, defined by (n) = |{k : 1 k n, gcd(k, n) = 1}|,

for all n N. This function is popularly known as the totient /phi /Euler phi function
and it counts the number of positive integers less than or equal to n that are co-prime
to n. For example, (1) = 1, (2) = 1, (3) = 2, (4) = 2, . . . , (10) = 4, . . ..

1.4. INTEGERS

27

i. Let m, n N with gcd(m, n) = 1. Then, use Exercise 1.4.6.1b to prove that

(mn) = (m) (n). So, the function is a multiplicative function. Hence, we

need to only determine (pn ) for every prime p and n N.



ii. Show that if p is a prime and n N then (pn ) = pn1 (p 1) = pn 1 1p .

iii. [Eulers Theorem] Let n N and let a Z such that gcd(a, n) = 1. Then,
a(n) 1 (mod n).A generalization of Fermats little theorem.

(f ) Consider the function : N C, defined by (n) = |{k : 1 k n, k is a prime}|,

for all n N. This function counts the number of primes less than or equal to n. For

example, (1) = 0, (2) = 1, (3) = 2, (4) = 2, . . . , (10) = 4, . . ..

P
1, for all n N. This function
(g) Consider the function d : N C, defined by d(n) =
r|n

counts the number of positive divisors of n. For example, d(1) = 1, d(2) = 2, d(3) =

2, d(4) = 3, . . . , d(10) = 4, . . ..
(h) Consider the function : N C, defined by (n) =

d|n

d, for all n N. This function

gives the sum of the positive divisors of n. For example, (1) = 1, (2) = 3, (3) =
4, (4) = 7, . . . , (10) = 18, . . ..
(i) Fix a positive integer k and consider the function k : N C, defined by k (n) =
P k
d , for all n N. This function gives the sum of the k-th powers of the positive
d|n

divisors of n. For example, k (1) = 1, k (2) = 1 + 2k , (3) = 1 + 3k , . . . , (10) =

1 + 2k + 5k + 10k , . . .. Also, note that 0 (n) = d(n) and 1 (n) = (n), for all n N.

(j) Let f be an arithmetic function. Then, we define a function D from the set of
P
f (d).
arithmetic functions to itself, called the divisor sum function, by (Df )(n) =
d|n

28

CHAPTER 1. BASIC SET THEORY

Chapter 2

Advanced topics in set theory

2.1

Families of Sets

Definition 2.1.1. Let A be a set. For each x A, take a new set Ax . Then, the collection


{Ax }xA := Ax | x A is a family of sets indexed by elements of A (index set). Unless
otherwise mentioned, we assume that the index set for a class of sets is nonempty.
Definition 2.1.2. Let {B }S be a nonempty class of sets. We define their
1. union as B = {x | x B , for some }, and
S

2. intersection as B = {x | x B , for all }.

S

3. Convention: Union of an empty class is . The intersection of an empty class of subsets

of X is X 1 .

1. Take A n= [3], A1 = {1, 2}, o

A2 = {2, 3} and A3 = {4, 5}. Then, {A |
A} = {A1 , A2 , A3 } = {1, 2}, {2, 3}, {4, 5} . Thus, A = [5] and A = .

Example 2.1.3.

2. n
Take A = N and An = {n,
o n + 1, . . .}. Then, the family {A | A} = {A1 , A2 , . . .} =
{1, 2, . . .}, {2, 3, . . .}, . . . . Thus, A = N and A = .
A
A
T
[ n1 , n2 ] = {0}.
3. Prove that the intersection
nN

We now give a set of important rules whose proofs are left for the reader.

Theorem 2.1.4. [Algebra of union and intersection] Let {A }L be a nonempty class of

subsets of X and B be any set. Then, the following statements are true.


1. B A = (B A ).
L
L


2. B A = (B A ).
3.
1

L

c
A

= Ac .
L

The way we see this convention is as follows: First we agree that the intersection of an empty class of subsets

is a subset of X. Now, let x X such that x 6 B . This implies that there exists an S such that x 6 B .
Since S is empty, such an does not exist.

29

30

CHAPTER 2. ADVANCED TOPICS IN SET THEORY

A

4.

c

= Ac .
L

Proof. We give the proofs for Part 1 and 4. For Part 1, we see that


x B A x B and x A x B and x A , for some L
L

x B A , for some L x (B A ).
L

For Part 4, we have

x

c

x 6 A x 6 A , for some L x Ac , for some L

L

x Ac .
L

Proceed in similar lines to complete the proofs of the other parts.


Exercise 2.1.5.
1. Consider Ax }xR , where Ax = [x, x + 1]. What is Ax and Ax ?
xR

xR

2. For x [0, 1] write Zx := {zx | z Z} and Ax = R \ Zx. What is Ax and Ax ?

xR

xR

3. Write the closed interval [1, 2] = In , where In are open intervals.

nN

4. Write R as a union of infinite number of pairwise disjoint infinite sets.

5. Write the set [4] as the intersection of infinite number of infinite sets.
6. Suppose that AB = B. Is A = ?
7. Prove Theorem 2.1.4.

2.2

More on Relations

Proposition 2.2.1. [Properties of union and intersection under a relation] Let f : X Y

be a relation and {A }L P(X). Then, the following statements hold.


1. f A = f (A ).
L
L


2. f A f (A ). Give an example where the inclusion is strict.
L

Proof. Part 1:
yf

A (x, y) f, for some x A (x, y) f with x A , for some L
L

y f (A ), for some L y f (A ).
L

For Part 2, we assume that A 6= . Then,

L

yf

A (x, y) f, for some x A (x, y) f with x A , for all L
L

y f (A ), for all L y f (A ).
L

Thus, the required result follows.

2.2. MORE ON RELATIONS

31

Remark 2.2.2. It is important to note the following in the proof of the above theorem:
y f (A ), for all L implies that for each L, we can find some x A such that

(x , y) f . That is, the x s need not be the same. This gives you an idea to construct a

counterexample.

Define f : {1, 2, 3, 4} {a, b} by f = {(1, a), (2, a), (2, b), (3, b), (4, b)}. Take A1 = {1, 3} and

A2 = {1, 2, 4} and verify that the inclusion in Part 2 of Theorem 2.2.1 is strict. Also, find the

xi s for b.

Definition 2.2.3. [Composition of relations] We define the composition g f of relations f

and g as

n
o
g f = (x, z) | (x, y) f and (y, z) g for some y rng(f ) dom(g) .

It is a relation. The composition f g is defined similarly. In case, both f and g are functions

then (f g)(x) = f (g(x)).

Example 2.2.4. Take f = {(, a), (3, b), (3, c)} and g = {(b, ), (c, )}. Then, g f = {(3, )}
and f g = {(b, a), (c, a)}.

Definition 2.2.5. [Identity relation] A relation from X to X is called a relation on X. The

relation I := {(x, x) | x X} is called the identity relation on X. It is also a function if

X 6= .

Definition 2.2.6. [Equivalence relation] Let f be a relation on X. We call f reflexive if

I f . We call f symmetric if f = f 1 . We call f transitive if f f f . An equivalence
relation on X is a relation which is reflexive, symmetric and transitive.
Example 2.2.7. On X = [5],
1. I is an equivalence relation,
2. f := I {(1, 2), (2, 1)} is also an equivalence relation, and
3. g := I {(1, 2), (2, 1), (1, 3)(3, 1)} is reflexive, symmetric but not transitive ((3, 2)
/ g).

4. Let f = {(a, b) Z Z : 10a b}. Then, f is an equivalence relation.

5. Fix a positive integer n and let f = {(a, b) Z Z : na b}. Then, f is an equivalence
relation.

Example 2.2.8. [Some more examples of relations]

1. Let X 6= . Then, is a symmetric and transitive relation on X, but not reflexive.
2. Let A = {a, b, c, d}. Then, some of the relations R on A are:
(a) R = A A.
(b) R = {(a, a), (b, b), (c, c), (d, d), (a, b), (a, c), (b, c)}.
(c) R = {(a, a), (b, b), (c, c)}.

32

CHAPTER 2. ADVANCED TOPICS IN SET THEORY

(d) R = {(a, a), (a, b), (b, a), (b, b), (c, d)}.
(e) R = {(a, a), (a, b), (b, a), (a, c), (c, a), (c, c), (b, b)}.
(f) R = {(a, b), (b, c), (a, c), (d, d)}.
3. Some of the relations on Z are as follows:
(a) R = {(a, b) Z2 | 7 divides a b}.
(b) R = {(a, b) Z2 | a b}.
(c) R = {(a, b) Z2 | a > b}.
(d) R = {(a, b) Z Z | a|b}, where Z = Z \ {0}.
4. Consider the set R2 . Also, let us write x = (x1 , x2 ) and y = (y1 , y2 ). Then, some of the
relations on R2 are as follows:
(a) R = {(x, y) R2 R2 | |x|2 = x21 + x22 = y12 + y22 = |y|2 }.
(b) R = {(x, y) R2 R2 | x = y for some R }.
(c) R = {(x, y) R2 R2 | 4x21 + 9x22 = 4y12 + 9y22 }.
(d) R = {(x, y) R2 R2 | x y = (1, 1) for some R }.
(e) Fix c R. Now, define R = {(x, y) R2 R2 | y2 x2 = c(y1 x1 )}.
(f) R = {(x, y) R2 R2 | |x| = |y|}, for some positive real number .
(g) R = {(x, y) R2 R2 | x1 x2 = y1 y2 }.
(h) Let S = {x R2 | x21 + x22 = 1}. Define a relation on S by
i. R = {(x, y) S S | x1 = y1 , x2 = y2 }.

ii. R = {(x, y) S S | x = y}.

5. Let A be the set of triangles in the plane. Then, R = {(a, b) A2 | a b}, where stands
for similarity of triangles.

6. In R, define a relation R = {(a, b) R2 | a b is an integer}.

7. Let A be any nonempty set and consider the set P(A). Then, one can define a relation R
on P(A) by R = {(S, T ) P(A) P(A) | S T }.

2.2. MORE ON RELATIONS

33

Picture for a relation on X

We can draw pictures for relations on X. We first put a point for each element x of X and
label it x. For each (x, y) in the relation, we put a directed line from x to y. If we have an
(x, x) we draw a loop at x. The following are some pictures for the relations in Example
2.2.7.
5

2
f

2
g

Definition 2.2.9. [Equivalence class] Let f be an equivalence relation on X and a X. The

set {x X | (x, a) f } is called the equivalence class of a and is denoted Ea .

Example 2.2.10. Consider the relations in Example 2.2.7.

1. For the relation I, we have 5 equivalence classes, namely, {1}, {2}, {3}, {4} and {5}.
2. For the relation f , we have 4 equivalence classes, namely, {1, 2}, {3}, {4} and {5}.
3. For g {(3, 2), (2, 3)}, we have 3 equivalence classes, namely, {1, 2, 3}, {4} and {5}.
4. Notice that in all the three cases, the different equivalence classes are disjoint sets and in
the picture they appear disconnected even after joining 2 and 3 by directed edges in g.
Proposition 2.2.11. [Equivalence relation divides a set into disjoint classes] Let f be an
equivalence relation on X. Then, the following statements are true.
1. (a, b) f if and only if Ea = Eb .
2. (a, b)
/ f if and only if Ea Eb = .
3. Furthermore, X = Ea .
aX

Thus, an equivalence relation f on X divides X into disjoint equivalence classes.

Example 2.2.12. Let f be an equivalence relation on [5] whose equivalence classes are {1, 2},
{3, 5} and {4}. Then, f must be I {(1, 2), (2, 1), (3, 5), (5, 3)}.

Proposition 2.2.13. [Constructing equivalence relation from equivalence classes] Let f be

an equivalence relation on X 6= whose disjoint equivalence classes are {Ea | a A}. Then,
f = I {(x, y) | x, y Ea }.
aA

34

CHAPTER 2. ADVANCED TOPICS IN SET THEORY

Exercise 2.2.14.

1. Let X and Y be two nonempty sets and f : X Y be a relation. Let

Ix and Iy be the identity relations on X and Y , respectively. Then,

(a) is it necessary that f 1 f Ix ?

(b) is it necessary that f 1 f Ix ?

(c) is it necessary that f f 1 Iy ?

(d) is it necessary that f f 1 Iy ?

2. Suppose now that f is a function. Then,

(a) is it necessary that f f 1 Iy ?

(b) is it necessary that f 1 f Ix ?

3. Write down the equivalence classes for the equivalence relations that appear in Examples 2.2.7 and 2.2.8.
4. Take A 6= . Is A A an equivalence relation on A? If yes, what are the equivalence
classes?

5. On a nonempty set A, what is the smallest equivalence relation (in the sense that every
other equivalence relation will contain this equivalence relation; recall that a relation is a
set)?
Exercise 2.2.15. [Optional]
1. Let X = [5] and f be a relation on X. By checking whether f is reflexive or not, whether
f is symmetric or not and whether f is transitive or not, we see that there are 8 types of
relations on X. Give one example for each type.
2. Let A = B = [3]. Then, what is the number of
(a) relations from A to B?
(b) relations f from [3] to {a, b, c} such that dom f = {1, 3}?
(c) relations f from [3] to [3] such that f = f 1 ?

(d) single valued relations from [3] to [3]? How many of them are functions?
(e) equivalence relations on [5].
3. [Important] Let f : X Y be a single valued relation, A X, B Y and {B }I be a
nonempty family of subsets of Y . Then, show that


(a) f 1 B = f 1 (B ).
I
I


1
(b) f
B = f 1 (B ).
I
1
f (B c )

= dom f \ f 1 (B).


(d) f f 1 (B) A = B f (A). Note that this equality fails if f is not single valued.
(c)

4. Let f, g be two non-equivalence relations on R. Then, is it possible to have f g as an

equivalence relation? Give reasons for your answer.

2.3. MORE ON FUNCTIONS

35

5. Let f, g be two equivalence relations on R. Then, prove/disprove the following statements.

(a) f g is necessarily an equivalence relation.
(b) f g is necessarily an equivalence relation.
(c) f g is necessarily an equivalence relation.

(d) f g c is necessarily an equivalence relation.

6. Show that each set can be written as a union of finite sets.
7. Give an example of an equivalence relation on N for which there are 7 equivalence classes,
out of which exactly 5 are infinite.
8. Show that union of finitely many finite sets is a finite set.

2.3

More on functions

Recall that a function f : A B is a single valued relation with dom f = A. The readers need

to carefully read the following important remark before proceeding further.

Remark 2.3.1.

1. If A = , then by convention, one assumes that there is a function, called

the empty function, from A to B.

2. If B = , then it can be easily observed that there is no function from A to B.

3. Some books use the word map in place of function. So, both the words are used interchangeably throughout the notes.
4. Throughout these notes, whenever the phrase let f : A B be a function is used, it will
be assumed that both A and B are nonempty sets.

Example 2.3.2.

1. Let A = {a, b, c}, B = {1, 2, 3} and C = {3, 4}. Then, verify that the

examples given below are indeed functions.

(a) f : A B, defined by f (a) = 3, f (b) = 3 and f (c) = 3.

(b) f : A B, defined by f (a) = 3, f (b) = 2 and f (c) = 2.
(c) f : A B, defined by f (a) = 3, f (b) = 1 and f (c) = 2.
(d) f : A C, defined by f (a) = 3, f (b) = 3 and f (c) = 3.
(e) f : C A, defined by f (3) = a, f (4) = c.
2. Note that the following relations f : Z Z are indeed functions.
(a) f = {(x, 1) | x is even} {(x, 5) | x is odd}.
(b) f = {(x, 1) | x Z}.
(c) f = {(x, x (mod 10)) | x Z}.

36

CHAPTER 2. ADVANCED TOPICS IN SET THEORY

(d) f = {(x, 1) | x < 0} {(0, 0)} {(x, 1) | x > 0}.

Exercise 2.3.3. Do the following relations represent functions? Give reasons for your answer.
1. Let f : Z Z be defined by
(a) f = {(x, 1) | 2 divides x} {(x, 5) | 3 divides x}.
(b) f = {(x, 1) | x S} {(x, 1) | x S c }, where S is the set of perfect squares in Z.
(c) f = {(x, x3 ) | x Z}.

2. Let f : R+ R be defined by f = {(x, x) | x R+ }.

3. Let f : R R be defined by f = {(x,
4. Let f : R C be defined by f = {(x,

x) | x R}.
x) | x R}.

5. Let f : R R be defined by f = {(x, log e |x|) | x R }.

6. Let f : R R be defined by f = {(x, tan x) | x R}.




Exercise 2.3.4.
1. Define f : N Z by f = { x, x
| x is even} { x, x+1
| x is odd}.
2
2
Is f one-one? Is it onto?

2. Define f : N Z and g : Z Z by f = {(x, 2x) | x N} and g = { x, x2 | x is even}
{(x, 0) | x is odd}. Are f, g, and g f one-one? Are they onto?

3. Let A be the class of subsets of [9] of size 5 and B be the class of 5 digit numbers with
strictly increasing digits. For a A, define f (a) the number obtained by arranging the
elements of a in increasing order. Is f one-one and onto?

Proposition 2.3.5. [Algebra of composition of functions] Let f : A B, g : B C and

h : C D be functions.

1. Then, (h g) f : A D and h (g f ) : A D are functions. Moreover, (h g) f =

h (g f ) (associativity holds).

2. If f and g are injections then g f : A C is an injection.

3. If f and g are surjections then g f : A C is a surjection.
4. Let A and B be sets with at least two elements each and let f : A B be a bijection.
Then, the number of bijections from A to B is at least 2.

5. [Extension] Let f : A B and g : C D be bijections. Suppose that dom f dom g =

and rng f rng g = . Then, (f g) : (A C) (B D) is a bijection, where
f g = {(x, f (x)) | x A} {(x, g(x)) | x C}.
Definition 2.3.6. Fix a set A and let IdA : A A be defined by IdA (a) = a, for all a A.
Then, the function eA is called the identity function or map on A.

2.3. MORE ON FUNCTIONS

37

The subscript A in Definition 2.3.6 will be removed, whenever there is no chance of confusion
about the domain of the function.
Theorem 2.3.7. [Properties of identity function] Let A and B be nonempty sets, f : A B
and g : B A be any two functions. Then, for the Id map defined above, one has
1. Id is a one-one and onto map.
2. the map f Id = f .
3. the map Id g = g.
Proof. Part 1: Let Id = IdA . Then, by definition, Id(a) = a, for all a A and hence it is clear
that Id is one-one and onto.

Part 2: By definition, (f Id)(a) = f (Id(a)) = f (a), for all a A. Hence, f Id = f .

Part 3: The readers are advised to supply the proof.

Theorem 2.3.8. [bijection principle] Let f : A B and g : B A be functions. Prove that

if

1. g f (i) = i, for each i A, then f is one-one.

2. f g(j) = j, for each j B, then f is onto.
Proof. Let g f (i) = i, for each i A. Then, the assumption f (i) = f (j) implies that

i = g f (i) = g f (j) = j. Thus, f is one-one and the proof of the first part is over.

For the second part, let f g(j) = j, for each j B. To see that f is onto, let b B and put

a = g(b) A. Then, by the given assumption, f (a) = f (g(b)) = b.

Exercise 2.3.9.

1. Let f, g : N N be defined by f = {(x, 2x) | x N} and g = { x, x2 |

x is even} {(x, 0) | x is odd}. Then, verify that g f is the identity map on N, whereas

f g maps even numbers to itself and maps odd numbers to 0.

2. Let f : A B be a function. Then, f 1 is a function if and only if f is a bijection.

Cantors Experiment for the student: Why does it happen?
Take a plain paper.
1. On the left draw an oval (of vertical length) and write the elements of [4] inside it, one
below the other. On the right draw a similar but large oval and write the elements of
P([4]) inside it, one below the other.
2. Now draw a directed line from 1 (on the left) to any element on the right. Repeat
this for 2, 3 and 4. We have drawn a function. Call it f .
3. Notice that f (1), f (2), f (3) and f (4) are sets. Find out the set A = {i : i
/ f (i)}.
Locate this set on the right.

4. It is guaranteed that you do not have a directed line touching A. Why?

38

CHAPTER 2. ADVANCED TOPICS IN SET THEORY

Lemma 2.3.10. [Cantor] Let S be a set and f : S P(S) be a function. Then, there exists

A P(S) which does not have a pre-image. That is, there is no surjection from S to P(S).

Proof. On the contrary assume that f : S P(S) is a surjection. Then, for A = {x : x

/ f (x)}

there exists a S such that f (a) = A. So, A = f (a) = {x : x

/ f (x)}. We now show that a

neither belongs to A nor to Ac .

If a A, then by definition of A, a
/ f (a) = A. Similarly, if a
/ A means that a f (a) = A.

Thus, a
/ A Ac = S, a contradiction.
Exercise 2.3.11.

1. Define f : N2 N by f (m, n) = 2m1 (2n 1). Is f a bijection?

2. Are the following relations single valued functions, one-one, onto, and/or bijections?
(a) f : R R defined by f = {(x, sin x) | x R}.

(b) f = {(x, y) | x2 + y 2 = 1, x, y R} from dom(f ) to rng(f )

(c) f = {(x, y) | x2 + y 2 = 1, x 0, y 0} from dom(f ) to rng(f )

(d) f : R R defined by f = {(x, tan x) | 2 < x < 2 }.

3. Let f : X Y be one-one and {A }L be a nonempty family of subsets of X. Is

f A = f (A )?
L

4. Let f : X Y be a bijection and A X. Is f (Ac ) = (f (A))c ?

5. Let f : X Y and g : Y X be two functions such that
(a) (f g)(y) = y holds, for each y Y .
(b) (g f )(x) = x holds, for each x X.

Show that f is a bijection and g = f 1 . Can we conclude the same without assuming the
second condition?

2.4

Supplying bijections

Experiment 1:
Make a horizontal list of the elements of N using only once. Now, horizontally list

the elements of Z just below the list of N using once. Draw vertical lines to supply a
bijection from N to Z. Can you supply another by changing the second list a little bit?

Experiment 1:
Suppose that you have an open interval (a, b). Its center is c =
center from one end is

l
2

ba
2 .

a+b
2

and the distance of the

View this as a line segment on the real line. Stretch (a, b)

uniformly without disturbing the center and make its length equal to L.
Where is c now (in R)? Where is c 2l ? Where is c + 2l ? Where is c 2l , for a fixed

(1, 1)?

Now, use the above idea to find a bijection from (a, b) to (s, t)? [Hint: Fix the center first.]

2.4. SUPPLYING BIJECTIONS

Exercise 2.4.1.

39

1. Supply two bijections from (1, ) to (5, ), one by scaling and the

other by translating.

2. Take reciprocal to supply a bijection from (0, 1) to (1, ). You can also use the exponential
function to get this.

3. Supply a bijection from (1, 1) to (, ).

4. Supply a bijection from (0, 1) (0, 1) to R R.
Train-Seat argument to find a bijection
Let f : P = (0, 1) T = (3, 5) be a bijection. Imagine elements of P as PERSONS and

elements of T as seats in a TRAIN. So, f assign a seat to each person and the train is full.

1. Now suppose a new person 0 is arriving. He wants a seat. To manage it, let us un-seat
two persons 12 , 13 . So, two seats f ( 12 ), f ( 13 ) are vacant. But we have 3 persons to take
those seats. Giving each person a seat is not possible.
1
2. Suppose that we un-seat 21 , 13 , , 30
? Can we manage it?

3. Suppose that we un-seat 12 , 13 , ? Can we manage it now?

4. What do we do if we had two new persons arriving? Fifty new persons arriving? A
set {a1 , a2 , } of new persons arriving?
The next result is left as an exercise for the students.
Theorem 2.4.2. Let A be a set containing the set {a1 , a2 , . . . , } and let f : A B be a bijection.
Then, prove that, for any collection

1. {c1 , . . . , ck } of elements that are

h(x) =

outside A, the function

f (x)

if x A \ {a1 , a2 , . . .}

f (ai+k ) if x = ai , i N

f (ai )

if x = ci , i = 1, 2, . . . , k.

is a bijection from A {c1 , . . . , ck } to B.

2. {c1 , c2 , . . .} of elements that are outside A,

f (x)
h(x) =
f (a2n1 )

f (a )
2n

the function
if x A \ {a1 , a2 , . . .}
if x = an , n N

if x = cn , n N

is a bijection from A {c1 , c2 , . . .} to B.

Proof. Exercise.
Exercise 2.4.3. In the following, give bijections from A to B, where
1. A = [0, 1) and B = (0, 1).
2. A = (0, 1) {1, 2, 3, 4} and B = (0, 1).

40

CHAPTER 2. ADVANCED TOPICS IN SET THEORY

3. (0, 1) N to (0, 1).
4. A = [0, 1] and B = [0, 1] \ { 11 , 13 , 15 , }.
5. A = R and B = R \ N.
6. A = (0, 1) and B = R \ N.
7. A = [0, 1] and B = R \ N.
8. A = (0, 1) and B = (1, 2) (3, 4).
9. A = R \ Z and B = R \ N.
Creating bijections from injections
Let X = Y = N. Take injections f : X Y and g : Y X defined as f (x) = x + 2 and
g(x) = x + 1. In the picture, we have X on the left and Y on the right. If (x, y) f , we
draw a solid line joining x and y. If (y, x) g, we draw a dotted line joining y and x.
1
b

b
b

..

..

We want to create a bijection h from X to Y by erasing some of these lines.

1. Thus, h(1) must be 3. So, the dotted line (3, 4) cannot be used for h.
2. So, h(4) must be 6. So, the dotted line (6, 7) cannot be used for h.
3. So, h(7) must be 9. Continue two more steps to realize what is happening.
(
f (x),
if x = 3n 2, n N
So, the bijection h : X Y is given by h(x) =
g1 (x), otherwise.
Exercise 2.4.4. Take X = Y = N. Supply bijections using the given injections f : X Y and
g : Y X.

1. f (x) = x + 1 and g(x) = x + 2.

2. f (x) = x + 1 and g(x) = x + 3.
3. f (x) = x + 1 and g(x) = 2x.
Theorem 2.4.5. [Schr
oder-Bernstein: Creating a bijection] Let A and B be two non-empty
sets and let f : A B and g : B A be injections. Then, there exists a bijection from A to B.

2.4. SUPPLYING BIJECTIONS

41

Proof. If g is onto, we have nothing to prove. So, assume that g is not onto. Put O = A \ g(B),
= g f and E = O (O) 2 (O) . Use 0 (O) to denote O. Notice that





g f (E) = (E) = n (O) = n (O) = E \ O,

n=0

n=1

as g does not map to O. Hence, g maps f (E) to E \ O bijectively. Recall that O is the set of

points in A that are not mapped by g, O E and g has already mapped

f (E) onto E \ O.
(
1
g (x) if x E c ,
is a
Hence, g must map f (E)c to E c bijectively. So, the function h(x) =
f (x)
if x E,
bijection from A to B.
Alternate.

If g is onto, we have nothing to prove. So, assume that g is not onto. Put

O = A \ g(B), = g f and E = O (O) 2 (O) . Use 0 (O) to denote O. Notice that






(E) = g f (E) = (E) = n (O) = n (O) = E \ O,

n=0

n=1

as g does not map to O. Observe that : E E \ O is a bijection. Define h : A A \ O as

(
x,
if x A \ E,
h(x) =
(x), if x E.
Then, note that h is a bijection and hence h1 g is a bijection from B to A.

Alternate. Let F = {T A | g (f (T )c ) T c }.
g
g(f (T )c )

f (T )c

f (T )

Note that F . Put U = T . Then, U F , as

T F




g f (U )c = g f

T F

c 

=g





c 
= g f (T )c = g (f (T )c ) T c = U c .
f (T )

T F

T F

T F

T F

Thus, U is the maximal element of F . We claim that U c g (f (U )c ). To see this, take

x U c \ g (f (U )c ) and put V = U {x}. Then, f (U ) f (V ) and so f (V )c f (U )c . Thus

g f (V )c g f (U )c U c {x}c = V c ,

a contradiction to the maximality of U in F . So, g f (U )c = U c . Now, define h : A B as
(
f (x)
if x U,
h(x) =
1
g (x) else.
It is easy to see that h is a bijection.

42

CHAPTER 2. ADVANCED TOPICS IN SET THEORY

Exercise 2.4.6.

1. Give a one-one

r s

2 3
f (x) =
5r 3s

function from N to Q. Define f from Q to N as

if x = rs , gcd(r, s) = 1, r > 0, s > 0,
if x =

r
s , gcd(r, s)

= 1, r > 0, s > 0,

if x = 0

Argue that f is one-one. Apply Schr

oder-Bernstein theorem to prove that Q is equivalent
to N.
2. Give a one-one map from (0, 1) (0, 1) (0, 1). For each x (0, 1), let .x1 x2 be the
nonterminating decimal representations1 of x. For x = .x1 x2 x3 , y = .y1 y2 y3 , define

f (x, y) = .x1 y1 x2 y2 x3 y3 . Argue that f is an injection from (0, 1) (0, 1) to (0, 1).

Hence, show that (0, 1) is equivalent to (0, 1) (0, 1). Hence, show that R R is equivalent
to R.

3. Fix k N. Supply a one-one map from N to Nk . Use k distinct primes to supply a one-one
map from Nk to N. Conclude that Nk is equivalent to N.

4. Supply a bijection from (0, 1) to (1, 2) (3, 4) (5, 6) (7, 8) .

5. Show using Schr
oder-Bernstein that (0, 1) is equivalent to (0, 1].

Recall that every real number has a unique nonterminating decimal representation.

Chapter 3

Countability, cardinal numbers* and

partial order
3.1

Countable-uncountable

Definition 3.1.1. A set which is either finite or equivalent to N is called a countable set. A
set which is not countable is an uncountable set.
Definition 3.1.2. Let A be a countably infinite set. Then, by definition, there is a bijection
f : N A. So, we can list all the elements of A as f (1), f (2), . . .. This list is called an

enumeration of the elements of A.

Example 3.1.3.

1. We know that Z is a countably infinite set.

2. The set N \ [99] is a countably infinite set.

3. The set P(N) is uncountable by Cantors lemma.
4. Let S be the set of all 0-1-sequences x = x1 , x2 , . . .. Define f : S P(N) as f (x) = {n |
xn = 1}. Then f is a bijection. Hence, S is uncountable by Cantors lemma.

5. Let T = {x (0, 1) | x has a decimal expansion containing the digits 0 and 1 only}. Then
T is uncountable.

Proof. One proof follows by the previous idea.

Alternate. [Cantors diagonalization] If S is countable (clearly infinite), let x1 , x2 ,

be an enumeration. Let xn = .xn1 xn2 , where xni {0, 1}. Put ynn = 1 if xnn = 0

and ynn = 0, otherwise. Consider the number y = .y11 y22 S. Notice that for each n,
y 6= xn . That is, y S but it is not in the enumeration list. This is a contradiction.

Theorem 3.1.4. A set A is infinite if and only if it has a countably infinite subset.
Proof. Follows from Fact 1.2.23.12.
Lemma 3.1.5. Let A = {a1 , a2 , . . .} be countably infinite and B A. Then B is countable.
43

44

CHAPTER 3. COUNTABILITY, CARDINAL NUMBERS* AND PARTIAL ORDER

Proof. If B is finite then by definition, it is countable. So, assume that B is infinite. Hence, by
Theorem 3.1.4, B has a countable infinite subset, say C = {c1 , c2 , . . .}. Thus, f : A C B,

defined by f (ai ) = ci , i = 1, 2, . . . is a one-one map. On the other hand, IdB : B A is a

one-one map. Hence, by Schr
oder-Bernstein theorem, B and A are equivalent.
As a corollary, we have the following result.
Corollary 3.1.6. Let A be uncountable and A B. Then B is uncountable.
Proof. If B is countable, then by Lemma 3.1.5, A must be countable, a contradiction.
Theorem 3.1.7. If S is infinite, then P(S) is uncountable.
Proof. As S is infinite, there is a one-one map, say f : N S. Now, define a map g : P(N)


P(S) as g(A) = f (A). Then, g is clearly one-one and hence g P(N) is uncountable (by Cantors

lemma). Hence, P(S), being a superset is uncountable, by Corollary 3.1.6.

Theorem 3.1.8. Countable union of countable sets (union of a countable class of countable
sets) is countable.
Proof. Let {Ai }iN be a countable class of countable sets and put X = Ai . If X is finite
i

then we are done. So, let X be infinite. Hence, by Fact 1.2.23.12, there is a one-one map

f : N X. Define g : X N as g(x) = 2i 3k , if i is the smallest positive integer for which

x Ai and x appears at the k-th position in the enumeration of Ai . Then g is one-one. Now,

by Schr
oder-Bernstein theorem A is equivalent to N.

Theorem 3.1.9. The set P(N) is equivalent to [0, 1). Furthermore, P(N) is equivalent to R.
Proof. We already know a one-one map f : P(N) [0, 1) (see Examples 3.1.3.4 and 3.1.3.5).

Let r (0, 1). Consider the nonterminating binary representation of r. Denote by Fr the set of

positions of 1 in this representation. Now, define g : [0, 1) P(N) by g(r) = Fr , if r 6= 0 and

g(0) = . Then g is one-one. Now, by Schr
oder-Bernstein theorem P(N) is equivalent to [0, 1).

The next statement follows as [0, 1) is equivalent to (0, 1) (see Exercise 2.4.6.5) and (0, 1) is

equivalent to R.
Definition 3.1.10. [Cardinal numbers]
1. Cardinal numbers are symbols which are associated with sets such that equivalent sets
get the same symbol. By A we denote the cardinal number associated with A.
2. If there is an injection f : A B, then we write A B. By A B, we mean that B A.
3. If there is a bijection f : A B, then we write A = B.
4. We write [n] as n and as 0. Thus, for a finite set A, we have A = |A|.
5. We use 0 to denote N. If x = A is a cardinal number by 2x we mean P(A).
Fact 3.1.11.

1. If x, y, z are cardinal numbers such that x y and y z, then x z. In

other words it says, if there is a one-one map from A to B and a one-one map from B to
C, then there is a one-one map from A to C.

3.1. COUNTABLE-UNCOUNTABLE

45

2. Let x be any cardinal number. Then x  2x . This is Cantors lemma.

3. The cardinal numbers we know till now are 0, 1, 2, 3, . . . , 0 = N, 20 = R, 22 0 , . . ..

4. The cardinal numbers 0 = N, 20 = R, 22 0 , . . . are called the infinite cardinal numbers.

5. The generalized continuum hypothesis says that there is no cardinal number between an
infinite cardinal number x and 2x .
Example 3.1.12.

1. Let A be the set of all infinite sequences formed using 0, 1 and B be

the set of all infinite sequences formed using 0, 1, 2. Which one has larger cardinality and
why?
Ans: For (x) = x1 , x2 , A, let us define f (x) = .x1 x2 (binary). Then f : A [0, 1]

is a surjection and hence A [0, 1]. For (y) = y1 , y2 , B, let us define g(y) = .y1 y2

(decimal). Then g : B [0, 1] is one-one. So, B [0, 1] and hence B A. Also,

IdA : A B is an injection. Thus, A B.

2. Write R as a union of pairwise disjoint sets of size 5.

Ans: Note that R = (, 2] (2, 3] (3, 6] (6, 7) [7, ) and these five sets have the

same cardinality. Let f, g, h and t be bijections from (, 2] to (2, 3], (3, 6], (6, 7), [7, ),

respectively. Then R =

{r, f (r), g(r), h(r), t(r)}.

r(,2)

3. Let S be a countable set of points on the unit circle in R2 . Consider the line segments
Ls with one end at the origin and the other end at a point s S. Fix these lines. We

are allowed to rotate the circle anticlockwise (the lines do not move). Let T be another
countable set of points on the unit circle. Can we rotate the circle by an angle so that
no line Ls touches any of the points of T ?
Ans: Let ij be the angle of rotation required so that point pi touches line lj . The set of
all ij is countable and the set [0, 2) is uncountable.
4. A complex number is algebraic if it is a root of a polynomial equation with integer
coefficients. All other numbers are transcendental. Show that the set of algebraic
numbers is countable.
Ans: For each point a = (a0 , a1 , a2 , . . . , ak ) Zk (Z \ {0}), let Sk be the roots of the

S
S
polynomial equation a0 + a1 x + + ak xk = 0. Take Ak =
Sk and A =
Ak .
aZk (Z\{0})

k=1

Then A is the set of all algebraic numbers. The set A is countable as each Ak is countable
and the union is over a countable set.

5. Give a bijection from R to R \ Q.

Ans: Recall that Q can be enumerated. First get a bijection from R \ Q to itself. Now,
use train-seat argument to adjust Q.

46

CHAPTER 3. COUNTABILITY, CARDINAL NUMBERS* AND PARTIAL ORDER

3.2

Partial orders

Definition 3.2.1. Let f be a relation on X. We call f antisymmetric if (x, y) f and x 6= y

implies (y, x)
/ f . That is, both (x, y) and (y, x) cannot be in f , whenever x and y are distinct.

A relation on X is called a partial order if it is reflexive, transitive and antisymmetric. Let f

be a partial order on X and a, b X. We say a and b are comparable if either (a, b) f or

(b, a) f .

Example 3.2.2.

1. Let X = [5].

(a) The identity relation I is reflexive, transitive and antisymmetric. So, it is a partial
order.
(b) The relation I {(1, 2)} is also a partial order.
(c) The relation I {(1, 2), (2, 1)} is reflexive, transitive. But it is not antisymmetric, as
(1, 2) and (2, 1) are both in f .

(d) The relation I {(1, 2), (3, 4)} is also a partial order.
2. Let X = N. Then f := {(a, b) | a divides b} is a partial order.
3. Let X be a nonempty class of sets. Then f := {(A, B) | A B} is a partial order on X.
4. On R the set f := {(a, b) | a b 0} is a partial order. It is called the usual partial
order on R. List 5 elements of f . Usual partial order on a subset of R is defined similarly.

Exercise 3.2.3. Give a partial order on [5] with the

1. maximum number of elements in it.
2. minimum number of elements in it.
Definition 3.2.4.

1. [Partially ordered sets] The tuple (X, f ) is called a partially or-

dered set (in short, poset) if f is a partial order on X. It is common to use instead

of f . We say x y to mean that x and y are related. We say x < y to mean that x y
and x 6= y.

2. [Linear order] A partial order f on X is called a linear/complete/total order if either

(x, y) f or (y, x) f , for each pair x, y X.
3. [Linear ordered set] The tuple (X, f ) is said to be a linearly ordered set if f is a linear
order on X. You may imagine the elements of a linearly ordered set as points on a line.
4. [Chain] A linearly ordered subset of a poset is called a chain. The maximum size of a
chain is called the height of a poset.
5. [Anti-chain] Let (X, f ) be a poset and A X. Suppose that no two elements in A are
comparable. Then A is called an anti-chain. The maximum size of an anti-chain is called

the width of the poset.

Example 3.2.5.

1. The poset in Example 3.2.2.1a has height 1 (resp. chain is {1}) and

width 5 (respectively, anti-chain is {1, 2, 3, 4, 5}).

3.2. PARTIAL ORDERS

47

2. The poset in Example 3.2.2.1b has height 2 (resp. chain is {1, 2}) and width 4 (resp.
anti-chain is {2, 3, 4, 5} or {1, 3, 4, 5}).

3. The poset in Example 3.2.2.1d has height 2 (resp. chain is {1, 2} or {3, 4}) and width 3
(resp. anti-chain is {1, 3, 5}). Find other anti-chains?

4. The set N with the usual order is a linearly ordered set.

5. If (X, f ) is a nonempty linearly ordered set, then the height of X is X and the width of
X is 1.
6. The set N with a b if a divides b, is not linearly ordered. However, the set {1, 2, 4, 8, 16}

is a chain. This is just a completely ordered subset of the poset. There are larger chains,
for example, {2k | k = 0, 1, 2, . . .}. It has height N and width N .

7. The poset (P([5]), ) is not linearly ordered. However, {, [2], [5]} is a chain in it. So, is
{, {2}, [2], [3], [4], [5]}. Its height is 6. What is its width?

Definition 3.2.6. Let (, ) be a nonempty finite linearly ordered set (like the English al-

phabets with a b c z) and be the set of all words of elements of . For

a a1 a2 an , b b1 b2 bm define a b if
(a) a1 < b1 or
(b) ai = bi for i = 1, . . . , k and ak+1 < bk+1 or
(c) ai = bi for i = 1, . . . , n.

Then ( , ) is a linearly ordered set. This ordering is called the lexicographic or dictionary

ordering. Sometimes is called the alphabet set and is called the dictionary.

Exercise 3.2.7. Let D1 be the dictionary of words made from a, b, c and D2 be the dictionary
of words made from a, b, d. Are these two sets equivalent?
Discussion 3.2.8.

1. [Directed graph representation of a finite poset] Often we represent

a nonempty finite poset (X, ) by a picture. The process is described below.

(a) Put a dot for each element of X and label it.
(b) If a b, then join the dot for a and the dot for b by an arrow (a directed line).
(c) Put a loop at the dot of a, for each a X.
2. A directed graph representation of A = {1, 2, 3, 9, 18} with the divides relation (a b if
a | b) is given below.

48

CHAPTER 3. COUNTABILITY, CARDINAL NUMBERS* AND PARTIAL ORDER

18
9
2

3
1

Definition 3.2.9. [Hasse diagram] The Hasse diagram of a nonempty finite poset (X, ) is
a picture drawn in the following way.

1. Each element of X is represented by a point and is labeled with the element.

2. If a b then the point representing a must appear at a lower height than the point
representing b and further the two points are joined by a line.

3. If a b and b c then the line between a and c is removed.

Later, we shall show that for every nonempty finite poset (X, ) a Hasse diagram can be

drawn.

Example 3.2.10. Hasse diagram for A = {1, 2, 3, 9, 18} with the divides relation.
18
9
2

3
1

Exercise 3.2.11. Draw the Hasse diagram for [3] [4] under lexicographic order.
Proposition 3.2.12. Let F be a nonempty family of single valued relations such that either
f g or g f , that is, F is linearly ordered. Let h = f . Then the following are true.
f F

1. h is single valued.
2. dom(h) = dom(f ).
f F

3. rng(h) = rng(f ).
f F

4. If every element of F is one-one (from its domain to its range) then h is also one-one.
Proof. We shall only prove the first two items.

3.2. PARTIAL ORDERS

49

1. Let x dom(h) and (x, y), (x, z) h. Then there are f, g F, such that (x, y) f and
(x, z) g. As F is a chain, either f g or g f , say f g. Then, g is not single valued,

a contradiction.

2. Note that x dom(h) means (x, y) h for some y. This means (x, y) f for some f .
That is, x dom(f ), for a function f . This means x dom(f ).
f F

1. [Bounds] Let (X, f ) be a poset and A X. We say x X is an

Definition 3.2.13.

upper bound if for each z A, (z, x) f . In words, it means each element of A is x.

The term lower bound is defined analogously.

2. [Maximal] An element x A is maximal, if whenever there exists a z A with (x, z) f

then x = z. In other words, it means no element in A is strictly larger than x. The term
minimal is defined analogously.

3. [Maximum] An element x A is called the maximum of A, if x is an upper bound of A.

In other words, it means an upper bound of A which is contained in A. Such an element,
when it exists, is unique. The term minimum is defined analogously.
4. [Least upper bound] An element x X is called the least upper bound (lub) of A if
x is an upper bound of A and for each upper bound y of A, we have (x, y) f . In other

words x is the minimum/least of the set of all upper bounds of A. The term greatest

lower bound (glb) is defined analogously.

Example 3.2.14. Consider the two posets described by the following picture.
d
c

(X)

(Y)
Figure 3.1: Posets X and Y

1. Consider the poset in Figure 3.1 and let X = A = {a, b, c}. Then
(a) the maximal elements of A are b and c,
(b) the only minimal element of A is a,
(c) a is the lower bound of A in X,
(d) A has no upper bound in X,
(e) A has no maximum element,
(f) a is the minimum element of A,
(g) no element of X is the lub of A and
(h) a is the glb of A in X.

50

CHAPTER 3. COUNTABILITY, CARDINAL NUMBERS* AND PARTIAL ORDER

2. Consider the posets in Figures 3.1. Then, the definitions are illustrated in the following
table. Note that X = {a, b, c} and Y = {a, b, c, d}.
A = {b, c} X

A = {a, c} X

A = {b, c} Y

Minimal element(s) of A

b, c

b, c

Lower bound(s) of A in X/Y

Upper bound(s) of A in X/Y

doesnt exist

Maximum element of A

doesnt exist

doesnt exist

Minimum element of A

doesnt exist

doesnt exist

lub of A in X/Y

doesnt exist

glb of A in X/Y

Maximal element(s) of A

b, c

b, c

Exercise 3.2.15. Determine the maximal elements, minimal elements, lower bounds, upper
bounds, maximum, minimum, lub and glb of A in the following posets (X, f ).
1. Take X = Z with usual order and A = Z.
2. Take X = N, f = {(i, i) : i N} and A = {4, 5, 6, 7}.
Discussion 3.2.16. [Bounds of empty set] Let (X, f ) be a nonempty poset. Then each x X

is an upper bound for as well as a lower bound for . So, an lub for may or may not exist.

For example, if X = [3] and f is the usual order, then lub = 1. Whereas, if X = Z and f is
the usual order, then an lub for does not exist. Similar statements hold for glb.

Definition 3.2.17. A linear order f on X is said to be a well order if each nonempty subset
A of X has a minimal element (in A). We call (X, f ) a well ordered set to mean that f is a well
order on X. Note that a minimal element, if it exists, is a minimum in this case.
Example 3.2.18.
1. The set Z with usual ordering is not well ordered, as {1, 2, . . . , } is a nonempty subset
with no minimal element.

2. The ordering 0 1 1 2 2 3 3 describes a well order on Z.

3. The set N with the usual ordering is well ordered.
4. The set R with the usual ordering is not well ordered as the set (0, 1) doesnt have its
minimal element in (0, 1).
Exercise 3.2.19. Consider the dictionary order on N2 . Show that this is a well order.
Definition 3.2.20. Let (W, ) be well ordered and a W . The initial segment of a is
defined as I(a) := {x | x W, x < a}.

Example 3.2.21. Take N with the usual order. Then I(5) = [4] and I(1) = .
Theorem 3.2.22. [Principle of transfinite induction] Let (W, ) be a nonempty well ordered
set. Let A W which satisfies whenever I(w) A then w A. Then A = W .

3.2. PARTIAL ORDERS

51

Proof. If A 6= W , then Ac 6= . As W is well ordered, let s be the minimal element of Ac . So,

any element x < s is in A. That is, I(s) A. By the hypothesis s A, a contradiction.

Fact 3.2.23. The principle of transfinite induction is the principle of mathematical induction
when W = N.
Proof. To see this, let p(n) be a statement which needs to be proved by mathematical induction.
Put A = {n N | p(n) is true}. Assume that we have been able to show that I(n) A n

A. It means, we have shown that 1 A, as = I(1) A. Also we have shown that for n 2,

if {p(1), . . . , p(n 1)} are true then p(n) is true as well, as I(n) = [n 1].

Definition 3.2.24. [Product of sets] Recall that the product A1 A2 = {(x1 , x2 ) | xi Ai }

may be written as




f (1), f (2) f : [2] A1 A2 is a function with f (1) A1 , f (2) A2 .

Moreover, if A1 and A2 are finite sets then |A1 A2 | = |A1 | |A2 |. In general, we define the

product of the sets in {A }L , L 6= , as

Y



A = f | f : L A is a function with f () A , for each L .

Example 3.2.25.

1. Take L = N and An = {0, 1}. Then

A is the class of functions

f : L {0, 1}. That is, it is the class of all 0-1-sequences.

2. By definition, product of a class of sets among which one of them is is empty.

What about product of a class of sets in which no one is empty? Is it nonempty? This could
not be proved using the standard set theory. In fact, it is now proved that this question cannot
be answered using the standard set theory. So, a new axiom, called the axiom of choice, was
introduced.
Axiom 3.2.26. [Axiom of Choice] The product of a nonempty class of nonempty sets is
nonempty.
Proposition 3.2.27. [injection-surjection] Let A and B be nonempty sets. Then, there is a
surjection g : A B if and only if there is an injection f : B A.
Proof. Let g : A B be onto. We shall find an injection from B to A. To start with, notice that
Q 1
Q 1
g (b).
g (b) 6= . Let f
for each b B, the set g1 (b) 6= . Then, by axiom of choice
bB

bB

Then, by Definition 3.2.24, f : B A is a function. As g is a function, g1 (b)s are disjoint and

hence f is one-one.

Conversely, let f : B A be one-one. Fix an element b B. Define g : A B as

(
f 1 (x), if x f (B),
g(x) =
b,
if x A \ f (B).
Observe that g is onto.

52

CHAPTER 3. COUNTABILITY, CARDINAL NUMBERS* AND PARTIAL ORDER

Definition 3.2.28. A class F of sets is called a family of finite character if it satisfies:

A F if and only if each finite subset of A is also in F .
Example 3.2.29.

1. { } is a family of finite character.

2. Power sets are families of finite character.

3. {, {1}, {2}} is a family of finite character.
4. If A B = , then P(A) P(B) is a family of finite character.
5. The set {} {{a} | a 6= 0, a R} is a family of finite character. This is the class of
linearly independent sets in R.

6. Let V be a non trivial vector space and F be the class of linearly independent subsets of
V. Then F is a family of finite character.
Exercise 3.2.30.

1. Let L = A1 = A2 = A3 = [3]. Is the set

A equal to the class of

functions f : [3] [3]? Give reasons for your answer.

Q
An has 6 elements. Give another.1
2. Give sets An , n N such that
nN

Some equivalent axioms of axiom of choice

[Axiom of choice] Cartesian product of a nonempty collection of nonempty sets is

nonempty.
[Zorns lemma] A partially ordered set in which every chain has an upper bound, has

a maximal element.
[Zermelos well ordering principle] Every set can be well ordered.
[Hausdorff s maximality principle] Every nonempty partially ordered set contains a

maximal chain.
[Tukeys lemma] Every nonempty family of finite character has a maximal element.

Exercise 3.2.31.

1. Does there exist a poset with exactly 5 maximal chains of size (number

of elements in it) 2, 3, 4, 5, 6, respectively and 2 maximal elements? If yes, draw the Hasse
diagram. If no, argue it.
2. Let (X, f ) be a nonempty poset and =
6 Y X. Define fY = {(a, b) f | a, b Y }. Show
that fY is a partial order on Y . This is the induced partial order on Y .

3. Apply induction to show that a nonempty finite poset has a maximal element and a minimal
element.
Discussion 3.2.32. [Drawing the Hasse diagram of a finite poset (X, f )] Let x1 , . . . , xk be the
minimal elements of X. Draw k points on the same horizontal line and label them x1 , . . . , xk .
Now consider Y = X \ {x1 , . . . , xk } and fY . By induction, the picture of (Y, fY ) can be drawn.

Put it above those k dots. Let y1 , . . . , ym be the minimal elements of Y . Now, draw the lines
(xi , yj ) if (xi , yj ) f . This is the Hasse diagram of (X, f ).
1

When we ask for more than one example, we encourage the reader to get examples of different types, if

possible.

3.2. PARTIAL ORDERS

53

Discussion 3.2.33. [Existence of Hamel basis] Let V be a vector space with at least two
elements. Recall that the collection F of linearly independent subsets of V is a family of finite

character. Recall that a basis or a Hamel basis is a maximal linearly independent subset of V.
As V has at least 2 elements, it has a nonzero element, say a. Then {a} F. Hence, F 6= .

Thus, by Tukeys lemma, the set F has a maximal element. This maximal set is the required
basis. Hence, we have proved that every vector space with at least 2 elements has a Hamel basis.
Exercise 3.2.34.

1. Let n N. Define Pn = {k N | k divides n}. Define a relation n

on Pn as n = {(a, b) | a divides b}. Show that (Pn , n ) is a poset, for each n N. Give

a necessary and sufficient condition on n so that (Pn , n ) is a completely ordered set.

n
o n
o
2. Take X = (1, 1), (1, 2), (1, 3), . . . (2, 1)(3, 1), (4, 1), . . . . The ordering defined is
n
n

o [

o
f=

(1, m), (1, n)

(m, 1), (n, 1) .
m, n N
mn

m, n N
mn

Does X have any maximal or minimal elements? Is X linearly ordered? Is it true that
every nonempty set has a minimal element? Is it true that every nonempty set has a
minimum? What type of nonempty sets always have a minimum?
3. Prove or disprove:
(a) There are at least 5 functions f : R R which are partial orders.

(b) Let S be the set of sequences (xn ), with xn {0, 1, . . . , 9}, for each n N, such that
if xk < xk+1 , then xk+1 = xk+2 = }. Then S is countable.

(c) Take N with usual order. Then the dictionary order on N2 is a well order.
(d) Let S be the set of all non-increasing sequences made with natural numbers. Then S
is countable.
(e) Let S be the set of all nondecreasing sequences made with natural numbers. Then S
is countable.
(f ) Take N with usual order and N2 with the dictionary order. Then any nonempty subset
of N2 which is bounded above has a lub.
(g) Every nonempty countable linearly ordered set is well ordered with respect to the same
ordering.
(h) Every nonempty countable chain which is bounded below, in a partially ordered set,
is well ordered with respect to the same ordering.
(i) The set Q can be well ordered.
(j) For a fixed n N, let An and Bn be non-empty sets and let Rn be a one-one relation
from An to Bn . Then, Rn is a one-one relation.
n

(k) Let S be the set of words with length at most 8 using letters from {3, A, a, b, C, c}. We
want to define a lexicographic order on S to make it a dictionary. There are more
than 500 ways to do that.
(l) An infinite poset in which each nonempty finite set has a minimum, must be linearly
ordered.

54

CHAPTER 3. COUNTABILITY, CARDINAL NUMBERS* AND PARTIAL ORDER

(m) A nonempty finite poset in which each nonempty finite set has a minimum, must be
well ordered.
(n) An infinite poset in which each nonempty finite set has a minimum, must be well
ordered.
4. Let S = {(x, y) : x2 + y 2 = 1, x 0}. It is a relation from R R. Draw a picture of the
inverse of this relation.

5. Construct the Hasse diagram for the relation on P({a, b, c}).

6. Draw the Hasse diagram for the partial order describing the divides relations on the set
{2, 3, 4, 5, 6, 7, 8}.
7. Draw the Hasse diagram of {1, 2, 3, 6, 9, 18} with divides relation.
(a) What is its height? What is its width.
(b) Let A = {2, 3, 6}. What are the maximal elements, minimal elements, maximum,
minimum, lower bounds, upper bounds, glb and lub of A.

8. Show that the following three definitions are equivalent.

(a) A set A is finite if either A = or A = [n], for some n N.

(b) [Tarski] A set A is finite if and only if every nonempty family of subsets of A has a
minimal element.
(c) [Dedekind] A set is infinite if it is equivalent to a proper subset of itself. A set is
finite if it is not infinite.

9. Let (X, f ) be a nonempty poset. Show that there exists a linear order g on X such that
f g.
10. Let G be a non-Abelian group and H be an Abelian subgroup of G. Show that there is a
maximal Abelian subgroup J of G such that H J.
11. Let F be a family of finite character and B be a chain in F . Show that A F .
AB

12. Let A 6= and F be a field. Let

FA

FA

:= {f : f is a function from A to F}. Let := {f

: {a A : f (a) 6= 0} is finite}. Show that is a vector space over F with respect to

point-wise addition of functions and point-wise scalar multiplication. Also show that every
vector space V is isomorphic to for some suitable choice of A.
13. Let X be a vector space and A be a nonempty linearly independent subset of X. Let S X
satisfy span(S) = X. Show that a Hamel basis B such that A B S.

14. Let (L, ) be a nonempty linearly ordered set. Prove that W L such that well orders

W and such that for each x L, there is a y W satisfying x y. For example, for

L = R, we can take W = N.

15. Show that R is not a finite dimensional vector space over Q. Hint: Assume that R as
a vector space over Q has dimension k. Argue that R is isomorphic to Qk and so it is
countable, a contradiction.

3.2. PARTIAL ORDERS

55

16. Let A be a nonempty set. Then there is an element a which is not in A.

17. Let A be a nonempty set. Then there exists B such that A B = and A = B.
18. Let A and B be two nonempty sets. Show that there is a set C such that C A = and
C = B.

19. Let A and B be nonempty sets. Put a = A and b = B. Then show that either a b or
b a.

20. Let a = A and b = B, where A B = . Then we define a + b as A B and ab as A B.

(a) Let a be an infinite cardinal number. Show that a + a = a and aa = a.
(b) Let a, b, c be cardinal numbers. Show that a b {a + c b + c, ac bc}.
21. Suppose that u v are two infinite cardinal numbers. Then show that u + v = v and
uv = v.

56

CHAPTER 3. COUNTABILITY, CARDINAL NUMBERS* AND PARTIAL ORDER

Chapter 4

Logic
4.1

Propositional logic

We study logic to differentiate between valid and invalid arguments. An argument is a set of
statements which has two parts: premise and conclusion. There can be many statements in the
premise. Conclusion is just one statement. An argument has the structure
premise: Statement1 , . . ., Statementk ;

therefore conclusion: Statementc .

Consider the following examples.

Statement1 : If today is Monday, then Mr. X gets Rs. 5.
Statement2 : Today is Monday.

Statementc : Therefore, Mr. X gets Rs. 5 (statementc ).

Statement1 : If today is Monday, then Mr. X gets Rs. 5.
Statement2 : Mr. X gets Rs. 5.

Statementc : Therefore, today is Monday.

Statement1 : If today is Monday, then Mr. X gets Rs. 5.
Statement2 : Today is Tuesday.

Statementc : Therefore, Mr. X gets Rs. 5.

Statement1 : If today is Monday, then Mr. X gets Rs. 5.
Statement2 : Today is Tuesday.

Statementc : Therefore, Mr. X does not get Rs. 5.

We understand that the first one is a valid argument, whereas the next three are not. In order
to differentiate between valid and invalid argument, we need to analyze an argument. And in
order to do that, we first have to understand what is a statement. A simple statement is an
expression which is either false or true but not both. We create complex statements from the
old ones by using and, or and not.
For example, today is Monday is a statement. Today is Tuesday is also a statement. Today
is Monday and today is Tuesday is also a statement. Today is not Monday is also a statement.
57

58

CHAPTER 4. LOGIC

One way to analyze an argument is by writing it using symbols. The following definition
captures the notion of a statement.
Definition 4.1.1.

1. [Atomic formulae and truth values] Consider a nonempty finite set

of symbols F. We shall call an element of F as an atomic formula (also called atomic

variable). (These are our simple statements). The truth value of each element in F is
exactly one of T (for TRUE) and F (for FALSE). Normally, we use symbols p, q, p1 , p2 , . . .

for atomic formulae.

2. [Operations to create new formulae] We use three symbols (called disjunction/or),
(called conjunction/and), and (called negation) to create new formulae. The
way they are used and the way we attribute the truth value to such a new formula is
described below.
If p and q are formulae, then p q, p q, and p are formulae. The truth value of p q is

defined to be T when the truth values of both p and q are T . Its truth value is defined to be F

in all other cases. The truth value of p q is defined to be T when the truth values of at least
one of p and q are T . Its truth value is defined to be F when the truth values of both p and q

are F . The truth value of p is defined to be T if the truth value of p is F . The truth value of

p is defined to be F if the truth value of p is T .

Understanding , and

The following tables describe how we attribute the truth values to p q, p q and p.
pq

pq
T

p
F
T

How do we read these tables? Look at row 3 of the leftmost table (exclude the header). It
tells that the formula p q takes the truth value F if p takes truth value F and q takes T .
Remark 4.1.2. We use brackets while creating new formulae to make the meaning unambiguous.
For example, the expression p q r is ambiguous, where as p (q r) is unambiguous.
Definition 4.1.3.

1. Sometimes we write f (p1 , . . . , pk ) is a formula to mean that f is a

formula involving the atomic formulae p1 , . . . , pk .

2. Let f (p1 , . . . , pk ) be a formula. Then, the truth value of f is determined based on the truth
values of the atomic formulae p1 , . . . , pk . Since, there are 2 assignments for each pi , 1 i

k, there are 2k ways of assigning truth values to these atomic formulae. An assignment of
truth values to these atomic formulae is nothing but a function A : {p1 , . . . , pk } {T, F }.

3. By saying T F T is an assignment to the atomic variables p, q, r, we mean that the truth

value of p is T , that of q is F and that of r is T . Keeping this in mind, all possible

4.1. PROPOSITIONAL LOGIC

59

assignments to p, q, r are listed below. (Notice that, it is in the dictionary order, that is,
F F F appears before F F T in the list as if they are words in a dictionary. The reader will
notice that in the table given above, we have followed the reverse dictionary order while
writing a truth table, which is natural to us. This should not create any confusion.)
p

4. A truth table for a formula f (p1 , . . . , pk ) is a table which systematically lists the truth
values of f under every possible assignment of truth values to the involved atomic formulae.
The following is a truth table for the formulae p (q r).
p

q r p (q r)
F

5. In the previous table, if we fill the fourth column arbitrarily using T s and F s, will it be
a truth table of some formula involving p, q and r? We shall talk about it later.
We have already noted that we use , and to create new formulae from old ones. Some of

them will indeed be very important.

Definition 4.1.4. [Conditional formulae]

1. [p implies q] If p and q are formulae, then the formula (p) q is denoted by p q (read
as p implies q). Its truth table is

(p) q

60

CHAPTER 4. LOGIC
Observe
a) p q takes the truth value F if and only if p takes the truth value T and q takes
the truth value F .

b) If under some assignment p q takes the truth value T and that in this assignment

p is T , then it follows that in this assignment q must be T . This is why p q is called

if p then q.

c) Other phrases used for if p then q are p is sufficient for q or p only if q or q is a

necessary condition for p.
d) We sometimes use p q to mean q p.
2. [p if and only if q] The formula (p q) (called p if and only if q) means (p q)(q p).
Note that (p q) takes the truth value T whenever p and q take the same truth values

and takes the truth value F whenever p and q take different truth values. Its truth table
is
pq

3. [Converse/Contrapositive] The formula q p is called the converse of p q and the

formula q p is called the contrapositive of p q.

Discussion 4.1.5. [Understanding a conditional formula] When we assign different English

statements to the involved atomic formulae, we get an English statement corresponding to those
formulae. For example, for the formula p q, consider the following statements:

p: you attend the class.

q: you understand the subject.

Then, p q is the statement if you attend the class, then you understand the subject. The

formula p q is true under the following three cases.

1. p is true and q is true: this means you attend the class and understand the subject.
2. p is false and q is false: this means you do not attend the class and do not understand
the subject.
3. p is false and q is true: this means you do not attend the class but understand the subject.
The formula p q is false if p is true and q is false, which means you attend the class and do

not understand the subject.

Definition 4.1.6. [Connectives] The symbols , , , and are called connectives. The
set of well formed formulae (wff ) are defined inductively. Each atomic variable is a wff. If f

and g are two wff, then f g, f g, f , f g, and f g are wff. Brackets are used to avoid
ambiguity.

4.1. PROPOSITIONAL LOGIC

Example 4.1.7.

61

1. p q, q, p q are not wff, as they do not make sense.

2. p q r is not a wff as it is not clear what it means. We use brackets to get p (q r) or

(p q) r which are wff.

3. (p q) r, (p q) r, (p q) are wff.
Did you notice?
The connectives , , , and always connect two old formulae to create a new one. This
is why they are called binary connectives. The connective is used on a single old formula

to give a new one. So, it is called a unary connective.

Definition 4.1.8. Let A be the set of assignments to the variables p1 , . . . , pk . A function

k

f : A {T, F } is called a truth function. Since |A| = 2k , there are 22 such truth functions.

Example 4.1.9. The table on the left describes a truth function f and that on the right
describes the truth table for a particular formula.

Exercise 4.1.10.

(p q) (p (q))

1. Draw a truth table for the formula p p (p q) .

2. Can both the formulae p q and q p be false for some assignment on p and q?
Definition 4.1.11.

1. [Contradiction and tautology] A contradiction (F ) is a formula

which takes truth value F under each assignment. For example, p p. A tautology (T )

is a formula which takes truth value T under each assignment. For example, p p.

2. [Equivalence of formulae] Two formulae f and g are said to be equivalent, denoted

f g, if they have the same truth table involving all the atomic variables of both f and g.

That is, if both f and g carry the same truth values under each assignment to the involved
atomic variables.
Example 4.1.12.

1. Is p q q p? Yes, because they have the same truth tables.

p

f = p q g = q p
T

62

CHAPTER 4. LOGIC
2. Is p p (q (q))? Yes, because they have the same truth tables.

Remark 4.1.13.

f = p g = p (q (q))
T

1. There is another way to establish equivalence of two formulae f and g.

We show that f has a truth value T (or F ) if and only if g has the same truth value. For
example, to show that p q q p, proceed in the following way.
Step 1: Suppose that p q has a truth value F for an assignment a. Then a(p) = T and

a(q) = F . But then, under that assignment, we have q is T and p is F . That is, under

a, we have q p is F .

Step 2: Suppose that p q has a truth value T for an assignment a.

Then a

{T T, F T, F F }. Under T T , we have p is F and q is F , so that q p is T . Under

F T , we have p is T and q is F , so that q p is T . Under F F , we have p is T and

q is T , so that q p is T .
Thus, both are equivalent.

2. Let f (p1 , . . . , pk ) be a formula and q1 , . . . , qr be some new atomic variables. Then f

f (q1 (q1 )) (qr (qr )). This can be argued using induction. Thus f can be

viewed as a formula involving atomic variables p1 , . . . , pk , q1 , . . . , qr .

3. We have seen that
(a) p q p q, and

(b) p q (p q) (q p).
Thus, the connectives , and are enough for writing a formula in place of the 5

connectives , , , and .

4. Recall that a formula on variables p, q and r is a truth function. So there are exactly
3

22 = 28 nonequivalent formulae on variables p, q and r.

Exercise 4.1.14. Is p p q q?
Definition 4.1.15. [Substitution instance] Suppose B is a formula which involves some variables including p. Then, substituting a formula A for each appearance of the variable p in
B, gives us a new formula. This new formula is called a substitution instance of B. We
may substitute more than one variables, simultaneously. Note that A may involve old and new
variables.
Example 4.1.16. Let B: (p q) p. We substitute p q for p, and p for q, in B to obtain

the following substitution instance of B.

(p q) p (p q)

4.1. PROPOSITIONAL LOGIC

63

The following result is one of the most fundamental results of the subject.
Theorem 4.1.17. Any substitution instance of a tautology is a tautology.
Proof. Let P (p1 , . . . , pk ) be a tautology. Suppose that we replace each occurrence of p1 by a
formula f to obtain the formula R. Consider all the atomic variables involved in P and f .
View P and R as formulae involving all these atomic variables. Let a be an assignment to these
atomic variables.
If f takes the value T on a, then the value of R on a is nothing but the value of P (T, p2 , . . . , pk )
on a, which is T as P is a tautology.
If f takes the value F on a, then the value of R on a is nothing but the value of P (F, p2 , . . . , pk )
on a, which is T as P is a tautology. Thus, R takes the value T under each assignment.
Exercise 4.1.18. Show that any substitution instance of a contradiction is a contradiction.
Definition 4.1.19. [Functionally complete] A subset S of connectives is called functionally complete/adequate, if each formula has an equivalent formula written only using the
connectives in S.
Example 4.1.20. We already know that S = {, , } is adequate.
Exercise 4.1.21.

1. Determine which are adequate. (i) {, } (ii) {, }.

2. Fill in the blanks to prove that f g if and only if f g is a tautology.

Proof. Assume that f g. Let b be an assignment. Then, the value of f and g are
under b. Thus, the value of f g is

we see that f g is a

under b. As b is an

assignment,

Therefore, if f is T under b, then g is T under b. That is, f g and g f are both T

under b. Thus, f g is T under the assignment b.
Conversely, suppose that f g is a
under which

and

. Assume that f 6 g. Then, there is

take different

So, suppose that f takes T and g takes F under b. Then

is F under b and hence

f g takes F under b, a contradiction. A similar contradiction is obtained if f takes F

and g takes T under b.

The proof of the next result is left as an exercise for the readers.
Proposition 4.1.22. [Rules] If p, q, r are formulae, then
1. p q q p, p q q p (commutative)
2. p (q r) (p q) r, p (q r) (p q) r (associative)
3. p (q r) (p q) (p r), p (q r) (p q) (p r) (distributive)
4. (p q) p q, (p q) p q (De Morgans law)
5. p p p, p p p (idempotence)

64

CHAPTER 4. LOGIC
6. F p p, F p F
7. T p T, T p p
8. (p) p
9. p (p q) p, p (p q) p (absorption law)

Proof. First six may be proved suing direct arguments and the rest by using the first six.

Exercise 4.1.23. Does the absorption law imply p (p (q)) p and p (p (q)) p?

Discussion 4.1.24. The above rules can be used to simplify a formula or to show equivalence
of formulae. For example,

p (q r) p (q r)

as p p (p) q

p q r

Associativity

(p q) r

De Morgans law

(p q) r

as p p (p) q

Did you notice?

There are 3 ways to prove f g.
1. Using truth table.
2. Arguing that f is false under an assignment (of the variables involved in both) if and
only if g is false under the same assignment.
3. Using some of the above rules and by reducing f to g or g to f .

4.1. PROPOSITIONAL LOGIC

65

Experiment
Consider the variables p, q, r.
Give a formula which takes value T only on the assignment T T T .
Give a formula which takes value T only on the assignment T T F .
Give a formula which takes value T only on the assignment F T F .

(p q (r))

Give a formula which takes value T only on the assignments T T F and F T F .

Give a formula which takes value T only on the assignments T F T , T T F and T F F .
Give a formula f which takes value T only on the assignments F T F and F F F or whose
truth table is the following
p

Lemma 4.1.25. Let f be a truth function involving the variables p1 , . . . , pk . Then, there is a
formula g involving p1 , . . . , pk , whose truth table is described by f .
Proof. If T
/ rng f , then write q = p1 p1 p2 pk . Otherwise, collect all those assignments

b such that f (b) = T . Call this set A1 . For each b A1 , define a formulae q = r1 r2 rk ,

where for 1 j k,

rj =

pj

if b(pj ) = T

pj otherwise.

Then, the formulae q takes the value T only on the assignment b. Thus, taking the disjunctions
of all such qs related to each b A1 , we get the required result.
Exercise 4.1.26. Illustrate 4.1.25 with the truth function f
p

Definition 4.1.27. [Normal forms] An atomic formula or its negation is called a literal. We
say that a formula f is in disjunctive normal form (in short, DNF) if it is expressed as a
disjunction of conjunctions of literals. We say that a formula f is in conjunctive normal form
(in short, CNF) if it is expressed as a conjunction of disjunctions of literals.

66

CHAPTER 4. LOGIC

Example 4.1.28. p, p q, p q, (p q) r, (p q) (q r) (r s) are in DNF. Write

5 formulae in CNF involving p, q, r.

Theorem 4.1.29. Any formula is equivalent to a formulae in DNF. Similarly, Any formula is
equivalent to a formulae in CNF.
Proof. The proof of the first assertion follows from Lemma 4.1.25. For the second assertion, we
can write one proof in a similar way.
An alternate proof: take f , consider f , get a DNF P for f , and consider P .
Exercise 4.1.30. Write all the truth functions on two variables and write formulae for them.
Definition 4.1.31. [Principal connectives] Let h be a formula. A principal connective in
h is defined in the following way.
1. If h is expressed in a format f , then is the principal connective of h.
2. If h is expressed in a format f g, then is the principal connective of h.
3. If h is expressed in a format f g, then is the principal connective of h.
Exercise 4.1.32. Use induction on the number of connectives to show that any formula is
equivalent to a formulae in DNF and a formula in CNF.
Definition 4.1.33. [Dual] The dual P of a formula P involving the connectives , , is
obtained by interchanging with and the special variable T with the special variable F.
Example 4.1.34. Note that the dual of (p q) r is (p q) r.
Lemma 4.1.35. Let A(p1 , . . . , pk ) be a formula in the atomic variables pi involving connectives
, and . If A(p1 , . . . , pk ) is obtained by replacing pi with pi in A, then A(p1 , . . . , pk )
A (p1 , . . . , pk ).

Proof. Use induction on the number of connectives. If A = B C, then

A = B C B(p1 , . . . , pk ) C(p1 , . . . , pk )
(B C)(p1 , . . . , pk ) = A(p1 , . . . , pk ).
The remaining parts are similar and hence left for the reader.
Theorem 4.1.36. Let f, g be formulae using connectives , and . If f g, then f g .
Proof. By Lemma 4.1.35, we note that
f (b) = f (b) = g(b) = g (b) for any assignment b.
Thus, f g .

4.1. PROPOSITIONAL LOGIC

67

Discussion 4.1.37. [Tree representation] A formula can be represented by a tree. For example,
(r q) (q p) has the following representation.

q
Definition 4.1.38. [Polish notation] A formula may be expressed using Polish notation. It
is defined inductively as follows.
Let P (f ) denote the Polish notation of f . Then P (f g) is P (f )P (g), P (f g) is P (f )P (g),
and P (f ) is P (f ).

This notation does not use brackets. Here the connectives are written in front of the expressions
they connect. Advantage: it takes less space for storage. Disadvantage: its complicated look.
Example 4.1.39. In Polish notation (r q) (q p) becomes rq qp.
Exercise 4.1.40. Write a formula involving 8 connectives and the variables p, q, r. Draw its
tree. Write its Polish notation.
Definition 4.1.41.

1. [Satisfiable] A formula is satisfiable if it is not a contradiction.

2. [Order of operations] To reduce the use of brackets, we fix the order of operations: ,
, , , .

Discussion 4.1.42. There is another way of making a truth table for a formula. Consider
(p q) r. Draw a table like the following and give the truth values to the atomic formulae.

Evaluate the connectives for the subformulae one by one. In this example, the sequence of
column operations is: 5, 2, 4.
(p q)

q)

(p

Definition 4.1.43. [Inference] We say g is a logical conclusion of {f1 , , fn } if (f1

f2 fn ) g is a tautology. We denote this by {f1 , . . . , fn } g. At times, we write

f1 , . . . , fn g to mean {f1 , . . . , fn } g. Here, g is called the conclusion and {f1 , . . . , fn } is
called the hypothesis/premise.

68

CHAPTER 4. LOGIC

Example 4.1.44.

1. Consider the following three statements.

A : if x = 4, then discrete math is bad;

B : discrete math is bad;
C : x = 4.
Does C logically follow from A, B?
Ans: No. Denote x = 4 by p and discrete mathematics is bad by q. Then, the
above question is the same as asking whether {p q, q} p is true. That is, whether

P (p, q) := ((p q) q) p is a tautology.

To find that, suppose that there is an assignment for which P takes the value F . So, for
that assignment, p must be F and (p q) q must be true.
As (p q) q is true, q must be T . So, the assignment must be F T . Notice that p q

has a value T with this assignment. Thus, P (p, q) takes F under F T . Hence, it is not a
tautology. So, C does not logically follow from A, B.

2. Consider the following three statements.

A : if discrete math is bad, then x = 4;
B : discrete math is bad;
C : x = 4.
Does C logically follow from A, B?
Ans: Yes. Denote x = 4 by p and discrete mathematics is bad by q. Then, the
above question is the same as asking whether {q p, q} p is true. That is, whether

P (p, q) := ((q p) q) p is a tautology.

To find that, suppose that there is an assignment for which P takes the value F . So, for
that assignment, p must be F and (q p) q must be true.
As (q p) q is true, q must be T and q p must be T . So, the assignment must be

F T . But we see that q p has a value F with this assignment. This is a contradiction.

Thus, there is no assignment for which P (p, q) takes F . Hence, it is a tautology. So C

logically follows from A, B.
Definition 4.1.45. We write f g to mean f g and g f .
Did you notice?
Let f, g, h be some formulae. Then f, g h means that whenever f and g are T , h is also
T . That is, if f and g are T under an assignment, then h is T under that assignment.
Thus f g is the same as f g.
Example 4.1.46.

1. Show that { , , } .

4.1. PROPOSITIONAL LOGIC

69

Ans: Suppose is F . Then is T and is F . Assume that all the propositions in

the hypothesis are true. As is F and is T , must be F . Continuing, we get is

F , a contradiction.

2. Determine validity of the argument.

The meeting can take place if all members are informed in advance and there is quorum
(a minimum number of members are present). There is a quorum if at least 15 members
are present. Members would have been informed in advance if there was no postal strike.
Therefore, if the meeting was canceled, then either there were fewer than 15 members
present or there was a postal strike.
A : Let us denote the different statements with symbols, say
m: the meeting takes place;
a: all members are informed;
f : at least fifteen members are present;
q: the meeting had quorum;
p: there was a postal strike.


So, we reformulate the problem: whether (qa) m, f q, p a m (f p)?

From first two statements, we get (f a) m. Considering the third statement, we get
(f p) m. The conclusion is the contrapositive of this statement.

Alternate. Suppose that conclusion is F . This means that m (f p) takes the



value F and (q a) m, f q, p a takes the value T .

The first one implies that f p takes the value F and m takes the value T . Hence, we
see that the variables m, f and p take values F, T and F , respectively.

The second one implies that all the three expressions (q a) m, f q, and p a

take the value T . Since the second statement takes the value T and f has the value T ,

we see that q has to take the value T . Similarly, using the third statement, we see that a
has to take the value T . So, we see that the first statement (q a) m takes the value

T with the assignment of both q and a being T . So, we must have m to have the value T ,
contradicting the value F taken by m in the previous paragraph.
Exercise 4.1.47.

1. List all the nonequivalent formulae involving variables p and q which

take truth value T on exactly half of the assignments.

2. We assume F T . Let f and g be two truth functions on the variables p1 , . . . , p9 . Suppose
that for each assignment a, we have f (a) g(a). Does this imply f g is a tautology ?

3. Let f and g be two formulae involving the variables p1 , . . . , pk . Prove that f g (the
same truth table) if and only if f g is a tautology.

4. Without using , write an equivalent simplified statement of (p q) p (q r) .
5. Determine which of the following are logically equivalent.




(a) p (r s) (q r) s .

70

CHAPTER 4. LOGIC




(b) (p r) (s p) p (s r) .
(c) q s.

(d) s (q r) (q s) r .




(e) (p s) (q p) p (q s) .

6. Let p be a formula written only using connectives , and and involving the atomic

variables p1 , , pk , for some k. Show that the truth value of p is T under the assignment

f (pi ) = T , for all i.

7. Is {, , } adequate?
8. Verify the following assertions.
(a) P Q P

(b) P P Q

(c) P P Q

(d) (P Q) P

(e) P, P Q Q
(f ) P, P Q Q

(g) Q, P Q P

(h) P Q, Q R P R

(i) P Q, P R, Q R R

(j) P Q (P Q) (P Q)

(k) {p q, p q} q r
(l) {p q, p} q

(m) {p0 p1 , p1 p2 , . . . , p9 p10 } p0 p5 .



(n) (p q) r, s q, t, p t, (p r) s q.


(o) p q, r s, s t, q s, s, (p r) u, w t u w.

9. If H is a set of formulae, then H if and only if H {} .

10. Prove the equivalence of the following in three different ways (truth table, simplification,
each is a logical consequence of the other): p (q r) (p q) r.
11. Determine which of the following conclusions are correct.
(a) If the lecture proceeds, then either black board is used or the slides are shown or the
tablet pc is used. If the black board is used, then students at the back bench are not
comfortable in reading the black board. If the slides are shown, then students are
not comfortable with the speed. If the tablet pc is used, then it causes lots of small
irritating disturbances to the instructor. The lecture proceeds and the students are
comfortable. So, it is deduced that the instructor faces disturbances.
(b) There are three persons Mr X, Mr Y and Mr Z making statements. If Mr X is wrong,
then Mr Y is right. If Mr Y is wrong, then Mr Z is right. If Mr Z is wrong, then Mr
X is right. Therefore, some two of them are always right.

4.2. PREDICATE LOGIC

71

12. Consider the set S of all nonequivalent formulae written using two atomic variables p and
q. For f, g S, define f g if f g. Prove that this is a partial order on S. Draw its
Hasse diagram.

13. Consider the set S of all nonequivalent formulae written using three atomic variables p, q, r.
For f, g S define f g if f g. Let f1 and g1 be two formulae having the truth tables
p
T
T
T
T
F
F
F
F

q
T
T
F
F
T
T
F
F

r
T
F
T
F
T
F
T
F

f1
T
F
T
T
F
T
F
F

p
T
T
T
T
F
F
F
F

q
T
T
F
F
T
T
F
F

r
T
F
T
F
T
F
T
F

g1
T
F
T
F
T
F
T
F

How many nonequivalent formulae h are there such that {f1 , g1 } h?

14. How many assignments of truth values to p, q, r and w are there for which (p q)


r w is true? Guess a formula in terms of the number of variables.

15. Check the validity of the argument. If discrete math is bad, then computer programming

is bad. If linear algebra is good, then discrete math is good. If complex analysis is good,
then discrete math is bad. If computer programming is good, then linear algebra is bad.
Complex analysis is bad and hence, at least one more subject is bad.

4.2

Predicate logic

Definition 4.2.1. A k-place predicate or propositional function p(x1 , . . . , xk ) is a statement involving the variables x1 , . . . , xk . A truth value can be assigned to a predicate p(x1 , , xk )
for each assignment of x1 , . . . , xk from their respective universe of discourses (in short, UD)
(the set of values that xi s can take is the i-th UD).
Example 4.2.2. Let p(x) mean x > 0. Then p(x) is a 1-place predicate on some UD. Let
p(x, y) mean x2 + y 2 = 1. Then p(x, y) is a 2-place predicate on some UD.
Definition 4.2.3. [Quantifiers] We call the symbols and , the quantifiers. Formulae

involving them are called quantified formulae. The statement x p(x) is true if for each x (in

the UD) the property p(x) is T . The statement x p(x) is T if p(x) is T for some x in the UD.

Example 4.2.4. Let UD be the set of all human beings. Consider the 2-place predicate F (x, y):
x runs faster than y. Then
1. x y F (x, y) means each human being runs faster than every human being.
2. x y F (x, y) means for each human being there is a human being who runs slower.
3. x y F (x, y) means there is a human being who runs faster than some human being.
4. x y F (x, y) means there is a human being who runs faster than every human being .

72

CHAPTER 4. LOGIC

Definition 4.2.5.

1. [Scope of quantifier] In the quantified formulae x p(x) or x p(x)

the formula p(x) is called the scope of the quantifier (extent to which that quantification
applies).

2. An x-bound part in a formula is a part of the form x p(x) or x q(x). Any occurrence

of x in an x-bound part of the formula is a bound occurrence of x. Any other occurrence

of x is a free occurrence of x.
Example 4.2.6. In x p(x, y) the occurrence of y is free and both the occurrences of x are

bound. In y x p(x, y) all the occurrences of x and y are bound.

Definition 4.2.7.

1. A quantified formulae is well formed if it is created using the following

rules.
(a) Any atomic formula (of the form P , P (x, y), P (x, b, y)) is a wff.
(b) If A and B are wffs, then A B, A B, A B, A B, and A are wffs.
(c) If A is a wff and x is any variable, then x A and x A are wffs.
2. Let f be a formula. An interpretation (for f ) means the process of specifying the UD,
specifications of the predicates, and assigning values to the free variables from the UD. By
an interpretation of f , we mean the formula f under a given interpretation.
Example 4.2.8. Consider the wff x p(x, y).
1. Take N as UD. Let p(x, y) specify x > y. Let us assign 1 to the free variable y. Then, we
get the interpretation each natural number is greater than 1 which has the truth value
F.
2. Take N as UD. Let p(x, y) mean x + y is an integer, and take y = 2. Then, we get an
interpretation when we add 2 to each natural number we get an integer which has a truth
value T .
Discussion 4.2.9. [Translation] We expect to see that our developments on logic help us
in drawing appropriate conclusions. In order to do that, we must know how to translate an
English statement into a formal logical statement that involves no English words. We may
have to introduce appropriate variables and required predicates. We may have to specify the
UD, but normally we use the most general UD.
Example 4.2.10.

1. Translate: each person in this class room is either a BTech student or

an MSc student.
A: Does the statement guarantee that there is a person in the room?
No. All it says is, if there is a person, then it has certain properties. Let P (x) mean x is
a person in this class room; B(x) mean x is a BTech
student; and M (x) mean x is an



MSc student. Then, the formal expression is x P (x) B(x) M (x) .

2. Translate: there is a student in this class room who speaks Hindi or English.

4.2. PREDICATE LOGIC

73

A: Does the statement guarantee that there is a student in the room?

Yes. Let S(x) mean x is a student in this class room; H(x) mean
 x speaks Hindi; and

E(x) mean x speaks English. Then, the formal expression is x S(x) (H(x) E(x)) .


Note that x S(x) (H(x) E(x)) is not the correct expression. Why?
Remember
x (S(x) T (x)) never asserts S(x) BUT x(S(x) T (x)) asserts both S(x) and T (x).
Practice 4.2.11. Translate into formal logic.
1. Every natural number is either the square of a natural number or its square root is irrational.
2. For every real number x there is a real number y such that x + y = 0.
3. A subset S Rn is called compact, if -write the formal statement here-.
4. A function f : R R is called continuous at a point a, if -write the formal statement
here-.

5. A function f : R R is called continuous, if -write the formal statement here-.

6. A function f : R R is called uniformly continuous, if -write the formal statement
here-.

7. A subset S Rn is called connected, if -write the formal statement here-.

8. A set S is called a group, if -write the formal statement here-.
9. A subset S Rn is called a subspace, if -write the formal statement here-.
10. A function f : S T is called a bijection, if -write the formal statement here-.
11. A function f : Rn Rk is called a linear transformation, if -write the formal statement
here-.

12. A function f : (S, ) (T, +) is called a group isomorphism, if -write the formal
statement here-.

13. A function f : V W is called a vector space isomorphism, if -write the formal

statement here-.

Definition 4.2.12. A quantified formula is called valid if every interpretation of it has truth
value T . Two quantified formulae A and B are called equivalent (A B) if A B is valid.
Example 4.2.13.

1. x P (x) x P (x) is valid.

2. Is x y p(x, y) y x p(x, y)?

A: Yes. Denote x y p(x, y) by L and y x p(x, y) by R. Suppose that L R is F . This
means, we have an interpretation in which L is T and R is F . As R is F , we see that

p(x, y) is F , for each x, y in the UD. In that case, L is F , a contradiction. So, L R is

T . Similarly, R L is T .

74

CHAPTER 4. LOGIC
3. x y p(x, y) y x p(x, y). !!
4. x y p(x, y) 6 y x p(x, y). To see this take p(x, y): x > y.
Did you notice?
Two quantified formulae A and B are equivalent if and only if their interpretations under
the same UD, the same specification of predicates, and the same values to the free variables
have the same truth value.






5. Is x r(x) y r(y) p(x, y) xy r(x) r(y) p(x, y) ?

A: We want to know if






x r(x) y r(y) p(x, y) xy r(x) r(y) p(x, y)

is valid. Let us see whether







x r(x) y r(y) p(x, y) xy r(x) r(y) p(x, y)

is valid. Suppose that this is invalid. So there is an interpretation such that Right hand

side is F and Left hand side is T.

As Right hand side is F, we see that x, say x0 , for




which y r(x) r(y) p(x, y) is F. That is, y the formula r(x0 ) r(y) p(x0 , y)
is F. That is, r(x0 ) is T and for each y we see that r(y) p(x0 , y) is F. That is, r(x0 ) is T

andy(r(y) p(x0 , y)) is F. That

is, the formula r(x0 ) y(r(y) p(x0 , y)) is F. That is,


x r(x) y r(y) p(x, y) is F, a contradiction. The other part is an exercise.





Alternate. Take A := r(x) y r(y) p(x, y) and B := y r(x) r(y) p(x, y) .

Consider an x0 in the UD. If r(x0 ) is F, Then A andB both have value T. If r(x
0 ) is T.




Then notice that r(x0 ) y r(y) p(x0 , y) and y r(x0 ) r(y) p(x0 , y) have the
same truth value.

Thus A B. Hence xA xB.

6. Any student who appears in the exam and gets a score below 30, gets an F grade. Mr x0
is a student who has not written the exam. Therefore, x0 should get an F grade. Do you
agree?
A: Let S(x) mean x is a student, E(x) mean x writes the exam, B(x) mean x gets a
score below 30, and F (x) mean x gets F grade.
n
o
We want to see whether x[S(x) E(x) B(x) F (x)], S(x0 ) E(x0 ) F (x0 )?

Take the following interpretation: S(x) is x is a positive real number, E(x) is x is a

rational number, B(x) is x is an integer, F (x) is x is a natural number, and x0 = 2.

In this interpretation, statements in the premise mean every positive integer is a natural

number and 2 is a positive real number which is not rational. They both are true.

Whereas the conclusion means 2 is a natural number which is false. So, the argument
is incorrect.

4.2. PREDICATE LOGIC

75

7. Translate the following into formal statements.

All scientists are human beings. Therefore, all children of scientists are children of human
beings.
A: Let Sx : x is a scientist; Hx : x is a human being and Cxy : x is a child of y.
Let the hypothesis be x(Sx Hx). Then, the possible translations of the conclusion are

the following.

(a) x(y(Sy Cxy) z(Hz Cxz)). It means for each x, if x has a scientist father,
then x has a human father.

(b) x[y(Sy Cxy) z(Hz Cxz)]. This is wrong, as the statement means for all x,

if x is a common child of all scientists, then x is a common child of all human beings.

(c) x(Sx y(Cyx z(Hz Cyz))). This means for each x, if x is a scientist, then
each child of x has a human father.

(d) xy(Sx Cyx) xy(Hx Cxy). What? This means if each x is a scientist and
each y is a child of x (including x it self!), then each x is a human being and each y

is a child of x.
Exercise 4.2.14.

1. Write a formal definition of lim f (x) 6= l.

xa

2. Is x [p(x) q(x)] x p(x) x q(x) valid? Is its converse valid?

3. [common ones] If r does not involve x, then establish the following assertions.
(a) x p(x) x p(x);
x p(x) x p(x)




(b) x p(x) q(x) x p(x) x q(x);
x p(x) q(x) x p(x) x q(x).




(c) x p(x) q(x) x p(x) x q(x);
x p(x) q(x) x p(x) x q(x).




(d) x r q(x) r x q(x);
x r q(x) r x q(x)




(e) x r q(x) r x q(x);
x r q(x) r x q(x).




(f ) x p(x) r x p(x) r ;
x p(x) r x p(x) r .

4. Translate and check for validity of the following arguments.

(a) Recall that the decimal representation of a rational number either terminates or begins
to repeat the same finite sequence of digits, whereas that of an irrational number
neither terminates nor repeats. The square root of a natural number either has a
decimal representation which is terminating or has a decimal representation which
is non-terminating and non-repeating. The square root of all natural numbers which
are squares have terminating decimal representation. Therefore, the square root of a
natural number which is not a square is an irrational number.
(b) For any two algebraic numbers a and b, a 6= 0, 1 and b irrational, we have that ab

is transcendental. The number i (imaginary unit) is irrational and algebraic. The

number i is not equal to 0 or 1. Therefore, the number ii is transcendental.




5. (a) Give an interpretation to show that x r(x) y r(y) p(x, y) is not valid.

76

CHAPTER 4. LOGIC


(b) Give an interpretation to show the incorrectness of x p(x) q(x) x p(x)


q(x) .

6. Write a formal statement taking UD:= all students in all IITs in India, for the following.
For each student in IITG there is a student in IITG with more CPI.

7. Let UD= R, p(x): x is an integer, and q(x): x is a rational number. Translate the
following statements into English.


(a) x p(x) q(x)


(b) x p(x) q(x)




(c) x p(x) (x > 2) x q(x) (x < 2)


(d) > 0 > 0(0 < |x a| < |f (x) l| < )

8. Take the most general UD. Check whether the following conclusion is valid or not.
Each student writes the exam using blue ink or black ink. A student who writes the exam
using black ink and does not write his/her roll number gets an F grade. A student who
writes the exam using blue ink and does not have his/her ID card gets an F grade. A
student who has his/her ID card has written the exam with black ink. Therefore, a student
who passes the exam must have written his roll number.
Ans: Let S(x) : x is a student, B(x) : x writes the exam using blue ink, Bl(x) : x writes
the exam using black ink, R(x) : x writes his/her roll number, I(x) : x has his/her ID
card, F (x) : x gets an F grade. We have to determine, whether the following conclusion
is valid. (Continue.)

Chapter 5

Lattices and Boolean Algebra

5.1

Lattices

Discussion 5.1.1. In a poset, is it necessary that two elements x, y should have a common
upper bound?
Ans: No. Take [6] with divides partial order. The elements 5 and 3 have no common upper
bound.
In a poset, if a pair {x, y} has at least one upper bound, is it necessary that {x, y} should have

a lub?

Ans: No. Consider the third poset described by its Hasse diagram in Figure 5.1. Then, the
pair {a, b} has c, d as upper bounds, but there is no lub of {a, b}.
1

c
b

a
b

b
b

a
b

0
a distributive lattice

0
a non-distributive lattice

a
b

not a lattice

Figure 5.1: Hasse diagrams

Definition 5.1.2. [Lattice]

1. A poset (L, ) is called a lattice if each pair x, y L has a lub denoted x y and a glb
denoted x y.

2. A lattice is called a distributive lattice if it satisfies the following two properties.

a (b c) = (a b) (a c).

a (b c) = (a b) (a c).

Example 5.1.3.

distributive laws

1. Let L = {0, 1} Z and define a b = max{a, b} and a b = min{a, b}.

Then, L is a chain as well as a distributive lattice.

77

78

CHAPTER 5. LATTICES AND BOOLEAN ALGEBRA

2. The set N with usual order and := max and := min is a distributive lattice. We

consider two cases to verify that a (b c) = (a b) (a c). The second distributive

identity is left as an exercise for the reader.

(a) Case 1: a min{b, c}. Then, either a b or a c, say a b. Hence,

a (b c) = max{a, min{b, c}}
= a = min{max{a, b} = a, max{a, c} a} = (a b) (a c).
(b) Case 2: a < min{b, c}. Then, a < b and a < c. Hence,
a (b c) = max{a, min{b, c}}
= min{b, c} = min{max{a, b} = b, max{a, c} = c} = (a b) (a c).
3. Prove that the first figure in Figure 5.1 is a distributive lattice.
4. Prove that the second figure in Figure 5.1 is a lattice but not a distributive lattice.
5. Let S = {a, b, c}. On P(S), we define A B = A B and A B = A B. Then, it can
be easily verified that P(S) is a lattice.

6. Fix a positive integer n and let D(n) denote the poset obtained using the divides partial
order with := lcm and := gcd. Then, prove that D(n) is a distributive lattice. For
example, for n = 12, 30 and 36, the corresponding lattices are shown below.
36
b

30

12
b

12

4
b

6
b

10

b
b

15

6
b

18

Exercise 5.1.4.
D(pn ).

1. Fix a prime p and a positive integer n. Draw the Hasse diagram of

Does this correspond to a chain? Give reasons for your answer.

2. Let n be a positive integer. Then, prove that D(n) is a chain if and only if n = pm , for
some prime p and a positive integer m.
3. Let (X, f ) be a nonempty chain with := lub and := glb. Is it a distributive lattice?
Proposition 5.1.5. [properties of a lattice]
statements are true.

Let (L, ) be a lattice. Then, the following

(a) The operations and are idempotent, i.e., lub{a, a} = a and glb{a, a} = a.
(b) commutative (so is ).
(c) is associative (so is ).

5.1. LATTICES

79

(d) a (a b) = a = a (a b) [absorption] , i.e., glb{a, lub{a, b}} = a = lub{a, glb{a, b}} .

(e) a b a b = b a b = a.
(f ) b c {a b a c, a b a c} [isotonicity] .
(f1) {a b, c d} {a c b d, a c b d}.
(g) a (b c) (a b) (a c), a (b c) (a b) (a c) [distributive inequalities] .
(h) a c a (b c) (a b) c [modular inequality] .
Proof. We prove only a few parts. The rest are left for the reader.
(c) Let d = a (b c). Then, d is the lub of {a, b c}. Thus, d is an upper bound of both
{a, b} and {a, c}. So, d a b and d a c. Therefore, d a b and d c and hence
d an upper bound of {a b, c}. So, d is greater or equals to the lub of {a b, c}, i.e.,

d (a b) c. Thus, the first part of the result follows.

(e) Let a b. As b is an upper bound of {a, b}, we have a b = lub{a, b} b. Also, a b is an

upper bound of {a, b} and hence a b b. So, we get a b = b. Conversely, let a b = b.

As a b is an upper bound of {a, b}, we have a a b = b. Thus, the first part of the
result follows.

(f) Let b c. Note that a c a and a c c b. So, a c is an upper bound for {a, b}.
Thus, a c lub{a, b} = a b and hence the prove of the first part is over.

(f1) Using isotonicity, we have a c b c b d. Similarly, using isotonicity again, we have

a c b c b d.

(g) Note that a a b and a a c. Thus, a = a a (a b) (a c). As b a b and

c a c, we get b c (a b) (a c). Now using (f1), we obtain the required result,
i.e., a (b c) (a b) (a c).

(h) Let a c. Then, a c = c and hence by the distributive inequality, we have a (b

c) (a b) (a c) = (a b) c. Conversely, let a (b c) (a b) c. Then,

a a (b c) (a b) c c and the required result follows.

Practice 5.1.6. Show that in a lattice one distributive equality implies the other.
Definition 5.1.7. If (Li , i ), i = 1, 2 are lattices with := lub and := glb. Then, (L1 L2 , )

is a poset with a = (a1 , a2 ) (b1 , b2 ) = b if a1 1 b1 and a2 2 b2 , that is, if b dominates a

entrywise. In this case, we see that a b = (a1 1 b1 , a2 2 b2 ) and a b = (a1 1 b1 , a2 2 b2 ).
Thus (L1 L2 , ) is a lattice, called the direct product of (Li , i ), for i = 1, 2.

Example 5.1.8.

1. Consider L = {0, 1} with usual order. The set of all binary strings Ln

of length n is a poset with the order (a1 , . . . , an ) (b1 , . . . , bn ) if ai bi , i. This is the

n-fold direct product of L. It is called the lattice of n-tuples of 0 and 1.

2. Consider the lattices [3] and [4] with usual orders. Hasse diagram of the direct product
[3] [4] is given below.

80

CHAPTER 5. LATTICES AND BOOLEAN ALGEBRA

(3, 4)
b

(1, 4)
b

(3, 1)

(1, 1)
Practice 5.1.9. Consider N with the usual order. The lattice order defined on N2 as a direct
product is different from the lexicographic order on N2 . Draw pictures for all (a, b) (5, 6) in

both the orders to see the argument.

Proposition 5.1.10. The direct product of two distributive lattices is a distributive lattice.
Proof. The direct product of two lattices is a lattice by definition. Note that
[(a1 , b1 ) (a2 , b2 )] (a3 , b3 ) = (a1 a2 , b1 b2 ) (a3 , b3 )


=
(a1 a2 ) a3 , (b1 b2 ) b3


=
(a1 a3 ) (a2 a3 ), (b1 b3 ) (b2 b3 )

 

=
(a1 a3 ), (b1 b3 ) (a2 a3 ), (b2 b3 )

 

=
(a1 , b1 ) (a3 , b3 ) (a2 , b2 ) (a3 , b3 )

Definition 5.1.11. Let (Li , i ), i = 1, 2 be two lattices. A function f : L1 L2 satisfying

f (a 1 b) = f (a) 2 f (b) and f (a 1 b) = f (a) 2 f (b) is called a lattice homomorphism.

Furthermore, if f is a bijection, then it is called a lattice isomorphism.

Example 5.1.12.

1. Let D be the set of all words in our English dictionary with dictionary

ordering. Then, prove that D is a lattice. Now, consider the set S of all words in D
which are of length at most six or first-part-words of length six. Note that S is a lattice
again. Define f : D S as f (d) = d if d has length at most six, otherwise f (d) is the

first-part-word of length 6 of d. Then, f is a homomorphism. It is not an isomorphism as

f (stupid) = f (stupidity).
2. Consider the lattice N with usual order. Let S = {0, 1, 2} with usual order. Let f : N S
be a homomorphism. If f (m) = 0 and f (n) = 1, then m n, or else, we have
0 = f (m) = f (m n) 6= f (m) f (n) = 0 1 = 1.

Thus, the map f must have one of the following forms. Draw pictures to understand this.
(a) f 1 (0) = N.
(b) f 1 (0) = [k] and f 1 (1) = {k + 1, . . .}.

(c) f 1 (0) = [k], f 1 (1) = [r] \ [k] and f 1 (2) = N \ [r + k].

5.1. LATTICES

81

Definition 5.1.13. A lattice (L, ) is complete if A (lub of A) and A (glb of A) exist in

L, for each nonempty subset A of L.

Example 5.1.14.

1. Verify that every finite lattice is complete.

2. Every complete lattice has a least element 0 and a greatest element 1. Any lattice with
these two elements is called a bounded lattice.
3. The set [0, 5] with usual order is a bounded and complete lattice. So, is the set [0, 1)[2, 3].
4. The set (0, 5] is a lattice which is neither bounded nor complete.
5. The set [0, 1) (2, 3] is a bounded lattice, though not complete.
6. The set R with usual order is a lattice. It is not complete in the lattice sense. It is
conditionally complete, that is, for every bounded nonempty subset glb and lub exist.
Can you think of a reason which implies the importance of the condition non-emptiness ?

(1, 1, 0)
b

(1, 0, 0)

(1, 1, 1)

30

(1, 0, 1)

{a, b, c}
b

(0, 1, 1)

6
b

10

b
b

15

{a, b}

(0, 0, 1)

{a}

(0, 1, 0)

{a, c}
b

{b}

{b, c}

b
b

(0, 0, 0)

{c}

7. Fix n N and let p1 , p2 , . . . , pn be n distinct primes. Prove that the lattice D(N ), for
N = p1 p2 pn is isomorphic to the lattice Ln (the lattice of n-tuples of 0 and 1) and to
the lattice P(S), where S = {1, 2, . . . , n}. The Hasse diagram for n = 3 is shown above.

Definition 5.1.15. [Complement] Let (L, ) be a bounded lattice. Then, a complement of

b L is an element (if it exists) c L such that b c = 1 and b c = 0. The lattice is called

complemented if every element has at least one complement. We shall use b to denote b, a

complement of b.

Example 5.1.16.

1. The interval [0, 1] with usual ordering is a distributive lattice but not

complemented.
2. Verify the captions of the two figures given below. Also, compute 0, a, b, c, and 1.
1
b

a
b

b
b

Complemented but NOT distributive

Distributive but NOT complemented

82

CHAPTER 5. LATTICES AND BOOLEAN ALGEBRA

Discussion 5.1.17. [The comparison table] Let (L, ) be a lattice and let a, b, c L. Then,

the following table lists the properties that hold (make sense) in the specified type of lattices.
Properties

Lattice type

, are idempotent

any lattice

, are associative

any lattice

abab=aab=b

any lattice

, are commutative

any lattice

[absorption] a (a b) = a = a (a b)

any lattice

[isotonicity] b c {a b a c, a b a c}

any lattice

[distributive inequalities]

any lattice

a (b c) (a b) (a c)

a (b c) (a b) (a c)

[modular inequality] a c a (b c) (a b) c

any lattice

0 is unique; 1 is unique

bounded lattice !!

if a is a complement of b, then b is also a complement of a

bounded lattice !!

0 is unique and it is 1; 1 is unique and it is 0

bounded lattice !!

an element a has a unique complement

n
o
a c = b c, a c = b c a = b
o
[cancellation] n
a c = b c, a c = b c a = b
[DeMorgan]

(a b) = a b

(a b) = a b

a b = 1 a b = a

a b = 0 a b = a

distributive complemented lattice !!

distributive complemented lattice
distributive complemented lattice
distributive complemented lattice

Proof. We will only prove the properties that appear in the last three rows. The other properties
are left as an exercise for the reader. To prove the cancellation property, note that
b = b 0 = b (c c) = (b c) (b c) = (a c) (a c) = a (c =
6 c) = a 0 = a
and
b = b 1 = b (c c) = (b c) (b c) = (a c) (a c) = a (c c) = a 1 = a.
To prove the DeMorgans property, note that
(a b) (a b) = (a b a) (a b b) = 1 1 = 1,
and
(a b) (a b) = (a a b) (b a b) = 0 0 = 0.
Hence, by Definition 5.1.15, we get (a b) = a b. Similarly, note that (a b) (a b) =
(aab)(bab) = 11 = 1 and (ab)(ab) = (aba)(abb) = 00 = 0.

5.2. BOOLEAN ALGEBRAS

83

Thus, by Definition 5.1.15, we again get (a b) = (a b). To prove the next assertion, note

that if a b = 1, then

a = a (b b) = (a b) (a b) = (a b) 1 = a b.
Conversely, if a = a b, then a b = (a b) b = 1. On similar lines, one completes the

proof of the second part and is left as an exercise for the reader.
Exercise 5.1.18.

1. Prove that every linearly ordered set is distributive.

2. Draw the Hasse diagrams of [3] [4] with dictionary order and the lattice order ((m, n)
(p, q) if m p and n q).

3. Give a partial order on N to make it a bounded lattice. You may draw Hasse diagram
representing it.
4. Does there exist a partial order on N for which each nonempty subset has finitely many (at
least one) upper bounds and finitely many (at least one) lower bounds?
5. Consider the lattice N2 with lexicographic order. Is it isomorphic to the direct product of
(N, ) with itself, where is the usual order?
6. Show that {0, 1, 2, . . .} is a complete lattice under divisibility relation (allow (0, 0) in the
relation). Characterize those sets A for which A = 0.

7. Draw as many Hasse diagrams of non-isomorphic lattices of size 6 as you can.

8. Is the lattice [2] [2] [2] [2] isomorphic to [4] [4]?
9. Prove/Disprove: If L is a lattice which is not complete, then L N .
10. Draw the Hasse diagram of a finite complemented lattice which is not distributive.
11. How many lattice homomorphisms are there from [2] to [9]?

5.2

Boolean Algebras

Definition 5.2.1. [Boolean algebra] A Boolean algebra is a set S which is closed under the
binary operations (called the join) and (called the meet) and for each x, y, z S, satisfies
the following properties.

1. x y = y x, x y = y x [commutative] .
2. x (y z) = (x y) (x z), x (y z) = (x y) (x z) [distributive] .
3. 0, 1 S such that x 0 = x, x 1 = x [identity elements] .
4. For each x S, y S such that x y = 1 and x y = 0 [inverse] .
Proposition 5.2.2. Let S be a Boolean algebra. Then, the following statements are true.
1. Elements 0 and 1 are unique.
2. For each s S, s is unique. Therefore, for each x S, x is called the inverse of x.
3. If y is the inverse of x, then x is the inverse of y. That is, x = (x).

84

CHAPTER 5. LATTICES AND BOOLEAN ALGEBRA

Proof.
1. Let 01 and 02 be two such elements. Then, 01 x = x and x = x 02 , for all x S.
Hence, 01 = 01 02 = 02 . Thus, the required result follows. A similar argument implies
that 1 is unique.

2. Suppose there exists t, r S such that s t = 1, s t = 0, s r = 1 and s r = 0. Then,

t = t 1 = t (s r) = (t s) (t r) = 0 (t r) = (s r) (t r) = (s t) r = 1 r = r.
3. It directly follows from the definition of inverse.
Example 5.2.3.
A =

Ac ,

1. Let S 6= . Then, P (S) is a Boolean algebra with = , = ,

0 = and 1 = S. So, we have Boolean algebras of finite size as well as of

uncountable size.

2. Take S = {n N : n|30} with a b = lcm(a, b), a b = gcd(a, b), a =

30
a,

0 = 1 and

1 = 30. It is a Boolean algebra.

3. Let B = {T, F } with 0 = F , 1 = T and with usual , , . It is a Boolean algebra.

4. Let B be the set of all truth functions involving the variables p1 , . . . , pn , with usual , , .
Take 0 = F and 1 = T. This is the free Boolean algebra on the generators p1 , . . . , pn .

5. The class of finite length formulae involving variables p1 , p2 , . . . is a countable infinite

Boolean algebra with usual operations.
Observation.
The rules of Boolean algebra treat (, 0) and (, 1) equally. Notice that the second part
of the rules in Definition 5.2.1 can be obtained by replacing with and 0 with 1. Thus,

any statement that one can derive from these rules has a dual version which is derivable
from the rules. This is called the principle of duality.

Theorem 5.2.4. [Rules] Let (S, , , ) be a Boolean algebra. Then, the following rules, as

well as their dual, hold true.

1. 0 = 1.

2. For each s S, s s = s [idempotence] .

3. For each s S, s 1 = 1.
4. For each s, t S, s (s t) = s [absorption] .
5. If s t = r t and s t = r t, then s = r [cancellation] .
6. (s t) r = s (t r) [associative] .
Proof. We give the proof of the first part of each item and that of its dual is left for the reader.
1. 1 = 0 (0) = 0.
2. s = s 0 = s (s s) = (s s) (s s) = (s s) 1 = (s s).

5.2. BOOLEAN ALGEBRAS

85

3. 1 = s s = s (s 1) = (s s) (s 1) = 1 (s 1) = s 1.
4. s (s t) = (s 1) (s t) = s (1 t) = s 1 = s.
5. s = s 0 = s (t t) = (s t) (s t) = (r t) (r t) = r (t t) = r 0 = r.
6. We will prove it using absorption and cancellation. Using absorption, (s t) s = s and




s (r s) = s. Thus, (s t) r s = (s t) s (r s) = s (r s) = s. Using


absorption, we also have s (t r) s = s and hence




s (t r) s = (s t) r s.

Now, we see that [s (t r)] s = 0 [(t r) s] = (t s) (r s) and on similar

lines, [(s t) r] s = (t s) (r s). Thus, we again have




s (t r) s = (s t) r s.

Hence, applying the cancellation property, the required result follows.

Example 5.2.5. Let (L, ) be a distributive complemented lattice. Then, by Definition 5.1.2,

L has two binary operations and and by Definition 5.1.15, the operation x. It can be
easily verified that (L, , , ) is a indeed a Boolean algebra.

Now, let (B, , , ) be a Boolean algebra. Then, for any two elements a, b B, we define

a b if a b = a. The next result shows that is a partial order in B. This partial order is
generally called the induced partial order. Thus, we see that the Boolean algebra B, with
the induced partial order, is a distributive complemented lattice.

Theorem 5.2.6. Let (B, , , ) be a Boolean algebra. Define, a b if a b = a. Then, is

a partial order on B. Furthermore, a b = lub{a, b} and a b = glb{a, b}.

Proof. We first verify that (B, ) is indeed a partial order.
Reflexive: By idempotence, s s and hence is reflexive.

Antisymmetry: Let s t and t s. Then, we have s = s t = t.

Transitive: Let s t and t r. Then, using associativity, sr = (st)r = s(tr) = st = s

and thus, s r.

Now, we show that a b = lub{a, b}. Since B is a Boolean algebra, using absorption, we get

(a b) a = a and hence a a b. Similarly, b a b. So, a b is an upper bound for {a, b}.

Now, let x be any upper bound for {a, b}. Then, by distributive property, (a b) x =

(a x) (b x) = a b. So, a b x. Thus, a b is the lub of {a, b}. The rest of the proof is
similar and hence is left for the reader.

Thus, we observe that there is one-to-one correspondence between the set of Boolean Algebras
and the set of distributive complemented lattice.
Definition 5.2.7. [Atom] Let B be a Boolean algebra. If there exists a b B, b 6= 0 such that
b is a minimal element in B, then b is called an atom.
Example 5.2.8.

1. In the powerset Boolean algebra, singleton sets are the only atoms.

86

CHAPTER 5. LATTICES AND BOOLEAN ALGEBRA

2. Atoms of the divides 30 Boolean algebra are 2, 3 and 5.
3. The {F, T } Boolean algebra has only one atom, namely T .

Exercise 5.2.9.

1. Determine the atoms of the free Boolean algebra with generators p1 , . . . , pn ?

2. Is it necessary that every Boolean algebra has at least one atom?

Definition 5.2.10. [Boolean homomorphism] Let B1 and B2 be two Boolean algebras. A
function f : B1 B2 is a Boolean homomorphism if it preserves 0, 1, , , and . That is,
f (01 ) = 02 , f (11 ) = 12 , f (a b) = f (a) f (b), f (a b) = f (a) f (b), and f (a) = f (a).
A Boolean isomorphism is a Boolean homomorphism which is a bijection.
Exercise 5.2.11. Let B1 and B2 be two Boolean algebras and let f : B1 B2 be a function

that satisfies the four conditions f (01 ) = 02 , f (11 ) = 12 , f (a b) = f (a) f (b) and f (a b) =
f (a) f (b). Then, prove that f also satisfies the fifth condition, namely f (a) = f (a).

Example 5.2.12. The function f : P (J4 ) P (J3 ) defined as f (S) = S \ {4} is a Boolean

homomorphism. We check just two properties and the rest is left as an exercise.

f (A B) = f (A B) = (A B) \ {4} = (A \ {4}) (B \ {4}) = f (A) f (B).

f (11 ) = f (J4 ) = J4 \ {4} = J3 = 12 .
Proposition 5.2.13. Let B be a Boolean algebra and p, q be two distinct atoms. Then, pq = 0.
Proof. Suppose that p q 6= 0. As p q p and p is an atom, we must have p q = p, i.e.,
q p. As p 6= q and q is an atom, it follows that p cannot be an atom.

Proposition 5.2.14. Let B be a Boolean algebra with three distinct atoms p, q and r. Then,
p q 6= p q r.
Proof. Let if possible p q = p q r. Then, we have
r = r 0 = r [(p q) (p q)] = [r p q] [r (p q)] = [p q] [r (p q)]
= [(p q) r] [(p q) (p q)] = (p q) r = (p r) (q r) = 0 0 = 0,
a contradiction to r being an atom, i.e., r is nonzero.
Example 5.2.15. Let B be a Boolean algebra having distinct atoms A = {p, q, r}. Then, B
has at least 23 elements.

To show this, we define f : P(A) B by f () = 0 and for S A, f (S) =

that f is a one-one function.

x and claim

xS

Suppose f (S) = f (T ). Then, f (S) = f (S) f (T ) = f (S T ). In view of Proposition 5.2.14,

we have S = S T , i.e., T S. Similarly, as f (T ) = f (T S), we have S T and hence S = T .

Thus, f is a one-one function. Therefore, f (S) is distinct, for each subset of A and thus B has
at least 23 elements.

5.2. BOOLEAN ALGEBRAS

87

Theorem 5.2.16. Let B be a Boolean algebra having distinct atoms A = {p, q, r, s}. Let b B,

b 6= 0. Suppose that S = {atoms x : x b} = {p, q, r}. Then, b = p q r.

Proof. It is clear that p q r b. Suppose that p q r < b. Then,

b = b [(p q r) (p q r)] = [b (p q r)] [b (p q r)] = (p q r) [b (p q r)].

Therefore, the above equality implies that [b (p q r)] 6= 0. So, there is an atom, say x,
such that x b (p q r). Thus, we have x b and x (p q r).

Notice that if x (p q r), then x 0, which is not possible. So, x 6= p, q, r is an atom in

S, a contradiction.

Theorem 5.2.17. [Representation] Let B be a finite Boolean algebra. Then, there exists a
set X such that B is isomorphic to P(X).
Proof. Put X = {atoms of B}. Note that X 6= . Define f : B P(X) by f (b) = {atoms b}.
We show that f is the required Boolean isomorphism.

Injection: Let b1 6= b2 . Then, either b1  b2 or b2  b1 . Without loss of generality, let b1  b2 .

[Now imagine the power set Boolean algebra. Saying b1  b2 is the same as b1 * b2 . In that case,
we have an element in b1 which is not in b2 . That is, b1 bc2 6= . That is, there is a singleton

subset of b1 bc2 . This is exactly what we are aiming for, i.e., to prove that b1 b2 6= 0.] Note

that b1 = b1 (b2 b2 ) = (b1 b2 ) (b1 b2 ). Also, the assumption b1  b2 implies b1 b2 6= b1

and hence b1 b2 6= 0. So, there exists an atom x (b1 b2 ) and hence x = x b1 b2 .

Therefore,

x b1 = (x b1 b2 ) b1 = x b1 b2 = x.
Thus, x b1 . Similarly, x b2 . As x 6= 0, we cannot have x b2 (the condition x b2 and

x b2 implies x b2 b2 = 0). Thus, f (b1 ) 6= f (b2 ).

Surjection: Let A = {x1 , . . . , xk } X and put b = x1 xk (if k = 0, then b = 0). Clearly,

A f (b). Need to show: A = f (b). So, let y f (b), i.e., y is an atom in B and
y = y b = y (x1 xk ) = (y x1 ) (y xk ).

Since y 6= 0, by Proposition 5.2.13, it follows that y xi0 6= 0, for some i0 {1, 2, . . . , k}. As

xi0 and y are atoms, we have y = y xi = xi and hence y A. Thus, f is a surjection.

Preserving 0, 1: Clearly f (0) = and f (1) = X.
Preserving , : By definition,

x f (b1 b2 ) x b1 b2 x b1 and x b2
x f (b1 ) and x f (b2 ) x f (b1 ) f (b2 ).
Now, let x f (b1 b2 ). Then, by definition, x = x (b1 b2 ) = (x b1 ) (x b2 ). So, there exists

i such that x bi 6= 0 (say, x b1 ). As, x is an atom, x b1 and hence x f (b1 ) f (b1 ) f (b2 ).

Conversely, let x f (b1 ) f (b2 ). Without loss of generality, let x f (b1 ). Thus, x b1 and
hence x b1 b2 which in turn implies that x f (b1 b2 ).
As a direct corollary, we have the following result.

88

CHAPTER 5. LATTICES AND BOOLEAN ALGEBRA

Corollary 5.2.18. Let B be a finite Boolean algebra having exactly k atoms. Then, B is
isomorphic to P({1, 2, . . . , k}) and hence has exactly 2k elements.
Exercise 5.2.19.

1. Determine the number of elements in a finite Boolean algebra.

2. Supply a Boolean homomorphism f from P (J4 ) to P (J3 ) such that the image of P (J4 ) has
4 elements.
3. Prove/Disprove: The number of Boolean homomorphisms from P (J4 ) to P (J3 ) is less than
the number of lattice homomorphisms from P (J4 ) to P (J3 ).
4. Show that a lattice homomorphism on a Boolean algebra which preserves 0 and 1 is a
Boolean homomorphism.
5. Consider the class of all functions f : R {, e}. Can we define some operations on this
class to make it a Boolean algebra?

6. Show that a finite Boolean algebra must have at least one atom. Is finite necessary?
7. A positive integer is called squarefree if it is not divisible by the square of a prime. Let
Bn = {k N : k|n}. For a, b Bn take the operations a b = lcm(a, b), a b = gcd(a, b)

and a = n/a. Show that Bn is a Boolean algebra if and only if n > 1 is squarefree.

8. Show that the set of subsets of N which are either finite or have a finite complement is a
countable infinite Boolean algebra. Find the atoms. Is it isomorphic to the Boolean algebra
of all finite length formulae involving variables p1 , p2 , ?
9. Let B be a Boolean algebra and xi B, i = 1, 2, . . .. We know that, for each n N, the
n

W
W
expression xi is meaningful in each Boolean algebra due to associativity. Is xi
i=1

necessarily a meaningful expression?

i=1

10. Prove/Disprove: Let f : B1 B2 be a Boolean homomorphism and a B1 be an atom.

Then, f (a) is an atom of B2 .

11. Fill in the blank: The number of Boolean homomorphisms from P (J4 ) to P (J3 ) is

12. Fill in the blank: The number of Boolean homomorphisms from P (J4 ) onto P (J3 ) is

13. How many atoms does divides 30030 Boolean algebra has? How many elements does it
have?
14. If B1 and B2 are Boolean algebras of size k (k > 100), then they must be isomorphic and
there must be more than k isomorphisms between them.
15. Give examples of two countably infinite non-isomorphic Boolean algebras.
16. Give examples of two uncountably infinite non-isomorphic Boolean algebras.

Chapter 6

Counting
Discussion 6.0.1. In the previous chapters, we had learnt that two sets, say A and B, have
the same cardinality if there exists a one-one and onto function f : A B. We also learnt the

following two rules of counting which play a basic role in the development of this subject.

1. [Multiplication rule] If a task has n compulsory parts, say A1 , A2 , . . . , An and the ith
part can be completed in mi = |Ai | ways, i = 1, . . . , n, then the task can be completed in

m1 m2 mn ways. In mathematical terms,

|A1 A2 An | = |A1 | |A2 | |An |.

2. [Addition rule] If a task consists of n alternative parts, say A1 , A2 , . . . , An , and the ith
part can be done in |Ai | = mi ways, i = 1, . . . , n, then the task can be completed in
m1 + m2 + + mn ways. In mathematical terms,

|A1 A2 An | = |A1 | + |A2 | + + |An |, whenever Ai Aj 6= , 1 i < j n.

Definition 6.0.2. We use the notation n! = 1 2 n. By convention, we take 0! = 1.

6.1

Permutations and combinations

Example 6.1.1. How many three digit natural numbers can be formed using digits 0, 1, , 9?
Identify the number of parts in the task and the type of the parts (compulsory or alternative).
Which rule applies here?
Ans: The task has three compulsory parts. Part 1: choose a digit for the leftmost place.
Part 2: choose a digit for the middle place. Part 3: choose a digit for the rightmost place.

Multiplication rule applies. Ans: 900.

Example 6.1.2. How many three digit natural numbers with distinct digits can be formed
using digits 1, , 9 such that each digit is odd or each digit is even? Identify the number of

parts in the task and the type of the parts (compulsory or alternative). Which rule applies here?
89

90

CHAPTER 6. COUNTING

Ans: The task has two alternative parts. Part 1: form a three digit number with distinct
numbers from {1, 3, 5, 7, 9} using the odd digits. Part 2: form a three digit number with distinct

numbers from {2, 4, 6, 8} using the even digits. Observe that Part 1 is a task having three

compulsory subparts. In view of 6.1.1, we see that Part 1 can be done in 60 ways. Part 2 is a
task having three compulsory subparts. In view of 6.1.1, we see that Part 2 can be done in 24
ways. Since our task has alternative parts, addition rule applies. Ans: 84.
Definition 6.1.3. [r-sequence] An r-sequence of elements of X is a sequence of length r with

elements from X. This may be viewed as a word of length r with alphabets from X or as a
function f : [r] X. We write an r-sequence of S to mean an r-sequence of elements of S.
Theorem 6.1.4. [Number of r-sequences] The number of r-sequences of [n] is nr .
Proof. Here the task has r compulsory parts. Choose the first element of the sequence, the
second element and so on.
Exercise 6.1.5.

1. In how many ways can r distinguishable/distinct balls be put into n

distinguishable/distinct boxes?
2. How many distinct ways are there to make a 5 letter word using the ENGLISH alphabet
(a) with no restriction?
(b) with ONLY consonants?
(c) with ONLY vowels?
(d) with a consonant as the first letter and a vowel as the second letter?
(e) if the vowels appear only at odd positions?
3. Determine the total number of possible outcomes if
(a) two coins are tossed?
(b) a coin and a die are tossed?
(c) two dice are tossed?
(d) three dice are tossed?
(e) k dice are tossed, where k N?
(f ) five coins are tossed?
4. How many 5-letter words using only As, Bs, Cs, and Ds are there that do not contain
the word CAD?
Definition 6.1.6. [r-permutation, n-set] By an n-set, we mean a set containing n elements.
An r-permutation of an n-set S is an arrangement of r distinct elements of S in a row. An
r-permutation may be viewed as a one-one mapping f : [r] S. An n-permutation of an n-set
is simply called a permutation.

6.1. PERMUTATIONS AND COMBINATIONS

91

Example 6.1.7. How many one-one maps f : [4] A = {A, B, . . . , Z} are there?

Ans: The task has 4 compulsory parts: select f (1), select f (2), select f (3) and select f (4).

Note that f (2) cannot be f (1), f (3) cannot be f (1) or f (2) and so on. Now apply the multiplication rule. Ans: 26 25 24 23 =

26!
22! .

Theorem 6.1.8. [Number of r-permutations] The number of r-permutation of an n-set S is

P (n, r) =

n!
(nr)! .

Proof. Let us view an r-permutation as a one-one map from f : [r] S. Here the task

has r compulsory tasks: select f (1), select f (2), . . ., select f (r) with the condition, for 2

k r, f (k) 6 {f (1), f (2), . . . , f (k 1)}. Multiplication rule applies. Hence, the number of

r-permutations equals n(n 1) (n r + 1) =

n!
(nr)! .

Definition 6.1.9. By P (n, r), we denote the number of r-permutations of [n]. By convention,
P (n, 0) = 1. Some books use the notation n(r) and call it the falling factorial of n. Thus, if
r > n then P (n, r) = n(r) = 0 and if n = r then P (n, r) = n(r) = n!.
Exercise 6.1.10.

1. How many distinct ways are there to make 5 letter words using the

ENGLISH alphabet if the letters must be different?

2. How many distinct ways are there to arrange the 5 letters of the word ROY AL?
3. Determine the number of ways to place 4 couples in a row if each couple seats together.
4. How many distinct ways can 8 persons, including Ram and Shyam, sit in a row, with Ram
and Shyam sitting next to each other?
Proposition 6.1.11. [principle of disjoint pre-images of equal size] Let A, B be finite sets
and f : A B be a function such that for each pair b1 , b2 B we have |f 1 (b1 )| = k = |f 1 (b2 )|
(recall that f 1 (b1 ) f 1 (b2 ) = ). Then, |A| = k|B|.

Discussion 6.1.12. Consider the word AABAB. Give subscripts to the three As and the two
Bs and complete the following list. Notice that each of them will give us AABAB if we erase
the subscripts.
A1 A2 B1 A3 B2
A1 A2 B2 A3 B1

A1 A2 B1 B2 A3

A1 A2 A3 B1 B2

A1 A2 A3 B2 B1

A1 A2 B2 B1 A3

B2 A3 B1 A1 A2

B2 A3 B1 A2 A1

Example 6.1.13. How many words of size 5 are there which use three As and two Bs?
Ans: Put A = {arrangements of A1 , A2 , A3 , B1 , B2 } and B = {words of size 5 which use three

As and two Bs}. For each arrangement a A, define f (a) to be the word in B obtained by
erasing the subscripts. Then, the function f : A B satisfies:

for each b, c B, b 6= c, we have |f 1 (b)| = |f 1 (c)| = 3!2! and f 1 (b) f 1 (c) = .

Thus, by Proposition 6.1.11, |B| =

|A|
3!2!

5!
3!2! .

92

CHAPTER 6. COUNTING

Remark 6.1.14. Let us fix n, k N with 0 k n and ask the question how many words of

size n are there which uses k many As and (n k) many Bs ?

Ans: Put A = {arrangements of A1 A2 . . . Ak B1 B2 . . . Bnk } and B = {words of size n which

uses k many As and (n k) many Bs} and proceed as above to get

|B| =

n!
|A|
=
k!(n k)!
k!(n k)!

as the required answer. Observe that the above argument implies

n!
k!(nk)!

Z. We denote this

number by P (n; k). Note that P (n; k) = P (n; n k), Also, as per convention, P (n; k) = 0,

whenever k < 0 or n < k.

The above idea is further generalized below.

Definition 6.1.15. A multiset is a collection of objects where an object can appear more than
once. So, a set is a multiset. Note that {a, a, b, c, d} and {a, b, a, c, d} are the same 5-multisets.
Theorem 6.1.16. [Arrangements] Let us fix n, k N with 1 k n and let S be a multiset
k
P
containing ni N objects of i-th type, for i = 1, . . . , k with n =
ni . Then, there are
n!
(n1 + + nk )!
=
arrangements of the objects in S.
n1 !n2 ! nk !
n1 !n2 ! nk !

i=1

Proof. Assume that S consists of ni copies of Ai , i = 1, . . . , k. Put

A = {A11 , . . . , A1n1 , A21 , . . . , A2n2
B = {words of size

} and

made using elements of

}. For each arrangement a A,

define f (a) to be the word in B obtained by erasing the right subscripts of the objects of a.

Then, the function f : A B satisfies:

for each b, c B, b 6= c, we have |f 1 (b)| = |f 1 (c)| =
Thus, by Proposition 6.1.11, |B| =

|A|
n1 !nk !

and f 1 (b) f 1 (c) = .

(n1 ++nk )!
n1 !nk !

n!
n1 !n2 !nk ! .

Theorem 6.1.17. [Allocation I: distinct locations; identical objects (ni of type i); at most
one per place] Fix a positive integer k and for 1 i k, let Gi s be boxes containing ni N

identical objects. If the objects in distinct boxes are non-identical and n

k
P

ni then, the number

i=1

of allocations of the objects in n distinct locations l1 , . . . , ln , each location receiving at most one
object, is

n! P
n1 !nk !(n ni )! .

Proof. Consider a new group Gk+1 with nk+1 = n

k
P
1

ni objects of a new type. Notice that

an allocation of objects from G1 , . . . , Gk to n distinct places, where each location receives at

most one object, gives a unique arrangement of elements of G1 , . . . , Gk+1 .1 Thus, the number
1

Take an allocation of objects from G1 , . . . , Gk to n distinct places, where each location receives at most one

object. There are nk+1 locations which are empty. Supply an object from Gk+1 to each of these locations.
We have created an arrangement of elements of G1 , . . . , Gk+1 . Conversely, take an arrangement of elements of
G1 , . . . , Gk+1 . View this as an allocation of elements of G1 , . . . , Gk+1 to n distinct places. Empty the places which
have received elements from Gk+1 . We have created an allocation of elements of G1 , . . . , Gk to n distinct places,
where each location receives at most one object.

6.1. PERMUTATIONS AND COMBINATIONS

93

of allocations of objects from G1 , . . . , Gk to n distinct places, where each location receives at

most one object, is the same as the number of arrangements of elements of G1 , . . . , Gk+1 . By
Theorem 6.1.16, this number is

n! P
n1 !nk !(n ni )! .

n!
P
is den1 ! nk !(n ni )!
noted by P (n; n1 , . . . , nk ). Thus, P (6; 1, 1, 1) = P (6, 3). As a convention, P (n; n1 , . . . , nk ) = 0
k
P
whenever either ni < 0; for some i, 1 i k, or
ni > n. Many texts use C(n; n1 , , nk ) to
Definition 6.1.18. Let n, n1 , n2 , . . . , nk N. Then, the number

i=1

mean P (n; n1 , , nk ). We shall interchangeably use them.

Definition 6.1.19. [r-combination] An r-combination of an n-set S is an r-subset of S.

The number of r-subsets of an n-set is denoted by C(n, r). Thus, for any natural number n,
C(n, 0) = C(n, n) = 1.
Theorem 6.1.20. [Combination] C(n, r) = P (n; r) =

n!
r!(nr)! .

Proof. By Theorem 6.1.17, the number of allocations of r identical objects in n distinct places
(p1 , . . . , pn ) with each place receiving at most 1 is P (n; r). Note that each such allocation A
uniquely corresponds to a r-subset of [n], namely to {i | pi receives an object by A}. Thus,
C(n, r) = P (n; r) =

n!
r!(nr)! .

Example 6.1.21. In how many ways can you allocate 3 identical passes to 10 students so that
each student receives at most one? Ans: C(10, 3)
Theorem 6.1.22. [Pascal] C(n, r) + C(n, r + 1) = C(n + 1, r + 1).
Proof. By Theorem 6.1.20, C(n, r) =

n!
r!(nr)! .

Now verify the above identity to get the result.

Experiment
Complete the following list by filling the left list with all 3-subsets of [5] and the right list
with 3-subsets of [4] as well as with 2-subsets of [4] as shown below.

{1, 2, 3}

{2, 3, 4}

{1, 2, 5}
C(5, 3)

{3, 4, 5}

{1, 2, 3}

C(4, 3)

{2, 3, 4}

{1, 2}

C(4, 2)

{3, 4}

Theorem 6.1.23. [Alternate proof of Pascals Theorem 6.1.22] Here we supply a combinatorial proof, i.e., by associating the numbers with objects. Let S = [n + 1] and A be an
(r + 1)-subset of S. Then, there are C(n + 1, r + 1) such sets with either n + 1 A or n + 1 6 A.
Note that n + 1 A if and only if A \ {n + 1} is an r-subset of [n]. So, the number of

(r + 1)-subsets of [n + 1] which contain the element n + 1 is, by definition, C(n, r).

94

CHAPTER 6. COUNTING

Also, n + 1
/ A if and only if A is an (r + 1)-subset of [n]. So, a set A which does not contain

n + 1 can be formed in C(n, r + 1) ways. Hence, an (r + 1)-subset of S can be formed, by

definition, in C(n, r) + C(n, r + 1) ways. Thus, the required result follows.
Experiment
Here we consider subsets of [4]. Complete the following list by using 0s, 1s, xs and ys,
where x and y are commuting (xy = yx) symbols.

{1}
{2}
{3}
{4}
{1, 2}

0000
1000
0100
0010
0001
1100

yyyy = y 4
xyyy = xy 3
yxyy = xy 3
yyxy = xy 3
yyyx = xy 3
xxyy = x2 y 2

{1, 2, 3, 4}

1111

xxxx = x4

Remark 6.1.24. [Another alternate proof of Pascals Theorem 6.1.22] Here we supply
another combinatorial proof. An (r + 1) subset of [n + 1] may be viewed as a string (word)
of size n + 1 made of (n r) many 0s and (r + 1) many 1s. The number of such strings which

end with a 1 is C(n, r). The number of such strings which end with a 0 is C(n, r + 1). So, the
required conclusion follows.
Practice 6.1.25. Give a combinatorial proof of C(n, r) = C(n, n r), whenever n, r N with
0 r n.

Theorem 6.1.26. [Allocation II: distinct locations; distinct objects; ni at place i] The
number of ways of allocating objects o1 , . . . , on into pockets p1 , . . . , pk so that pocket pi contains
ni objects, is P (n; n1 , . . . , nk ).
Proof. Task has k compulsory parts: select n1 for pocket p1 and so on. So, the answer is
C(n, n1 )C(n n1 , n2 ) C(n n1 nk1 , nk ) = P (n; n1 , . . . , nk ).
Alternate. Take an allocation of o1 , . . . , on into pockets p1 , . . . , pk so that the pocket pi gets
ni objects. This is an allocation of n1 copies of p1 , , nk copies of pk into locations o1 , . . . , on
where each location gets exactly one. Hence, the answer is P (n; n1 , . . . , nk ).

Exercise 6.1.27.

1. In a class there are 17 girls and 20 boys. A committee of 5 students is

to be formed to represent the class.

(a) Determine the number of ways of forming the committee consisting of 5 students.

6.1. PERMUTATIONS AND COMBINATIONS

95

(b) Suppose the committee also needs to choose two different people from among themselves, who will act as spokesperson and treasurer. In this case, determine the
number of ways of forming a committee consisting of 5 students. Note that two committees are different if
i. either the members are different, or
ii. even if the members are the same, they have different students as spokesperson
and/or treasurer.
(c) Due to certain restrictions, it was felt that the committee should have at least 3 girls.
In this case, determine the number of ways of forming the committee consisting of 5
students (no one is to be designated as spokesperson and/or treasurer).
2. Combinatorially prove the following identities:
(a) kC(n, k) = nC(n 1, k 1).

(b) Newtons Identity: C(n, r)C(r, k) = C(n, k)C(n k, r k).

(c) C(n, r) = C(r, r)C(n r, 0) + C(r, r 1)C(n r, 1) + + C(r, 0)C(n r, r).

(d) C(n, 0)2 + C(n, 1)2 + + C(n, n)2 = C(2n, n).

3. Determine the number of ways of selecting a committee of m people from a group consisting
of n1 women and n2 men, with n1 + n2 m.
4. Determine the number of ways of arranging the letters of the word
(a) ABRACADABARAARCADA.
(b) KAGART HALAM N AGART HALAM .
5. How many anagrams of M ISSISSIP P I are there so that no two S are adjacent?
6. How many rectangles are there in an n n square? How many squares are there?
7. Show that a product of n consecutive natural numbers is always divisible by n!.
8. Show that (m!)n divides (mn)!.
9. If n points are placed on the circumference of a circle and all the lines connecting them are
joined, what is the largest number of points of intersection of these lines inside the circle
that can be obtained?
10. Prove that C(pn, pn n) is a multiple of p in two ways. Hint: Newtons identity.
11. How many ways are there to form the word MATHEMATICIAN starting from any side
and moving only in horizontal or vertical directions?

M
A

M
A
T

M
A
T
H

M
A
T
H
E

M
A
T
H
E
M

M
A
T
H
E
M
A

M
A
T
H
E
M
A
T

M
A
T
H
E
M
A
T
I

M
A
T
H
E
M
A
T
I
C

M
A
T
H
E
M
A
T
I
C
I

M
A
T
H
E
M
A
T
I
C
I
A

M
A
T
H
E
M
A
T
I
C
I
A
N

M
A
T
H
E
M
A
T
I
C
I
A

M
A
T
H
E
M
A
T
I
C
I

M
A
T
H
E
M
A
T
I
C

M
A
T
H
E
M
A
T
I

M
A
T
H
E
M
A
T

M
A
T
H
E
M
A

M
A
T
H
E
M

M
A
T
H
E

M
A
T
H

M
A
T

M
A

96

CHAPTER 6. COUNTING

12. (a) In how many ways can one arrange n different books in m different boxes kept in a
row, if books inside the boxes are also kept in a row?
(b) What if no box can be empty?
13. Prove by induction that 2n |(n + 1) (2n).

6.1.1

Multinomial theorem

Definition 6.1.28. Let x, y and z be commuting symbols. Then, by an algebraic expansion1

of (x + y + z)n we mean an expansion where each term is of the form xi y j z k so that two terms
differ in the degree of at least one of x, y, or z. By a word expansion2 of (x + y + z)n we
mean an expansion where each term is a word of length n using symbols x, y, z. Expansions for
(x1 + + xr )n , whenever xi s are commuting symbols, may be defined in a similar way.
Example 6.1.29.

1. x3 + 3xy 2 + y 3 + 3yx2 is an algebraic expansion of (x + y)3 , where as

xxx + xxy + xyx + xyy + yxx + yxy + yyx + yyy is a word expansion of (x + y)3 .
2. Take the word expansion of (X + Y + Z)9 . A term with exactly two Xs and exactly three
Y s is nothing but an arrangement of two Xs, three Y s and four Zs. So, the coefficient
of X 2 Y 3 Z 4 in the algebraic expansion of (X + Y + Z)9 is P (9; 2, 3, 4).
3. Consider (x+y+z)n = (x + y + z) (x + y + z) (x + y + z). Then, in this expression,
|
{z
}
n times

we need to choose, say

(a) i places from the n possible places for x (i 0),

(b) j places from the remaining n i places for y (j 0), and
(c) the n i j left out places for z (with n i j 0).
Thus, we get
(x + y + z)n =

i,j0,i+jn

C(n, i)C(n i, j)xi y j z nij =

P (n; i, j)xi y j z nij .

i,j0,i+jn

Theorem 6.1.30. [Multinomial Theorem] Fix a positive integer n and let x1 , x2 , . . . , xn be a

collection of commuting symbols. Then, for n = n1 + + nk , the coefficient of xn1 1 xn2 2 xnk k
in the algebraic expansion of (x1 + + xk )n is P (n; n1 , , nk ). So
(x1 + + xk )n =

n1 , . . . , nk 0
n1 + + nk = n

P (n; n1 , , nk ) xn1 1 xnk k .

Proof. The proof is left as an exercise for the reader.

As a special case, we have the famous binomial theorem.
Corollary 6.1.31. [Binomial Theorem] (x + y)n =

n
P

k=0
1
2

Nonstandard notion
Nonstandard notion

C(n, k)xnk y k . !!

6.2. CIRCULAR PERMUTATIONS

97

Example 6.1.32. Form words of size 5 using letters from MATHEMATICIAN (including
multiplicity, that is, you may use M at most twice). How many are there?
P
Ans:
C(5; k1 , , k8 ).
k1 ++k8 =5

k1 2,k2 3,k3 2,k4 1,k5 1,k6 2,k7 1,k8 1

1. Show that |P([n])| = 2n in the following ways.

Exercise 6.1.33.

(a) By using Binomial Theorem.

(b) By using select a subset is a task with n compulsory parts.
(c) By associating a subset with a 0-1 string of length n and evaluating their values in
base-2.
(d) Arguing in the line of a subset of [n + 1] either contains n + 1 or not and using
induction.
2. Let S be a set of size n. Then, prove in two different ways that the number of subsets
of S of odd size is the same as the number of subsets of S of even size, or equivalently
P
P
C(n, 2k + 1) = 2n1 .
C(n, 2k) =
k0

k0

3. Prove the following identities on Binomial coefficients.

(a)

n
P

C(k, )C(n, k) = C(n, )2n .

k=

(b) C(m + n, ) =

k=0

(c) C(n, ) =

t
P

k=0

C(m, k) C(n, k).

n
P

C(t, k) C(n t, k) =

(d) C(n + r + 1, r) =
(e) C(n + 1, r + 1) =

r
P

=0
n
P

k=0

C(n + , ).

C(t, k) C(n t, k), for any t, 0 t n.

C(, r).

=r

4. Evaluate

n
P

(2k + 1) C(n, 2k + 1) and

n
P

(5k + 3) C(n, 2k + 1), whenever n 3.

k=0

k=0

5. [Generalized Pascal] Assume that k1 + + km = n. Show that

C(n; k1 , . . . , km ) = C(n 1; k1 1, . . . , km ) + + C(n 1; k1 , . . . , km 1).
6. What is

C(n; k1 , . . . , km )?

k1 +...+km =n

7. Put l = m
2 . What is

6.2

(1)k2 +k4 ++k2l C(n; k1 , . . . , km )?

k1 +...+km =n

Circular permutations

Definition 6.2.1. [Circular permutation/arrangement] A circular permutation is an arrangement of n distinct objects on a circle. Two circular arrangements are the same if each
element has the same clockwise adjacent element. When |S| = n, we write a circular arrange-

ment of S to mean a circular arrangement of elements of S. By [x1 , x2 , . . . , xn , x1 ] we shall

denote a circular arrangement keeping the anticlockwise direction in picture.

98

CHAPTER 6. COUNTING

Example 6.2.2. Exactly two pictures in Figure 6.1 represent the same circular permutation.
A4

A3

A1

A5

A2

A3

A5

A2

A2

A4

A1

A4

A1

A3

A5

[A1 , A2 , A3 , A4 , A5 , A1 ]
Figure 6.1: Circular permutations
Example 6.2.3. Determine the number of circular permutations of X = {A1 , A2 , A3 , A4 , A5 }?

Ans: 4!. Proof. Let B = {circular permutations of X} and A = {permutations of X}.

Now, define f : A B as f (a) = b if a is obtained by breaking the cycle b at some gap and
then following in the anticlockwise direction. For example, if we break the leftmost circular
permutation in Figure 6.1 at the gap between A and B, we get [A2 , A3 , A4 , A5 , A1 ]. Notice that
|f 1 (b)| = 5, for each b B. Further if b, c B, then f 1 (b) f 1 (c) = (why?1 ). Thus, by

the principle of disjoint pre-images of equal size, the number of circular permutations is 5!/5.

Theorem 6.2.4. [circular permutations] The number of circular permutations of [n] is (n1)!.
Proof. A proof may be obtained on the line of the previous example. Here we give an alternate
proof. Put A = {circular permutations of [5]}. Put B = {permutations of [4]}. Define f :
A B as f ([5, x1 , x2 , x3 , x4 , 5]) = [x1 , x2 , x3 , x4 ]. Define g : B A as g([x1 , x2 , x3 , x4 ]) =

[5, x1 , x2 , x3 , x4 , 5]. Then, g f (a) = a, for each a A and f g(b) = b, for each b B. Hence,

by the bijection principle (see Theorem 2.3.8) f is a bijection.

Example 6.2.5. Find the number of circular arrangements of {A, B, B, C, C, D, D, E, E}.

Ans: There is only one A. Cutting A out from a circular arrangement we get a unique

arrangement of {B, B, C, C, D, D, E, E}. So, the required answer is

8!
2!4 .

Definition 6.2.6. [Rotation, Orbit size]

1. Given an arrangement [X1 , . . . , Xn ], by a rotation R1 ([X1 , . . . , Xn ]), in short R1 (X1 , . . . , Xn ),
we mean [X2 , . . . , Xn , X1 ] and by R2 (X1 , . . . , Xn ) we mean [X3 , . . . , Xn , X1 , X2 ]. On similar lines, we define Ri , i N and put R0 (X1 , . . . , Xn ) = [X1 , . . . , Xn ]. Thus, for each

k N,

R0 (X1 , . . . , Xn ) = Rkn (X1 , . . . , Xn ) = [X1 , . . . , Xn ].

2. The orbit size of an arrangement [X1 , . . . , Xn ] is the smallest positive integer i which
satisfies Ri (X1 , . . . , Xn ) = [X1 , . . . , Xn ]. In that case, we call
n
o
R0 (X1 , . . . , Xn ), R1 (X1 , . . . , Xn ), . . . , Ri1 (X1 , . . . , Xn )

the orbit of [X1 , . . . , Xn ].

1

Think of creating the circular permutation from a given permutation.

6.2. CIRCULAR PERMUTATIONS

Example 6.2.7.

99

1. We have R1 (ABCABCABC) = [BCABCABCA], R2 (ABCABCABC) =

[CABCABCAB] and R3 (ABCABCABC) = [ABCABCABC]. Thus, orbit size of ABCABCABC

is 3.
2. An arrangement of S = {A, A, B, B, C, C} with orbit size 6 is [AABCBC]. An arrangement of S with orbit size 3 is [ACBACB].

3. There is no arrangement of {A, A, B, B, C, C} with orbit size 2. In fact, if [X1 X2 X6 ]

is an arrangement with orbit size 2 then, [X1 X2 X3 X4 X5 X6 ] = [X3 X4 X5 X6 X1 X2 ]. Thus,

X1 = X3 = X5 which is not possible.

4. There is no arrangement of {A, A, B, B, C, C} with orbit size 1 or 2 or 4 or 5.
5. There are 3! arrangements of {A, A, B, B, C, C} with orbit size 3.
6. Take an arrangement of {A, A, B, B, C, C} with orbit size 3. Make a circular arrangement

by joining the ends. How many distinct arrangements can we generate by breaking the
circular arrangement at gaps?
Ans: 3. They are the elements of the same orbit.

7. Take an arrangement of {A, A, B, B, C, C} with orbit size 6. Make a circular arrangement

by joining the ends. How many distinct arrangements can we generate by breaking the
circular arrangement at gaps?
Ans: 6. They are the elements of the same orbit.

8. Take an arrangement of n elements with orbit size k. Make a circular arrangement by

joining the ends. How many distinct arrangements can we generate by breaking the circular
arrangement at gaps?
Ans: k. They are the elements of the same orbit.
9. If we take the set of all arrangements of a finite multiset and group them into orbits (notice
that each orbit gives us exactly one circular arrangement), then the number of orbits is
the number of circular arrangements.
Example 6.2.8. Find the number of circular arrangements of S = {A, A, B, B, C, C, D, D, E, E}.
Ans: There are two types of arrangements of S: one of orbit size 10 and the other of orbit

size 5. The number of arrangements of S with orbit size 5 is 5!. So, they can generate 4! distinct
circular arrangements. The number of arrangements of S with orbit size 10 is
Hence, they can generate

10!
2!2!2!2!2!10

of circular arrangements is 4! +

10!
2!2!2!2!2!

5!.

5!
10
distinct circular arrangements. Thus, the total number

10!
2!2!2!2!2!10

5!
10 .

Example 6.2.9. Suppose, we are given an arrangement [X1 , . . . , X10 ] of five As and five Bs.
Can it have an orbit size 3?
Ans: No. To see this assume that its orbit size is 3. Then,
[X1 , . . . , X10 ] = R3 (X1 , . . . , X10 ) = R6 (X1 , . . . , X10 ) = R9 (X1 , . . . , X10 ) = R2 (X1 , . . . , X10 ).

100

CHAPTER 6. COUNTING

Since 3 was the least positive integer with R3 (X1 , . . . , X10 ) = [X1 , . . . , X10 ], we arrive at a
contradiction. Hence, the orbit size cannot be 3.
Proposition 6.2.10. The orbit size of an arrangement of an n-multiset is a divisor of n.
Proof. Suppose, the orbit size of [X1 , . . . , Xn ] is k and n = kp + r, for some r, 0 < r < k. Then,
Rk (X1 , . . . , Xn ) = R2k (X1 , . . . , Xn ) = = Rkp (X1 , . . . , Xn ) = Rkr (X1 , . . . , Xn ).
Thus, Rkr (X1 , . . . , Xn ) = [X1 , . . . , Xn ], contradicting the minimality of k. Hence, a contradiction and therefore r = 0. Or equivalently, k divides n.
Proposition 6.2.11. Let S1 = {Pi1 , Pi2 , . . . , Pik } and S2 = {Pj1 , Pj2 , . . . , Pjl } be any two orbits
of certain arrangements of an n-multiset. Then, either S1 S2 = or S1 = S2 .

Proof. If S1 S2 = , then there is nothing to prove. So, let there exists an arrangement

Pt S1 S2 . Then, by definition, there exist rotations R1 and R2 such that R1 (Pi1 ) = Pt and

R2 (Pj1 ) = Pt . Thus, R21 (Pt ) = Pj1 and hence R21 (R1 (Pi1 )) = R21 (Pt ) = Pj1 . Therefore, we
see that the arrangement Pj1 S1 and hence S2 S1 . A similar argument implies that S1 S2

and hence S1 = S2 .

Definition 6.2.12. [Binary operation] Let [X1 , . . . , Xn ] and [Y1 , . . . , Yn ] be two arrangements
of an n-multiset. Then, in the remainder of this section,
1. we shall consider expressions like [X1 , . . . , Xn ] + [Y1 , . . . , Yn ].
2. by [Ri + Rj ](X1 , . . . , Xn ), we mean the expression Ri (X1 , . . . , Xn ) + Rj (X1 , . . . , Xn ).
3. by Ri ([X1 , . . . , Xn ]+[Y1 , . . . , Yn ]) we denote the expression Ri (X1 , . . . , Xn )+Ri (Y1 , . . . , Yn ).
Example 6.2.13. Think of all arrangements P1 , . . . , Pn , n =

6!
3!3! ,

of three As and three Bs.

How many copies of [ABCABC] are there in [R0 + + R5 ](P1 + + Pn )?

Ans: Of course 6.

To see this, note that R0 takes [ABCABC] to itself; R1 will take

[CABCAB] to [ABCABC]; R2 will take [BCABCA] to [ABCABC]; and so on.

Example 6.2.14. Let P = [X1 , . . . , X12 ] be an arrangement of a 12-multiset with orbit size 3.
Since, the orbit size of P is 3, the set S = {P, R1 (P ), R2 (P )} forms the orbit of P . Thus, the

rotations R0 , R3 , R6 and R9 fix each element of S, i.e., Ri (Rj (P )) = Rj (P ) for all i {0, 3, 6, 9}

and j {0, 1, 2}. In other words, [R0 + + R11 ](P ) accounts for 4 counts of the same circular
arrangement, where 4 is nothing but the number of rotations fixing P . Thus, we see that
[R0 + R1 + +R11 ](P + R1 (P ) + R2 (P ))
=

[R0 + R1 + R11 ](P ) + [R0 + R1 + R11 ](R1 (P ))

+[R0 + R1 + R11 ](R2 (P ))

4(P + R1 (P ) + R2 (P )) + 4(P + R1 (P ) + R2 (P )) + 4(P + R1 (P ) + R2 (P ))

12(P + R1 (P ) + R2 (P ))

6.2. CIRCULAR PERMUTATIONS

101

The proof of the next result is similar to the idea in the above example and hence is omitted.
Proposition 6.2.15. Let P1 , . . . , Pn be all the arrangements of an m-multiset. Then,
[R0 + + Rm1 ](P1 + + Pn ) = m(P1 + + Pn ).
Let P be an arrangement of an m-multiset with orbit size k. Then, by Proposition 6.2.10
k divides m. Now, from the understanding obtained from the above example, we note that
m
m
[R0 + + Rm1 ](P ) accounts for
counts of the same circular arrangement, where
is
k
k
nothing but the number of rotations fixing P . Also, by Proposition 6.2.11, we know that two
orbits are either disjoint or the same and hence the next two results are immediate. Therefore,
the readers are supposed to provide a proof of the following results.
Discussion 6.2.16. Let P1 , . . . , Pn be all the arrangements of an m-multiset. Then,
X

the number of rotations fixing Pi =

Pi

X
[R0 + + Rm1 ](Pi )
Pi

= m(P1 + + Pn )

= m(the number of circular arrangements).

Discussion 6.2.17. Let P1 , . . . , Pn be all the arrangements of an m-multiset and {R0 , R1 , . . . , Rm1 }
the set of all rotations. Then,
X

the number of rotations fixing Pi =

Pi

X
Pi

X
Rj

|{Rj | Rj (Pi ) = Pi }| = |{(Pi , Rj ) | Rj (Pi ) = Pi }|

|{Pi | Rj (Pi ) = Pi }|
the number of Pi s fixed by Rj .

Rj

Hence, using Discussion 6.2.16, the number of circular arrangements is

1
m
Example 6.2.18.
Ans: R0 fixes

Rj

the number of Pi s fixed by Rj .

a rotation

1. How many circular arrangements of {A, A, A, B, B, B, C, C, C} are there?

9!
3!3!3!

arrangements, None of R1 , R2 , R4 , R5 , R7 andR8 fixes any arrange-

ment, R3 and R6 fixes 3! arrangements, namely the 3! arrangements of X, Y, Z, where

X = AAA, Y = BBB and Z = CCC.
Thus, the number of circular arrangements is

1
9!
9 3!3!3!

i
+ 3! + 3! =

5678+12
9

564
3

= 188.

2. Determine the number of circular arrangements of size 5 using the alphabets A, B and C.
Ans: In this case, R0 fixes all the 35 arrangements. The rotations R1 , R2 , R3 and R4
fixes the arrangements AAAAA, BBBBB and CCCCC. Hence, the required number is


1
5
5 3 + 4 3 = 51.

Verify that the answer will be 8 if we have just two alphabets A and B.

102

CHAPTER 6. COUNTING

Exercise 6.2.19.

1. If there are n girls and n boys then what is the number of ways of

making them sit around a circular table in such a way that no two girls are adjacent and
no two boys are adjacent?
2. Persons P1 , . . . , P100 are seating on a circle facing the center and talking. If Pi talks lie,
then the
(a) person to his right talks truth. So, the minimum number of persons talking truth is
.
(b) second person to his right talks truth ? So, the minimum number of persons talking
truth is

(c) next two persons to his right talk truth ? So, the minimum number of persons talking
truth is

3. Let us assume that any two garlands are same if one can be obtained from the other by
rotation. Then, determine the number of distinct garlands that can be formed using 6
flowers, if the flowers
(a) are of 2 colors, say red and blue.
(b) are of 3 different colors.
(c) are of k different colors, for some k N.
(d) of red color are 2 and that of blue color is 4.

6.3

Solutions in nonnegative integers

Definition 6.3.1. [Solution in nonnegative integers] Recall that N0 := N {0}. A point

p = (p1 , . . . , pk ) Nk0 with p1 + + pk = n is called a solution of the equation x1 + + xk = n

in nonnegative integers or a solution of x1 + + xk = n in N0 . Two solutions (p1 , . . . , pk )

and (q1 , . . . , qk ) are said to be the same if pi = qi , for each i = 1, . . . , k. Thus, (5, 0, 0, 5) and
(0, 0, 5, 5) are two different solutions of x + y + z + t = 10 in N0 .
Example 6.3.2. Determine the number of
1. words which uses 3 As and 6 Bs.
2. arrangements of 3 As and 6 Bs.
3. distinct strings that can be formed using 3 As and 6 Bs.
4. solutions of the equation x1 + x2 + x3 + x4 = 6, where each xi N0 and 0 xi 6.
5. ways of placing 6 indistinguishable balls into 4 distinguishable boxes.
6. 3 subsets of an 9-set.
Solution: Observe that all the problems correspond to forming strings using +s (or |s

or bars) and 1s (or balls or dots) in place of Aa and Bs, respectively?

6.3. SOLUTIONS IN NONNEGATIVE INTEGERS

BBABBBABA
ABBBBBAAB
ABBBABABB

11 + 111 + 1+ = 2 + 3 + 1 + 0
+11111 + +1 = 0 + 5 + 0 + 1
+111 + 1 + 11 = 0 + 3 + 1 + 2

103

| ||
| ||
| | |

Figure 6.2: Understanding the three problems

Note that the As are indistinguishable among themselves and the same holds for Bs.
Thus, we need to find 3 places, from the 9 = 3 + 6 places, for the As. Hence, the answer
is C(9, 3). The answer will remain the same as we just need to replace As with +s (or
|s) and Bs with 1s (or balls) in any string of 3 As and 6 Bs. See Figure 6.2 or note that

four numbers can be added using 3 +s or four adjacent boxes can be created by putting
3 vertical lines or |s.
In general, we have the following result.
Theorem 6.3.3. [solutions in N0 ] The number of solutions of x1 + + xr = n in N0 is
C(n + r 1, n).

Proof. Each solution (x1 , . . . , xr ) may be viewed as an arrangement of n dots and r 1 bars.

Put x1 many dots; put a bar; put x2 many dots; put another bar; continue; and end by

putting xr many dots.

For example, (0, 2, 1, 0, 0) is associated to | | || and vice-versa. Thus, there are C(n + r

1, r 1) arrangements of n dots and r 1 bars.

Theorem 6.3.4. (a) The number of solutions of x1 + + xr n in nonnegative integers is

C(n + r, n).

(b) The number of terms in the algebraic expansion of (x1 + + xr )n is C(n + r 1, n).
Proof. (a) Any solution of x1 + +xr n uniquely corresponds to a solution of x1 + +xr +y =
n in nonnegative integers..

(b) Note that each term in the algebraic expansion of (x1 + +xr )n has the form xi11 xi22 xirr ,

with i1 +i2 + +ir = n. Thus, each term uniquely corresponds to a solution of i1 +i2 + +ir = n

in nonnegative integers.

Theorem 6.3.5. [r-multiset] The number of r-multisets of elements of [n] is C(n+r1, n1).
Proof. Let A be an r-multiset. Let di be the number of copies of i in A. Then, any solution of
d1 + + dn = r in nonnegative integers gives A uniquely. Hence, the conclusion.
Alternate. Put A = {arrangements of n 1 dots and r bars}. Put B = {r-multisets of [n]}.

For a A, define f (a) to be the multiset

f (a) = {d(i) + 1 | where d(i) is the number of dots to the left of the i-th bar}.
For example, ||||| gives us {1, 1, 3, 4, 4}. It is easy to define g : B A so that f (g(b)) = b, for

each b B and g(f (a)) = a, for each a A. Thus, by the bijection principle (see Theorem 2.3.8),
|A| = |B|. Also, we know that |A| = C(n + r 1, n 1) and hence the required result follows.

104

CHAPTER 6. COUNTING

Example 6.3.6.

1. There are 5 kinds of ice-creams available in our market complex. In how

many ways can you buy 15 of them for a party?

Ans: Suppose you buy xi ice-creams of the i-th type. Then, the problem is the same as
finding the number of solutions of x1 + + x5 = 15 in nonnegative integers.
2. How many solutions in N0 are there to x + y + z = 60 such that x 3, y 4, z 5?
Ans: (x, y, z) is such a solution if and only if (x3, y4, z5) is a solution to x+y+z = 48
in N0 . So, answer is C(50, 2).
3. How many solutions in N0 are there to x + y + z = 60 such that 20 x 3, 30 y
4, 40 z 5?

Ans: We are looking for solution in N0 of x + y + z = 48 such that x 17, y 26 and

z 35. Let A = {(x, y, z) N30 | x + y + z = 48}, Ax = {(x, y, z) N30 | x + y + z = 48, x

18}, Ay = {(x, y, z) N30 | x + y + z = 48, y 27} and Az = {(x, y, z) N30 | x + y + z =

48, z 36}. We know that |A| = C(50, 2). Our answer is then C(50, 2) |Ax Ay Az |.

Very soon we will learn to find the value of |Ax Ay Az |.

Exercise 6.3.7.

1. Determine the number of solutions of x + y + z = 7 with x, y, z N?

2. Find the number of allocations of n identical objects to r distinct locations so that location
i gets at least pi 0 elements, i = 1, 2, , r.
3. In how many ways can we pick integers x1 < x2 < x3 < x4 < x5 , from [20] so that
xi xi1 3, i = 2, 3, 4, 5? Solve in three different ways.
4. Find the number of solutions in nonnegative integers of a + b + c + d + e < 11.
5. In a room, there are 2 distinct book racks with 5 shelves each. Each shelf is capable of
holding up to 10 books. In how many ways can we place 10 distinct books in two racks?
6. How many 4-letter words (with repetition) are there with the letters in alphabetical order?
7. Determine the number of non-decreasing sequences of length r using the numbers 1, 2, . . . , n.
8. In how many ways can m indistinguishable balls be put into n distinguishable boxes with
the restriction that no box is empty.
9. How many 26-letter permutations of the ENGLISH alphabets have no 2 vowels together?
10. How many 26-letter permutations of the ENGLISH alphabets have at least two consonants
between any two vowels?
11. How many ways are there to select 10 integers from the set {1, 2, . . . , 100} such that the
positive difference between any two of the 10 integers is at least 3.

12. How many 10-element subsets of the ENGLISH alphabets do not have a pair of consecutive
letters?
13. How many 10-element subsets of the ENGLISH alphabets have a pair of consecutive letters?

6.4. SET PARTITIONS

105

14. How many ways are there to distribute 50 balls to 5 persons if Ram and Shyam together
get no more than 30 and Mohan gets at least 10?
15. How many arrangements of the letters of KAGARTHALAMNAGARTHALAM have no 2
vowels adjacent?
16. How many arrangements of the letters of RECURRENCERELATION have no 2 vowels
adjacent?
17. How many ways are there to arrange the letters in ABRACADABARAARCADA such
that the first
(a) A precedes the first B?
(b) B precedes the first A and the first D precedes the first C?
(c) B precedes the first A and the first A precedes the first C?
18. How many ways are there to arrange the letters in KAGART HALAM N AGART HAT AM
such that the first
(a) A precedes the first T ?
(b) M precedes the first G and the first H precedes the first A?
(c) M precedes the first G and the first T precedes the first G?
19. In how many ways can we pick 20 letters from 10 As, 15 Bs and 15 Cs?
20. Determine the number of ways to sit 10 men and 7 women so that no 2 women sit next to
each other?
21. How many ways can 8 persons, including Ram and Shyam, sit in a row with Ram and
Shyam not sitting next to each other?
22. Evaluate

i1 P
i2
n P
P

i1 =1 i2 =1 i3 =1

6.4

iP
k1

1.

ik =1

Set partitions

Definition 6.4.1. [Set partition] A partition of a set S is a collection of pairwise disjoint

nonempty subsets whose union is S.





Example 6.4.2. (a) {1, 2}, {3}, {4, 5, 6} , {1, 3}, {2}, {4, 5, 6} and {1, 2, 3, 4}, {5}, {6} are
both partitions of [6] into 3 subsets.

(b) There are 2n1 1 partitions of [n], n 2 into two subsets. To see this, observe that for

each nontrivial subset A P([n]), the set {A, Ac } is a partition of [n] into two subsets.

Since, the total number of nontrivial subsets of P([n]) equals 2n 2, the required result

follows.

(c) Number of allocations of 7 students into 7 different project groups so that each group has
one student, is 7! = C(7; 1, 1, 1, 1, 1, 1, 1) but the number of partitions of a set of 7 students
into 7 subsets is 1.

106

CHAPTER 6. COUNTING

(d) In how many ways can I write

n
o
{1, 2}, {3, 4}, {5, 6}, {7, 8, 9}, {10, 11, 12} on a piece of

paper, with the condition that sets have to be written in a row in increasing size?
Ans: Let us write a few first.
n
o
{1, 2}, {3, 4}, {5, 6}, {7, 8, 9}, {10, 11, 12}
n
o
{2, 1}, {3, 4}, {5, 6}, {7, 8, 9}, {10, 11, 12}
n
o
{5, 6}, {3, 4}, {1, 2}, {10, 11, 12}, {9, 7, 8}
n
o
{2, 3}, {1, 4}, {5, 6}, {7, 8, 9}, {10, 11, 12}
n
o
{2, 1}, {3, 4}, {7, 8, 9}, {5, 6}, {10, 11, 12}

correct
correct
correct
incorrect, not the same partition
incorrect, not satisfying the condition

There are 3!(2!)3 2!(3!)2 ways. Notice that from each written partition, if I remove the
brackets I get an arrangement of elements of [12].

(e) How many arrangements do I generate from a partition with pi subsets of size ni , n1 <
< nk ?
Ans: p1 !(n1 !)p1 pk !(nk !)pk =

k
Y
[pi !(ni )pi ].
i=1

Theorem 6.4.3. [Set partition] The number of partitions of [n] with pi subsets of size ni ,
n1 < < nk is
(n1

!)p1 p

n!
.
p
1 ! (nk !) k pk !

Proof. Note that each such partition generates

k
Q

[pi !(ni )pi ] arrangement of elements of [n].

i=1

Conversely, for each arrangement of elements of [n] we can easily construct a partition of the
above type which can generate this arrangement. Thus, the proof is complete.
Definition 6.4.4. Stirling numbers of the second kind, denoted S(n, r), is the number of
partitions of [n] into r-subsets (r-parts). By convention, S(n, r) = 1, if n = r and 0, whenever
either n > 0 and r = 0 or n < r.
Theorem 6.4.5. [recurrence for S(n, r)] S(n + 1, r) = S(n, r 1) + rS(n, r).
Proof. Write an r-partition of [n + 1] and erase n + 1 from it. That is, if {n + 1} is an element

of an r-partition, then the number of such partitions become S(n, r 1); else n + 1 appears in

one of the element of an r-partition of [n], which gives the number rS(n, r).

Example 6.4.6. Determine the number of ways of putting n distinguishable/distinct balls into
r indistinguishable boxes with the restriction that no box is empty.
Ans: Let A be the set of n distinct balls and let the balls in i-th box be Bi , 1 i r.
1. Since each box is non-empty, each Bi is non-empty.
2. Also, each ball is in some box and hence

r
S

i=1

Bi = A.

6.4. SET PARTITIONS

107

3. As the boxes are indistinguishable, we arrange the boxes in non-increasing order, i.e.,
|B1 | |Br |.
Thus, B1 , B2 , . . . , Br is a partition of A into r-parts. Hence, the required number of ways is
given by S(n, r), the Stirling number of the second kind.
To proceed further, consider the following example.
Example 6.4.7. Let A = {a, b, c, d, e} and S = {1, 2, 3}. Define an onto function f : A S by

f (a) = f (b) = f (c) = 1, f (d) = 2 and f (e) = 3. Then, f gives a partition B1 = {a, b, c}, B2 =

{d} and B3 = {e} of A into 3-parts. Also, let A1 = {a, d}, A2 = {b, e} and A3 = {c} be a
partition of A into 3-parts. Then, this partition gives 3! onto functions from A into S, each of

them being a one-to-one function from {A1 , A2 , A3 } to S, namely,

f1 (a) = f1 (d) = 1, f1 (b) = f1 (e) = 2, f1 (c) = 3, f1 (A1 ) = 1, f1 (A2 ) = 2, f1 (A3 ) = 3
f2 (a) = f2 (d) = 1, f2 (b) = f2 (e) = 3, f2 (c) = 2, f2 (A1 ) = 1, f2 (A2 ) = 3, f2 (A3 ) = 2
f3 (a) = f3 (d) = 2, f3 (b) = f3 (e) = 1, f3 (c) = 3, f3 (A1 ) = 2, f3 (A2 ) = 1, f3 (A3 ) = 3
f4 (a) = f4 (d) = 2, f4 (b) = f4 (e) = 3, f4 (c) = 1, f4 (A1 ) = 2, f4 (A2 ) = 3, f4 (A3 ) = 1
f5 (a) = f5 (d) = 3, f5 (b) = f5 (e) = 1, f5 (c) = 2, f5 (A1 ) = 3, f5 (A2 ) = 1, f5 (A3 ) = 2
f6 (a) = f6 (d) = 3, f6 (b) = f6 (e) = 2, f6 (c) = 1, f6 (A1 ) = 3, f6 (A2 ) = 2, f6 (A3 ) = 1.
Lemma 6.4.8. The total number of onto functions f : [r] [n] is n!S(r, n).
Proof. f is onto means for all y [n] there exists x [r], such that f (x) = y. Therefore, the
number of onto functions is 0, whenever r < n. So, we assume that r n. Then,

1. for each i [n], f 1 (i) = {x [r] | f (x) = i} is a non-empty set (f is onto).

2. f 1 (i) f 1 (j) = , whenever 1 i 6= j n (f is a function).
3.

n
S

f 1 (i) = [r] (domain of f is [r]).

i=1

Therefore, f 1 (i)s give a partition of [r] into n-parts. Also, note that each such function f ,
gives a one-to-one function from {f 1 (1), . . . , f 1 (r)} to [n].

Conversely, for each partition A1 , A2 , . . . , An of [r] into n-parts, we get n! one-to-one function

from {A1 , A2 , . . . , An } to [n]. Hence,





{f : [r] [n] | f is onto} = {g : {A1 , A2 , . . . , An } [n] | g is one-to-one}


Partition of [r] into n-parts
= n! S(r, n).

Thus, the required result follows.

Lemma 6.4.9. Let r, n N and = min{r, n}. Then,
nr =

X
k=1

C(n, k)k!S(r, k).

(6.1)

108

CHAPTER 6. COUNTING

Proof. Let A = {f | f : [r] [n]}. We compute |A| by two different methods.

Method 1: By Theorem 6.1.4, |A| = nr .

[r] [n] be any function. Then, f0 is an onto function from [r] to

S
Ak , where
Im (f0 ) = f0 ([r]). Moreover, , for some k, 1 k = min{r, n}. Thus, A =
Method 2: Let f0 :

k=1

Ak = {f : [r] [n] | |f ([r])| = k} and Ak Aj = , whenever 1 j 6= k . Now, using

Theorem 6.1.20, a subset of [n] of size k can be selected in C(n, k) ways. Thus, for 1 k ,

 

|Ak | = {K : K [n], |K| = k} {f : [r] K | f is onto} = C(n, k)k!S(r, k).

Therefore,

[
 X
X


|A| = 
Ai  =
|Ak | =
C(n, k)k!S(r, k).


k=1

k=1

k=1

Hence, using the two counting methods, the required result follows.
Remark 6.4.10.

1. The following two problems are equivalent.

(a) Count the number of onto functions f : [r] [n].

(b) Count the number ways to put r distinguishable/distinct balls into n distinguishable/distinct boxes so that no box is empty.
2. The numbers S(r, k) can be recursively calculated using Equation (6.1). For example, we
show that S(m, 1) = 1, for all m 1.
Ans: Take n 1 and r = 1 in Equation (6.1) to get n = n1 =

C(n, 1)1!S(1, 1) = nS(1, 1). Thus, S(1, 1) = 1.

Take n = 1 and r 2 in Equation (6.1) to get 1 = 1r =

P1

k=1 C(n, k)k!S(1, k)

P1

k=1 C(1, k)k!S(r, k)

= S(r, 1).

3. As exercise, verify that S(5, 2) = 15, S(5, 3) = 25, ; S(5, 4) = 10, S(5, 5) = 1.
Exercise 6.4.11.

1. Determine the number of ways of

(a) selecting r distinguishable objects from n distinguishable objects, when n r.

(b) distributing 20 distinct toys among 4 children if each children gets 5 toys?
(c) placing r distinguishable balls into n indistinguishable boxes if no box is empty?
(d) placing r distinguishable balls into n indistinguishable boxes?
2. For n N, let b(n) denote the number of partitions of [n]. Then, b(n) =

n
P

S(n, r) is

r=0

called the nth Bell number. By definition, b(0) = 1 = b(1). Determine b(n), for 2 n 5.
3. Fix n N. Then, a composition of n is an expression of n as a sum of positive integers.
For example, if n = 4, then the distinct compositions are

4, 3 + 1, 1 + 3, 2 + 2, 2 + 1 + 1, 1 + 1 + 2, 1 + 2 + 1, 1 + 1 + 1 + 1.
Let Sk (n) denote the number of compositions of n into k parts. Then, S1 (4) = 1, S2 (4) =
P
Sk (n).
3, S3 (4) = 3 and S4 (4) = 1. Determine Sk (n), for 1 k n and
k1

6.4. SET PARTITIONS

109

4. Let S = {f | f : [r] [n]}. Compute |S| in two ways to prove (n + 1)r =

r
P

C(r, k)nk .

k=0

5. Suppose 13 people get on the lift at level . If all the people get down at some level, say
1, 2, 3, 4 and 5 then, calculate the number of ways of getting down if at least one person

gets down at each level.

Definition 6.4.12. [Partition of a number] Let n, k N. A partition of n into k parts is
a tuple (x1 , , xk ) Nk written in non-increasing order such that x1 + + xk = n. It may

be viewed as a k-multiset S N with sum n. By n (k), we denote the number of partitions

of n into exactly k parts and by n , the number of partitions of n. Conventionally 0 = 1 and

n (k) = 0, whenever k > n.
Remark 6.4.13. 7 (4) = 3 as the partitions of 7 into 4-parts are 4 + 1 + 1 + 1, 3 + 2 + 1 + 1
and 2 + 2 + 2 + 1. Verify that 7 (2) = 3 and 7 (3) = 4.
Example 6.4.14. Determine the number of ways of placing r indistinguishable balls into n
indistinguishable boxes
1. with the restriction that no box is empty.
Ans: As the balls are indistinguishable, we need to count the number of balls in each box.
As the boxes are indistinguishable, arrange them so that the number of balls inside boxes
are in non-increasing order. Also, each box is non-empty and hence the answer is r (n).
2. with no restriction.
Ans: Let us place one ball in each box. Now placing r indistinguishable ball into n
indistinguishable boxes with no restriction is same as placing r + n indistinguishable
balls into n indistinguishable boxes so that no box is empty. Therefore, the required
answer is m+n (n).
Exercise 6.4.15.

1. Calculate (n), for n = 1, 2, 3, . . . , 8.

2. Prove that 2r (r) = (r), for any r N.

3. For a fixed n N determine a recurrence relation for the numbers n (r)s for 1 r n.
Definition 6.4.16. [Stirling number of first kind] The Stirling number of the first kind,
denoted s(n, k), is the coefficient of xk in xn, where xn is called the falling factorial and equals
x(x 1)(x 2) (x n + 1). The rising factorial xn is defined as x(x+ 1)(x+ 2) (x+ n 1).
Exercise 6.4.17. Prove by induction that
1. s(n, m)(1)nm is the coefficient of xm in xn and |s(n, m)| = s(n, m)(1)nm .
2. Let a(n, k) denote the number of permutations of [n] which have k disjoint cycles. For
example, a(4, 2) = 11 as it corresponds to the permutations (12)(34), (13)(24), (14)(23),
(1)(234), (1)(243), (134)(2), (143)(2), (124)(3), (142)(3), (123)(4) and (132)(4). By convention, a(0, 0) = 1 and a(n, 0) = 0 = a(0, n), whenever n 1. Determine prove that the
numbers a(n, k)s satisfy

a(n, k) = (n 1)a(n 1, k) + a(n 1, k 1).

110

CHAPTER 6. COUNTING

3. Prove that a(n, m) = |s(n, m)| for all n, m N0 .

6.5

Lattice paths and Catalan numbers

Consider a lattice of integer lines in R2 and let S = {(m, n) | m, n = 0, 1, . . .} be the said of

points on the lattice. For a pair of points, say A = (m1 , n1 ) and B = (m2 , n2 ) with m1 m2

and n1 n2 , we define a lattice path from A to B to be a subset {e1 , . . . , ek } of S such that

if ei = (x, y) then ei+1 is either (x + 1, y) or (x, y + 1), for 1 i k 1. That is, at each step

we move either one unit right, denoted R, or one unit up, denoted U (see Figure 6.3).

(8, 7)

U = UP

(2, 3)

R = RIGHT

(0, 0)
Figure 6.3: A lattice with a lattice path from (2, 3) to (8, 7)
Example 6.5.1.

1. Determine the number of lattice paths from (0, 0) to (m, n).

Ans: As at each step, the unit increase is either R or U , we need to take n many R steps
and m many U steps to reach (m, n) from (0, 0). So, any arrangement of n many Rs and
m many U s will give such a path uniquely. Hence, the answer is C(m + n, m).
2. Use the method of lattice paths to prove

m
P

C(n + , ) = C(n + m + 1, m).

=0

Ans: Observe that C(n + m + 1, m) is the number of lattice paths from (0, 0) to (m, n + 1)
and the left hand side is the number of lattice paths from (0, 0) to (, n), where 0 m.
Fix , 0 m and let P be a lattice path from (0, 0) to (, n). Then, the path P Q,

where Q = U RR R with R appearing m times, gives a lattice path from (0, 0) to

(m, n + 1), namely

(0, 0)
(, n)
(, n + 1)
(m, n + 1).
These lattice paths for 0 m are all distinct and hence the result follows.
Exercise 6.5.2.

1. Give a bijection between the solution set of x0 + x1 + x2 + + xk = n

in non-negative integers and the number of lattice paths from (0, 0) to (n, k).
2. Use lattice paths to construct a proof of

n
P

k=0

C(n, k) = 2n .

6.5. LATTICE PATHS AND CATALAN NUMBERS

3. Use lattice paths to construct a proof of

n
P

111

C(n, k)2 = C(2n, n). [Hint: C(n, k) is the

k=0

number of lattice paths from (0, 0) to (n k, k) as well as from (n k, k) to (n, n).]

Discussion 6.5.3. As observed earlier, the number of lattice paths from (0, 0 to (n, n) is
C(2n, n). Suppose, we wish to take paths so that at no step the number of U s exceeds the
number of Rs. Then, what is the number of such paths?
Ans: Call an arrangement of n many U s and n many Rs a bad path if the number of U s
exceeds the number of Rs at least once. For example, the path RRU U U RRU is a bad path.
To each such arrangement, we correspond another arrangement of n + 1 many U s and n 1
many Rs in the following way: spot the first place where the number of U s exceeds that of

Rs in the bad path. Then, from the next letter onwards change R to U and U to R. For
example, the bad path RRU U U RRU corresponds to the path RRU U U U U R. Notice that this
is a one-one correspondence. Thus, the number of bad paths is C(2n, n 1). So, the answer to
C(2n, n)
the question is C(2n, n) C(2n, n 1) =
.
n+1
Definition 6.5.4. [Catalan number] The nth Catalan number, denoted Cn , is the number
of different representations of the product A1 An+1 of n + 1 square matrices of the same size
using n pairs of brackets. By convention C0 = 1.

Theorem 6.5.5. [Catalan number] Prove that Cn =

C(2n,n)
n+1

for all n N.

Proof. Claim: After the (n k)-th (, there are at least k + 2 many As. To see this pick the

substring starting right from the (n k)-th ( till we face (k + 1) many )s. This substring
represents a product of matrices. So, it must contain (k + 2) many Ai s.

Given one representation of the product, replace each Ai by A. Drop the right brackets to have
a sequence of n many (s and n + 1 many As. Thus, the number of As used till the n kth (

is at most n + 1 (k + 2) = n k 1. So, the number of As never exceeds the number of (.

Conversely, given such an arrangement, we can put back the )s: find two consecutive letters
from the last (; put a right bracket after them; treat (AA) as a letter; repeat the process. For

example,
((A((AAAA ((A((AA)AA ((A((AA)A)A ((A((AA)A))A = ((A((AA)A))A)
By previous example the number of such arrangements is

C(2n,n)
n+1 .

The readers who are interested in knowing more about Catalan numbers should look at the
book enumerative combinatorics by Stanley [11].
Exercise 6.5.6.

1. Give a recurrence relation for Cn s (i.e., a formula for Cn involving

C0 , . . . , Cn1 ). Hence, show that Cn = C(2n, n)/(n + 1).

2. Give an arithmetic proof of the fact that (n + 1) divides C(2n, n).
3. A man is standing on the edge of a swimming pool (facing it) holding a bag containing n
blue and n red balls. He randomly picks up one ball at a time and discards it. If the ball
is blue he takes a step back and if the ball is red, he takes a step forward. What is the
probability of his falling into the swimming pool?

112

CHAPTER 6. COUNTING

4. Consider a regular polygon with vertices 1, 2, , n. In how many ways can we divide the
polygon into triangles using (n 3) noncrossing diagonals?

5. How many arrangements of n blue and n red balls are there such that at any position in
the arrangement the number of blue balls (till that position) is at most one more than the
number of red balls (till that position)?
6. We want to write a matrix of size 10 2 using numbers 1, . . . , 20 with each number appearing exactly once. Then, determine the number of such matrices in which the numbers
(a) increase from left to right?
(b) increase from up to down?
(c) increase from left to right and up to down?

6.6

Some Generalizations

1. Let n, k N with 0 k n. Then, in Theorem 6.1.20, we saw that C(n, k) =

n!
.
k!(n k)!

n (n 1) (n k + 1)
. With this understanding, we
k!
generalize C(n, k) for any n R and k N0 as follows:

0,
if k < 0

0,
if n = 0, n 6= k
(6.2)
C(n, k) =
1,
if n = k

n (n 1) (n k + 1) ,
otherwise.
k!

Hence, we can think of C(n, k) =

With the notations as above, we give the generalized binomial theorem without proof.
Theorem 6.6.1. [Generalized binomial theorem] Let n be any real number. Then,
(1 + x)n = 1 + C(n, 1)x + C(n, 2)x2 + + C(n, r)xr + .
In particular, (1 x)1 = 1 + x + x2 + x3 + and if a, b R with |a| < |b|, then
 a r X

X
a n
C(n, r)
= bn
=
C(n, r)ar bnr .
(a + b)n = bn 1 +
b
b
r0

r0

Let us now understand Theorem 6.6.1 through the following examples.

1
(a) Let n = . In this case, for k 1, Equation (6.2) gives
2




1
1
1

1
k+1
1 (1) (3 2k)
(1)k1 (2k 2)!
1
2
2
2
=
=
.
C( , k) =
2
k!
2k k!
22k1 (k 1)!k!
Thus,
(1 + x)1/2 =

X
k0

X (1)k1 (2k 2)!

1
1
1
1
xk .
C( , k)xk = 1 + x + 3 x2 + 4 x3 +
2
2
2
2
22k1 (k 1)!k!
k4

6.6. SOME GENERALIZATIONS

113

The above expression can also be obtained by using the Taylor series expansion of
f (x) = (1 + x)1/2 around x = 0. Recall that the Taylor series expansion of f (x)

P f (k) (0) k
around x = 0 equals f (x) = f (0) + f (0)x + f 2!(0) x2 +
k! x , where f (0) = 1,
k3

f (0)

1
2,

f (0)

1
22

and in general

f (k) (0)

1
2

( 12

1) ( 21 k + 1), for k 3.

(b) Let n = r, where r N. Then, for k 1, Equation (6.2) gives C(r, k) =

r (r 1) (r k + 1)
= (1)k C(r + k 1, k). Thus,
k!
X
1
2
(1 + x)n =
=
1

rx
+
C(r
+
1,
2)x
+
C(r + k 1, k)(x)k .
(1 + x)r
k3

2. Let n, m N. Recall the identity nm =

Equation (6.1). Note that for

0m
m
C(0, 0)
1

m
C(1, 0)
2
C(2, 0)

X=
3m , A =
.

.
.
.
..

C(n, 0)
nm

m
P

C(n, k)k!S(m, k) =

n
P

C(n, k)k!S(m, k) in

k=0

k=0

each m N, the above identity equals X = AY , where

0

0!S(m, 0)

1!S(m, 1)

C(2, 1) C(2, 2)
0
and Y = 2!S(m, 2) .

..
..
..
..
..

.
.
.
.
.

n!S(m, n)
C(n, 1) C(n, 2) C(n, n)
C(1, 1)

As A is lower triangular with det(A) = 1, it has an inverse and each entry of A1 has a
similar form. So, Y = A1 X, where

C(0, 0)
0
0
0

C(1, 0)
C(1, 1)
0
0

C(2, 1)
C(2, 2)
0
C(2, 0)
A1 =
C(3, 0)
C(3, 1)
C(3, 2)
C(3, 3)

.
.
.
..

..
..
..
.

n
n1
n2
n3
(1) C(n, 0) (1)
C(n, 1) (1)
C(n, 2) (1)
C(n, 3)

Hence, for n, m N, we have

S(m, n) =

..
.

0
.
0

..
.

C(n, n)
0

1 X
(1)k C(n, k)(n k)m .
n!

(6.3)

k0

3. The above matrix inversion implies that for n N0 , the identity

a(n) =

C(n, k)b(k) holds if and only if b(n) =

k0

X
(1)k C(n, k)a(k) holds.
k0

We end this chapter with another set of exercises.

Exercise 6.6.2.

1. Prove that there exists a bijection between any two of the following sets.

(a) The set of words of length n on an alphabet consisting of m letters.

114

CHAPTER 6. COUNTING
(b) The set of maps of an n-set into an m-set.
(c) The set of distributions of n distinct objects into m distinct boxes.
(d) The set of n-tuples on m letters.

2. Prove that there exists a bijection between any two of the following sets.
(a) The set of n letter words with distinct letters out of an alphabet consisting of m letters.
(b) The set of one-one functions from an n-set into an m-set.
(c) The set of distributions of n distinct objects into m distinct boxes, subject to if an
object is put in a box, no other object can be put in the same box.
(d) The set of n-tuples on m letters, without repetition.
(e) The set of permutations of m symbols taken n at a time.
3. Prove that there exists a bijection between any two of the following sets.
(a) The set of increasing words of length n on m ordered letters.
(b) The set of distributions on n non-distinct objects into m distinct boxes.
(c) The set of combinations of m symbols taken n at a time with repetitions permitted.

Chapter 7

Advanced Counting Principles

7.1

Pigeonhole Principle

Discussion 7.1.1. [Pigeonhole principle (PHP)]

(PHP1) If n + 1 pigeons stay in n holes then there is a hole with at least two pigeons.
(PHP2) If kn + 1 pigeons stay in n holes then there is a hole with at least k + 1 pigeons.
(PHP3) If p1 + + pn + 1 pigeons stay in n holes then there is a hole i with at least pi + 1
pigeons.

Example 7.1.2.

1. Consider a tournament of n > 1 players, where each pair plays exactly

once and each player wins at least once. Then, there are two players with the same number
of wins.
Ans: Number of wins vary from 1 to n 1 and there are n players.
2. A bag contains 5 red, 8 blue, 12 green and 7 yellow marbles. The least number of marbles
to be chosen to ensure that there are
(a) at least 4 marbles of the same color is 13,
(b) at least 7 marbles of the same color is 24,
(c) at least 4 red or at least 7 of any other color is 22.
3. In a group of 6 people, prove that there are three mutual friends or three mutual strangers.
Ans: Let a be a person in the group. Let F be the set of friends of a and S the set of
strangers to a. Clearly |S| + |F | = 5. By PHP either |F | 3 or |S| 3.
Case 1: |F | 3. If any two in F are friends then those two along with a are three mutual

friends. Else F is a set of mutual strangers of size at least 3.

Case 2: |S| 3. If any pair in S are strangers then those two along with a are three

mutual strangers. Else S becomes a set of mutual friends of size at least 3.

4. If 7 points are chosen inside or on the unit circle, then there is a pair of points which are
at a distance at most 1.
115

116

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

Ans: To see this divide the circle into 6 equal cone type parts creating an angle of 60o
with the center. By PHP there is a part containing at least two points. The distance
between these two is at most 1.

5. If n + 1 integers are selected from [2n], then there is a pair which has the property that
one of them divides the other.
Ans: Each number has the form 2k O, where O is an odd number. There are n odd
numbers. If we select n + 1 numbers from S, by PHP some two of them (say, x, y) have
the same odd part, that is, x = 2i O and y = 2j O. If i j, then x|y, otherwise y|x.
6. (a) Let r1 , r2 , , rmn+1 be a sequence of mn + 1 distinct real numbers. Then, prove that

there is a subsequence of m + 1 numbers which is increasing or there is a subsequence

of n + 1 numbers which is decreasing.
Ans: Define li to be the maximum length of an increasing subsequence starting at
ri . If some li m + 1 then we have nothing to prove. So, let 1 li m. Since

(li ) is a sequence of mn + 1 integers, by PHP, there is one number which repeats at

least n + 1 times. Let li1 = li2 = = lin+1 = s, where i1 < i2 < < in+1 . Notice

that ri1 > ri2 , because if ri1 ri2 , then ri1 together with the increasing sequence
of length s starting with ri2 gives an increasing sequence of length s + 1. Similarly,

ri2 > ri3 > > rin+1 and hence the required result holds.
Alternate.

Let S = {r1 , r2 , , rmn+1 } and define a map f : S Z Z by

f (ai ) = (s, t), for 1 i mn + 1, where s equals the length of the largest increasing

subsequence starting with ai and t equals the length of the largest decreasing subsequence ending at ai . Now, if either s m + 1 or t n + 1, we are done. If not,

then note that 1 s m and 1 t n. So, the number of tuples (s, t) is at most

mn. Thus, the mn + 1 distinct numbers are being mapped to mn tuples and hence
by PHP there are two numbers ai 6= aj such that f (ai ) = f (aj ). Now, proceed as in
the previous case to get the required result.

(b) Does the above statement hold for every collection of mn distinct numbers? No.
Consider the sequence:
n, n1, , 1, 2n, 2n1, . . . , n+1, 3n, 3n1, , 2n+1, , mn, mn1, , mnn+1.
7. Given any 1010 integers, prove that there is a pair that either differ by, or sum to, a
multiple of 2017. Is this true if we replace 1010 by 1009?
Ans: Let the numbers be n1 , n2 , . . . , n1010 and S = {n1 nk , n1 + nk : k = 2, . . . , 1010}.

Then, |S| = 2018 and hence, at least two of them will have the same remainder when di-

vided by 2017. Then, consider their difference. For the later part, consider {0, 1, 2, . . . , 1008}.

8. Let a Qc . Then, there are infinitely many rational numbers

p
q

such that |a pq | <

1
.
q2

Ans: Enough to show that there are infinitely many (p, q) Z2 with |qa p| < 1q . Note

that for every m N, 0 < ia ia < 1, for i = 1, . . . , m + 1. Hence, by PHP there exist

7.1. PIGEONHOLE PRINCIPLE

117

i, j with i < j such that

|(j i)a (ja ia)| <

1
1

.
m
ji

Then, the tuple (p1 , q1 ) = (ja ia, j i) satisfies the required property. To generate

another tuple, find m2 such that

1
p1
< |a |
m2
q1

and proceed as before to get (p2 , q2 ) such that |q2 a p2 | <

|a

p1
q1 |,

we have

p1
q1

6=

p2
q2 .

1
m2

1
q2 .

Since |a pq22 | <

1
m2

<

Now use induction to get the required result.

9. Prove that there exist two powers of 3 whose difference is divisible by 2017.
Ans: Let S = {1 = 30 , 3, 32 , 33 , . . . , 32017 }. Then, |S| = 2018. As the remainders of any

integer when divided by 2017 is 0, 1, 2, . . . , 2016, by PHP, there is a pair which has the
same remainder. Hence, 2017 divides 3j 3i for some i, j.
10. Prove that there exists a power of three that ends with 0001.
Ans: Let S = {1 = 30 , 3, 32 , 33 , . . .}. Now, divide each element of S by 104 . As |S| > 104 ,

by PHP, there exist i > j such that the remainders of 3i and 3j , when divided by 104 , are

equal. But gcd(104 , 3) = 1 and thus, 104 divides 3 1. That is, 3 1 = s 104 for some
positive integer s. That is, 3 = s 104 + 1 and hence the result follows.

Exercise 7.1.3.

1. Consider the poset (X = P([4]), ). Write 6 maximal chains P1 , . . . P6

(need not be disjoint) such that Pi = X. Let A1 , . . . , A7 be 7 distinct subsets of [4]. Use
i

PHP, to prove that there exist i, j such that Ai , Aj Pk , for some k. That is, {A1 , . . . , A7 }

cannot be an anti-chain. Conclude that this holds as the width of the poset is 6.
9
P
2. Let {x1 , . . . , x9 } N with
xi = 30. Then, there exist i, j, k [9] with xi + xj + xk 12.
i=1

3. Pick any 6 integers from [10], then there exists a pair with odd sum.

4. Any 14-subset of [46] has four elements a, b, c, d such that a + b = c + d.

5. In a row of 12 chairs 9 are filled. Then, some 3 consecutive chairs are filled. Will 8 work?
6. Every n-sequence of integers has a consecutive subsequence with sum divisible by n.
7. Let n > 3 and S [n] of size m = n+2
2 + 1. Then, there exist a, b, c S such that
a + b = c.

8. Let a, b N, a < b. Given more than half of the integers in the set [a + b], there is a pair
which differ by either a or b.

9. Consider a chess board with two of the diagonally opposite corners removed. Is it possible to
cover the board with pieces of rectangular dominos whose size is exactly two board squares?
10. Mark the centers of all squares of an 8 8 chess board. Is it possible to cut the board with
13 straight lines not passing through any center, so that every piece had at most 1 center?

11. Fifteen squirrels have 100 nuts. Then, some two squirrels have equal number of nuts.

118

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

12. Suppose that f (x) is a polynomial with integer coefficients. If

(a) f (x) = 2 for three distinct integers, then for no integer x, f (x) can be equal to 3.
(b) f (x) = 14 for three distinct integers, then for no integer x, f (x) can be equal to 15.
(c) f (x) = 11 for five distinct integers, then for no integer x, f (x) can be equal to 9.
13. Choose 5 points at random inside an equilateral triangle of side 1 unit, then there exists a
pair which have distance at most 0.5 units.
14. Prove that among any 55 integers 1 x1 < x2 < x3 < < x55 100, there is a pair with
difference 9, a pair with difference 10, a pair with difference 12 and a pair with difference
13. Surprisingly, there need not be a pair with difference 11.
15. Let {x1 , x2 , . . . , xn } Z. Prove that there exist 1 i < j n such that
(a) xi + xi+1 + + xj1 + xj is a multiple of 2017, whenever n 2017.
(b) xj + xi or xj xi is a multiple of 2017, whenever n 1009.
16. Let A and B be two discs, each having 2n equal sectors. On disc A, n sectors are colored
red and n are colored blue. The sectors of disc B are colored arbitrarily with red and blue
colors. Show that there is a way of putting the two discs, one above the other, so that at
least n corresponding sectors have the same colors.
17. There are 7 distinct real numbers. Is it possible to select two of them, say x and y such
that 0 <

xy
1+xy

<

1 ?
3

18. If n is odd then for any permutation p of [n] the product

n
Q

i=1

19. Fix a positive

Qc .

i p(i) is even.

Then, S = {m + n : m, n Z} is dense in R.

20. Take 25 points on a plane satisfying among any three of them there is a pair at a distance
less than 1. Then, some circle of unit radius contains at least 13 of the given points.
21. Five points are chosen at the nodes of a square lattice (view Z Z). Why is it certain that
a mid-point of some two of them is a lattice point?

22. Each of the given 9 lines cuts a given square into two quadrilaterals whose areas are in the
ratio 2 : 3. Prove that at least three of these lines pass through the same point.
23. If more than half of the subsets of [n] are selected, then some two of the selected subsets
have the property that one is a subset of the other.
24. Given any ten 4-subsets of [11], some two of them have at least 2 elements in common.
25. A person takes at least one aspirin a day for 30 days. If he takes 45 aspirin altogether, in
some sequence of consecutive days he takes exactly 14 aspirins.
26. If 58 entries of a 14 14 matrix are 1, then there is a 2 2 submatrix whose all entries 1.
Exercise 7.1.4.

1. If each point of a circle is colored either red or blue, then show that there

exists an isosceles triangle with vertices of the same color.

2. Each point of the plane is colored red or blue, then prove the following.

7.2. PRINCIPLE OF INCLUSION AND EXCLUSION

119

(a) There exist two points of the same color which are at a distance of 1 unit.
(b) There is an equilateral triangle all of whose vertices have the same color.
(c) There is a rectangle all of whose vertices have the same color.
3. Let S [100] be a 10-set. Then, some two disjoint subsets of S have equal sum.
4. For n N, prove that there exists a N such that n divides 2 1.
5. Does there exist a multiple of 2017 that is formed using only the digits
(a) 2? Justify your answer.
(b) 2 and 3 and the number of 2s and 3s are equal? Justify your answer.
6. Each natural number has a multiple of the form 9 90 0, with at least one 9.

7.2

Principle of Inclusion and Exclusion

We start this section with the following example.

Example 7.2.1. How many natural numbers n 1000 are not divisible by any of 2, 3?

Ans: Let A2 = {n N | n 1000, 2|n} and A3 = {n N | n 1000, 3|n}. Then,

|A2 A3 | = |A2 | + |A3 | |A2 A3 | = 500 + 333 166 = 667. So, the required answer is
1000 667 = 333.

We now generalize the above idea whenever we have 3 or more sets.

Theorem 7.2.2. [Principle of inclusion and exclusion] Let A1 , , An be finite subsets of a
set U . Then,

n
 n
 X
 Ai  =
(1)k+1
i=1

k=1

1i1 <<ik n




Ai Ai  .
1
k

(7.1)

Or equivalently, the number of elements of U which are in none of A1 , A2 , . . . , An equals

n
X
 n

|U |  Ai  = |U |
(1)k
i=1

k=1

1i1 <<ik n




Ai1 Ai  .
k

Proof. Let x
/ Ai . Then, we show that inclusion of x in some Ai contributes (increases
i=1

the value) 1 to both sides of Equation (7.1). So, assume that x is included only in the sets
A1 , , Ar . Then, the contribution of x to |Ai1 Aik | is 1 if and only if {i1 , . . . , ik } [r].
P
|Ai1 Aik | is C(r, k). Thus, the contribution
Hence, the contribution of x to
1i1 <<ik n

of x to the right hand side of Equation (7.1) is

r C(r, 2) + C(r, 3) + (1)r+1 C(r, r) = 1.

The element x clearly contributes 1 to the left hand side of Equation (7.1) and hence the required
result follows. The proof of the equivalent condition is left for the readers.

120

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

Example 7.2.3. How many integers between 1 and 10000 are divisible by none of 2, 3, 5, 7?
Ans: For i {2, 3, 5, 7}, let Ai = {n N | n 10000, i|n}. Therefore, the required answer is

10000 |A2 A3 A5 A7 | = 2285.

Definition 7.2.4. [Euler totient function] For a fixed n N, the Eulers totient function

is defined as (n) = |{k N : k n, gcd(k, n) = 1}|.

Theorem 7.2.5. Let n =

k
Q

i=1

pi i , be a factorization of n into distinct primes p1 , . . . , pk . Then,

1

1

1

1
1
.
p1
p2
pk
Proof. For 1 i k, let Ai = {m N : m n, pi |m}. Then,

(n) = n 1

k
h
X
1
+
(n) = n | Ai | = n 1
i
pi
i=1

as |Ai | =

n
pi ,

1i<jk

1

1

1

= n 1
1
1
p1
p2
pk

|Ai Aj | =

n
pi pj

i
1
1
+ (1)k
pi pj
p1 p2 pk

and so on. Thus, the required result follows.

Definition 7.2.6. [Derangement]A derangement of objects in a finite set S is a permutation/arrangement on S such that for all x, (x) 6= x.
For example, 2, 1, 4, 3 is a derangement of 1, 2, 3, 4. The number of derangements of 1, 2, . . . , n
is denoted by Dn . By convention, D0 = 1. Also, we use a b to mean that b is an approximate

value of a.

Theorem 7.2.7. For n N, Dn = n!

n
X
(1)k
k=0

k!

. Thus,

Dn
1
.
n!
e

Proof. For each i, 1 i n, let Ai be the set of arrangements such that (i) = i. Then,

verify that |Ai | = (n 1)!, |Ai Aj | = (n 2)! and so on. Thus,

| Ai | = n.(n 1)! C(n, 2)(n 2)! + + (1)n1 C(n, n)0! = n!

i

So, Dn = n! Ai = n!
i

n
P

k=0

(1)k
k!

1
Dn
= .
. Furthermore, lim
n n!
e

n
X
(1)k1
k=1

k!

Example 7.2.8. For n N, how many squarefree integers do not exceed n?

Ans: Let P = {p1 , , ps } be the set of primes not exceeding n and for 1 i s, let Ai

be the set of integers between 1 and n that are multiples of p2i . It is easy to see that
n
n
|Ai | = 2 , |Ai Aj | = 2 2 ,
pi
pi pj
and so on. So, the number of squarefree integers not greater than n is
s
X
n
2 +
n | Ai | = n
i=1
pi
i=1
s

1i<js

2
pi p2j

p2 p2 p2
1i<j<ks i j k

For n = 100, we have P = {2, 3, 5, 7}. So, the number of squarefree integers not exceeding 100
is

100

100
100
100
100
100
100

+
+
= 61.
4
9
25
49
36
100

7.2. PRINCIPLE OF INCLUSION AND EXCLUSION

121

1. Let m, n N with gcd(m, n) = 1. Then, (mn) = (m)(n).

r
1 P
2. Let n N. Then, use inclusion-exclusion to prove S(n, r) =
(1)i C(r, i)(r i)n .
r! i=0

Exercise 7.2.9.

3. In a school there are 12 students who take an art course A, 20 who take a biology course
B, 20 who take a chemistry course C and 8 who take a dance course D. There are 5
students who take both A and B, 7 students who take both A and C, 4 students who take
both A and D, 16 students who take both B and C, 4 students who take both B and D
and 3 students who take who take both C and D. There are 3 who take A, B and C; 2
who take A, B and D; 3 who take A, C and D; and 2 who take B, C and D. Finally there
are 2 in all four courses and further 71 students who have not taken any of these courses.
Find the total number of students.
4. Find the number of nonnegative integer solutions of a + b + c + d = 27, where 1 a
5, 2 b 7, 3 c 9, 4 d 11.

5. Determine all integers n satisfying (n) = 13.

6. Determine all integers n satisfying (n) = 12.
7. For each fixed n N, use mathematical induction to prove that

(d) = n.

d|n

8. A function f : N N is said to be multiplicative if f (nm) = f (n)f (m), whenever

gcd(n, m) = 1.

(a) Let f, g : N N be functions satisfying f (n) =

g(d) and f (1) = g(1) = 1. If f is

d|n

multiplicative then use induction to show that g is also multiplicative.

(b) Imagine the fractions
P
(d).
that n =

1 2
n
n, n, . . . , n.

Cancel the common factors and regroup to show

d|n

(c) Conclude that is multiplicative.

1
9. Show that for n 1, Dn = n!
e + 2 .
n
P
10. Prove combinatorially:
C(n, i)Dni = n!.
i=0

11. Show that

m
P

(1)k C(m, k)(m k)n =

k=0

n! if m = n
0

if m > n.

12. Determine the number of 10-letter words using ENGLISH alphabets that does not contain
all the vowels.
13. Determine the number of ways to put
(a) 30 indistinguishable balls into 4 distinguishable boxes with at most 10 balls in each
box.
(b) 30 distinguishable balls into 10 distinguishable boxes such that at least 1 box is empty.
(c) r distinguishable balls into n distinguishable boxes such that at least 1 box is empty.
(d) r distinguishable balls into n distinguishable boxes so that no box is empty.

122

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

14. Determine the number of ways to arrange 10 digits 0, 1, . . . , 9, so that the digit i is never
followed immediately by i + 1.
15. Determine the number of strings of length 15 consisting of the 10 digits, 0, 1, . . . , 9, so that
no string contains all the 10 digits.
16. Determine the number of ways of permuting the 26 letters of the ENGLISH alphabets so
that none of the patterns lazy, run, show and pet occurs.
17. Let x be a positive integer less than or equal to 9999999.
(a) Find the number of xs for which the sum of the digits in x equals 30.
(b) How many of the solutions obtained in the first part consist of 7 digits?

7.3

Generating functions

This is one of the strongest tools in combinatorics. We start with the definition of formal power
series over Q and develop the theory of generating functions. This is then used to get closed
form expressions for some known recurrence relations and are then further used to get some
binomial identities.
Definition 7.3.1.
1. [Formal power series] An algebraic expression of the form f (x) =
P
n
an x , where an Q for all n 0, is called a formal power series in the indeterminate
n0

x over Q. By P(x), we denote" the set of all#formal power series in x and by cf[xn , f ], the
P
coefficient of xn in f , e.g., cf xn ,
an xn = an .
n0

2. [Equality of two formal power series] Two elements f, g P(x) are said to be equal if
cf[xn , f ] = cf[xn , g] for all n 0.

3. [Sum and Product in P(x)] Let f (x) =

an xn , g(x) =

n0

n0

bn xn P(x). Then, their

(a) sum/addition is defined by cf[xn , f + g] = cf[xn , f ] + cf[xn , g].

(b) product (called the Cauchy product) is defined by cf[xn , f g] = cn =
Before proceeding further, we consider the following examples.
Example 7.3.2.

n
P

ak bnk .

k=0

1. How many words of size 8 can be formed with 6 copies of A and 6 copies

of B?
Ans:

6
P

k=2

C(8, k), as we just need to choose k places for A, where 2 k 6.

Alternate. In any such word, we need m many As and n many Bs with m + n = 8,

8!
m 6 and n 6. Also, the number of words with m many As and n many Bs is
.
m!n!
8!xm y n
We identify this number with
and note that this is a term of degree 8 in
m!n!
h
y2 y3 y4 y5 y6 i
x2 x3 x4 x5 x6 ih
1+y+
.
+
+
+
+
+
+
+
+
8! 1 + x +
2!
3!
4!
5!
6!
2!
3!
4!
5!
6!

7.3. GENERATING FUNCTIONS

123

If we replace y by x, then our answer is

i
h
3
4
5
6
2
3
4
5
6
2
8!cf x8 , (1 + x + x2! + x3! + x4! + x5! + x6! )(1 + x + x2! + x3! + x4! + x5! + x6! )
h
i
2
3
4
5
6
2
3
4
5
6
=
8!cf x8 , ( x2! + x3! + x4! + x5! + x6! )( x2! + x3! + x4! + x5! + x6! )
i
h
3
2
3
2
=
8!cf x8 , ( x2! + x3! + )( x2! + x3! + )




8
= 8!cf x8 , (ex 1 x)2 = e2x + 1 + x2 2xex 2ex + 2x = 8! 28! 7!2 8!2 = 238.

2. How many anagrams are there of the word M ISSISSIP P I?

11!
. For another understanding, the readers
Ans: Using basic counting, the answer is
4!4!2!
should also note that




x2 x3 x4
2
x2

11!
= 11! cf x11 , 1 + x 1 + x +
+
+
1+x+
4!4!2!
2!
3!
4!
2!


2
4
2
5
3

x

2 x


x
x
x
11
= 11! cf x , x +
+
+
+
+
+
2!
4!
5!
2!
3!

x2
x4 x4
,
and
for the alphabets M, I, S and P , respectively.
4! 4!
2!
3. Prove that the number of nonnegative integer solutions of u + v + w + t = 10 equals


cf x10 , (1 + x + x2 + )4 .
as we need to have x,

Ans: Note that u can take any value from 0 to 10 which corresponds to 1 + x + + x10 .

Hence, using Theorem 6.6.1, the required answer is



4 5 13
cf x10 , f = (1 + x + x2 + )4 = (1 x)4 = C(13, 10) =
.
10!

Definition 7.3.3. [Generating functions] Let (br )

0 be a sequence of integers. Then, the
1. ordinary generating function (ogf) is the formal power series
b0 + b1 x + b2 x2 + b3 x3 + , and
2. exponential generating function (egf) is the formal power series
b0 + b1 x + b2

x2
x3
+ b3 + .
2!
3!

If there exists an M N such that br = 0 for all r M , then the generating functions have
finitely many terms.

Example 7.3.4. What is the number of nonnegative integer solutions of 2a + 3b + 5c = r,

r N0 ?

Ans: Note that a N0 and hence 2a corresponds to the formal power series 1 + x2 + x4 + .

Thus, we need to consider the ogf

(1 + x2 + x4 + )(1 + x3 + x6 + )(1 + x5 + x10 + ) =



Hence, the required answer is cf xr ,

(1

x2 )(1

1
.
x3 )(1 x5 )

(1

x2 )(1

1
.
x3 )(1 x5 )

124

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

P xn
xn
, g(x) =
bn
P(x). Then, in case of egf,
n!
n!
n0
n0
n
P
P
xn
n

dn , where dn =
their product equals
k ak bnk , for n 0.
n!
n0
k=0

Remark 7.3.5.

1. Let f (x) =

an

P (ex 1)n
P yn
x
implies that ee 1 =
and
n!
n0
n0 n!


m
x 1)n 
X (ex 1)n
X
 m ex 1 
(e
m
m
=
cf x , e
= cfx ,
cf x ,
.
n!
n!

2. Note that ee

x 1

P(x) as ey =

(7.2)

n=0

n0



x
That is, for each m 0, cf xm , ee 1 is a sum of a finite number of rational numbers.

x
x
Whereas, the expression ee 6 P(x) requires infinitely many computation for cf xm , ee ,
for all m 0.

With the algebraic operations as defined in Definition 1.3, it can be checked that P(x) forms a
Commutative Ring with identity, where the identity element is given by the formal power series
P
f (x) = 1. In this ring, the element f (x) =
an xn is said to have a reciprocal if there exists
n0
P
another element g(x) =
bn xn P(x) such that f (x)g(x) = 1. So, the question arises, under
n0

what conditions on cf[xn , f ], can we find g(x) P(x) such that f (x)g(x) = 1. The answer to

this question is given in the following proposition.



Proposition 7.3.6. The reciprocal of f P(x) exists if and only if cf x0 , f 6= 0.
P
P
Proof. Let g(x) =
bn xn P(x) be the reciprocal of f (x) =
an xn . Then, f (x)g(x) = 1 if
n0
n0


and only if cf x0 , f g = 1 and cf[xn , f g] = 0, for all n 1.




But, by definition of the Cauchy product, cf x0 , f g = a0 b0 . Hence, if a0 = cf x0 , f = 0


then cf x0 , f g = 0 and thus, f cannot have a reciprocal. However, if a0 6= 0, then the

coefficients cf[xn , g] = bn s can be recursively obtained as follows:

1
1
1
as 1 = c0 = a0 b0 ;
b1 =
(a1 b0 ) as 0 = c1 = a0 b1 + a1 b0 ;
b2 =
(a2 b0 + a1 b1 )
b0 =
a0
a0
a0
as 0 = c2 = a0 b2 + a1 b1 + a2 b0 ; and in general, if we have computed bk , for k r, then using
1
0 = cr+1 = ar+1 b0 + ar b1 + + a1 br + a0 br+1 ,
br+1 =
(ar+1 b0 + ar b1 + + a1 br ).
a0
Hence, the required result follows.
Note that, in Proposition 7.3.6, bn Q as a0 Q. We now look at the composition of formal
P
P
bn xn P(x) then the composition
an xn , g(x) =
power series. Recall that, if f (x) =
n0

n0

(f g)(x) = f (g(x)) =

an (g(x))n =

n0

an (

n0

bm xm )n

m0

may not be defined (just to compute the constant term of the composition, one may have to
look at an infinite sum of rational numbers). For example, let f (x) = ex and g(x) = x + 1. Note
that g(0) = 1 6= 0. Here, (f g)(x) = f (g(x)) = f (x + 1) = ex+1 . So, as function f g is well
P
ak xk P(x) (i.e., with ak Q)
defined, but there is no formal procedure to write ex+1 as
k0

and hence ex+1 is not a formal power series over Q. The next result gives the condition under
which the composition (f g)(x) is well defined.

7.3. GENERATING FUNCTIONS

125

Proposition 7.3.7. Let f, g P(x). Then, the composition (f g)(x) P(x) if either f is a




polynomial or cf x0 , g(x) = 0. Moreover, if cf x0 , f (x) = 0, then there exists g P(x), with


cf x0 , g(x) = 0, such that (f g)(x) = x. Furthermore, (g f )(x) P(x) and (g f )(x) = x.

P
cn xn and suppose that either f is a
Proof. As (f g)(x) P(x), let (f g)(x) = f (g(x)) =
n0




polynomial or cf x0 , g(x) = 0. Then, to compute ck = cf xk , (f g)(x) , for k 0, one just
k
P
P
needs to consider the terms
an xn . Hence, each ck Q and
ak (g(x))n , whenever f (x) =
n=0

n0

thus, (f g)(x) P(x). This completes the proof of the first part. We leave the proof of the

other part for the reader.

Proposition 7.3.8. [Basic tricks] Recall the following statements from Binomial theorem and
Theorem 6.6.1.


1. cf xn , (1 x)r = (1 + x + x2 + )r = C(n + r 1, n).

2. (1 xm )n = 1 C(n, 1)xm + C(n, 2)x2m + (1)n xnm .



1 xm n
2
m1
n
= (1 xm )n (1 + x + x2 + )n .
3. (1 + x + x + + x
) =
1x

We now define the formal differentiation in P(x) and give some important results. The proof
is left for the reader.
Definition 7.3.9. [Differentiation] Let f (x) =
ation of f (x), denoted f (x), is defined by

n0

an xn P(x). Then, the formal differenti-

f (x) = a1 + 2a2 x + + nan xn1 + =

nan xn1 .

n1

Proposition 7.3.10. [ogf: tricks] Let g(x), h(x) be the ogf s for the sequences (ar )
0 , (br )0 ,

respectively. Then, the following are true.

1. Ag(x) + Bh(x) is the ogf for (Aar + Bbr )
0 .
2. (1 x)g(x) is the ogf for the sequence a0 , a1 a0 , a2 a1 , .
3. (1+x+x2 + )g(x) = (1x)1 g(x) is the ogf for (Mr )
0 , where Mr = ar +ar1 + +a0 .
4. g(x)h(x) is the ogf for (cr )
0 , where cr = a0 br + a1 br1 + a2 br2 + + ar b0 .
5. xf (x) is the ogf for (rar )
1 .
Proof. For example, to prove (3), note that if g(x) = a0 + a1 x + a2 x2 + , then the coefficient

of x2 in (1 + x + x2 + )(a0 + a1 x + a2 x2 + ) is a2 + a1 + a0 .
Example 7.3.11.

1. Let ar = 1 for all r 0. Then, the ogf of the sequence (ar )

0 equals

1 + x + x2 + = (1 x)1 = f (x). So, for r 0, the ogf for

(a) ar = r is xf (x) and


(b) ar = r 2 is x f (x) + xf (x) .

(c) ar = 3r + 5r 2 is 3xf (x) + 5 xf (x) + x2 f (x) = 8x(1 x)2 + 10x2 (1 x)3 .

126

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

2. Determine the number of ways to distribute 50 coins among 30 students so that no student
gets more than 4 coins equals




cf x50 , (1 + x + x2 + x3 + x4 )30 = cf x50 , (1 x5 )30 (1 x)30

= C(79, 50) 30C(74, 45) + C(30, 2)C(69, 40) +

10
X
(1)i C(30, i)C(79 5i, 50 5i).
=
i=0

3. For n, r N, determine the number of solutions to y1 + +yn = r with yi N0 , 1 i n.

Ans: Recall that this number equals C(r + n 1, r) (see Theorem 6.3.3).
Alternate. We can think of the problem as follows: the above system can be interpreted
as coming from the monomial xr , where r = y1 + + yn . That is, the problem reduces
to finding the coefficients of xyk of a formal power series, for yk 0. Now, recall that


cf y xk , (1 y)1 = 1. Hence, the question reduces to computing




1
1
r
r
cf x ,
= C(r + n 1, r).
= cf y ,
(1 y)(1 y) (1 y)
(1 y)n

4. Evaluate

k=0

1
k.
2k

Put f (x) = (1 x)1 . Then, the required sum is

1
2 f (1/2)

= 2.

Alternately (rearranging terms of an absolutely convergent series) it is

1
2
1
4
1
8

+
+

1
4
1
8

1+

1
2

..
.

5. Determine a closed form expression for

+
+

Hence, (1 x)S =

= (1 x)1 =

xn , one has (1 x)2

nxn = x + 2x2 + 3x3 + . Then, xS = x2 + 2x3 + 3x4 + .

Thus, the closed form expression is

nxn P(x).

n0

Alternate. Let S =

+ = 2.

n0

Ans: As (1 x)1 =

1
8

n0
xk =

k1

k0

x
.
(1 x)2

xk =

n0

xn

x
x
. Thus, S =
.
1x
(1 x)2

nxn1 .

n0

6. Determine the sum of the first N positive integers.




Ans: Using previous example, note that k = cf xk1 , (1 x)2 . Therefore, by PropoN



P
k = cf xN 1 , (1 x)1 (1 x)2 and hence
sition 7.3.10, one has
k=1

N
X
k=1



N (N + 1)
.
k = cf xN 1 , (1 x)3 = C(N + 1, N 1) =
2

7.3. GENERATING FUNCTIONS

127

7. Determine the sum of the squares of the first N positive integers.





P
P
P 2 n
x(1 + x)
x
x
n
Ans: Recall
nxn = (1x)
.
nx
=
x
=
.
Thus,
n
x
=
x
2
(1x)2
(1 x)3
n0
n0
n0
Hence,
N
X

k=1


= cf xN ,






x(1 + x)
1
1
1
N 1
N 2

= cf x
,
+ cf x
,
1 x (1 x)3
(1 x)4
(1 x)4

N (N + 1)(2N + 1)
.
6
1. For n, r N, determine the number of solutions to x1 +2x2 + +nxn =

= C(N + 2, N 1) + C(N + 1, N 2) =
Exercise 7.3.12.

r with xi N0 , 1 i n.

P
1
2. Determine
C(n + k 1, k).
2k
k=0

3. Find the number of nonnegative integer solutions of a + b + c + d + e = 27, satisfying

(a) 3 a 8,
(b) 3 a, b, c, d 8
(c) c is a multiple of 3 and e is a multiple of 4.
4. Determine the number of ways in which 100 voters can cast their 100 votes for 10 candidates such that no candidate gets more than 20 votes.
5. Determine a closed form expression for

N
P

k3 .

k=1

6. Determine a closed form expression for

P n2 + n + 6
.
n!
n0

7. Verify the following table of formal power series.

ex =

xk

k!
P (1)r x2r
cos(x) =
(2r)!
r0
P x2r
cosh(x) =
r0 (2r)!
k0

1
1x
(1 + x)n
xr

1
1 4x

Table of Formal Power Series

P
C(n, k)xk , n N0
(1 + x)n =
r0

P (1)r x2r+1
r0 (2r + 1)!
P x2r+1
sinh(x) =
r0 (2r + 1)!
sin(x) =

Radius of convergence: |x| < 1

P xk
log(1 x) =
k1 k
P k
P
1
x
C(n + k 1, k)xk , n N
=
=
n
(1

x)
k0
k0
P
P
xn
C(n, r + k)xk
C(k, n)xk , n N0
=
=
n+1
(1

x)
kr
k0
1
Radius of convergence: |x| <
4

P
P 1
1

4x
=
C(2k, k)xk
=
C(2k, k)xk
2x
k0
k0 k + 1

128

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

(5, 3, 3, 2, 1, 1)

9 : I-hook
7 : II-hook

3 : III-hook

(5, 5, 4, 3, 2)
Figure 7.1: Ferrers diagram and its conjugate
(6, 4, 3, 1, 1)

Definition 7.3.13. [Ferrers diagram] For n, k N, let (n1 , n2 , , nk ) be a partition of n N

into k parts. Then, the Ferrers Diagram of (n1 , n2 , , nk ) is a pictorial representation

(pattern) using dots in the following way: place n1 dots in the first row. The n2 dots in the
second row are placed in such a way to cover the first n2 dots of the first row and so on (see
Figure 7.1).
Example 7.3.14.

1. (1, 1, 1, 1), (2, 2), (2, 1, 1) are a few partitions of 4.

2. Ferrers diagram for (5, 3, 3, 2, 1, 1) is

3. Let be a partition and its Ferrers diagram. Then, the diagram obtained by
interchanging the rows and columns of is called the conjugate of , denoted . Thus,
the conjugate of the partition (5, 3, 3, 2, 1, 1) is (6, 4, 3, 1, 1), another partition of 15.
Definition 7.3.15. [Self conjugate] A partition is said to be self conjugate if the Ferrers
diagram of and is the same.
Example 7.3.16. Find a one-one correspondence between self conjugate partitions and partitions of n into distinct odd terms.
Ans: Let be a self conjugate partition with k diagonal dots. For 1 i k, define ni =

number of dots in the i-th hook (dotted lines in Figure 7.1). Conversely, given any partition,

say (x1 , . . . , xk ) with odd terms, we can get a self conjugate partition by putting x1 dots in the
first hook, x2 dots in the second hook and so on. Since each xi is odd, the hook is symmetric
and xi xi1 + 2 for 2 i k implies that the corresponding diagram of dots is indeed a
Ferrers diagram and hence the result follows.

Theorem 7.3.17. [Euler: partition of n] The generating function for n is

(x) = (1 + x + x2 + )(1 + x2 + x4 + ) (1 + xn + x2n + ) =

1
.
(1 x)(1 x2 ) (1 xn )

Proof. Note that any partition of n has m1 copies of 1, m2 copies of 2 and so on till mn
n
P
copies of n, where mi N0 for 1 i n and
mi = n. Hence, uniquely corresponds to
i=1

(x1 )m1 (x2 )m2 (xn )mn in the word-expansion of

(1 + x + x2 + )(1 + x2 + x4 + ) (1 + xn + x2n + ).
Thus, n = cf[xn , (x)].

7.3. GENERATING FUNCTIONS

129

Example 7.3.18. Let f (n) be the number of partitions of n in which no part is 1. Then, note
that the ogf for f (n) is (1 x)(x). Hence, f (n) = n n1 .
Alternate. Let = (n1 , . . . , nk ) be a partition of n with nk = 1. Then, gives a partition
of n 1, namely (n1 , . . . , nk1 ). Conversely, if = (t1 , . . . , tk ) is a partition of n 1, then
(t1 , . . . , tk , 1) is a partition of n with last part 1, Hence, the required result follows.

The next result is the same idea as Theorem 7.3.17 and hence the proof is omitted.

r
Q
Theorem 7.3.19. The number of partitions of n with entries at most r is cf xn ,

i=1

1
1xi

Theorem 7.3.20. [ogf of n (r)] Fix n, r N. Then, the ogf for n (r), the number of partitions

of n into r parts, is

xr
.
(1x)(1x2 )(1xr )

Proof. Let be a partition of n into at most r parts. Then, corresponds to a partition of

n with entries at most r. Now, add a column of dots of height r on the left of the Ferrers
diagram of . Then, the new Ferrers diagram corresponds to a partition of n + r into r parts.
Conversely, given a partition of n + r into r parts, the inverse map gives a partition of n into at
most r parts. Thus, by Theorem 7.3.19, we get


1
nr
n (r) = cf x ,
.
(1 x)(1 x2 ) (1 xr )

xr
.
(1 x)(1 x2 ) (1 xr )
1. For n, r N, prove that n (r) is the number of partitions of n + C(r, 2)

Hence, the ogf for n (r) is

Exercise 7.3.21.

into r unequal parts.

2. Let P, M N and f (n) be the number of partitions of n where parts are from P and
multiplicities are from M . Find the generating function for the numbers f (n).

Theorem 7.3.22. Suppose there are k types of objects.

1. If there is an unlimited supply of each object, then the egf of the number of r-permutations
is ekx .
2. If there are mi copies of i-th object, then the egf of the number of r-permutations is
 


xm1
xmk
x2
x2
+ +
+ +
1+x+
1 + x +
.
2!
m1 !
2!
mk !
3. Moreover, n!S(r, n) is the coefficient of

xr
r!

in (ex 1)n .

Proof. Part 1: Since there are unlimited supply of each object, the egf for each object corresponds
xn
+ . Hence, the required result follows.
to ex = 1 + x + +
n!
Part 2: Argument is similar to that of Part 1 and is omitted.
Part 3: Recall that n!S(r, n) is the number of surjections from [r] to S = {s1 , , sn }. Each

surjection can be viewed as word of length r of elements of S, with each si appearing at least
n
P
once. Thus, we need a selection of ki N copies of si , with
ki = r. Also, by Theorem 6.1.26,
i=1

this number equals C(r; k1 , , kn ). Hence,

n 
 r

 
x
x2 x3
x
n
r
+
+
= cf
, (e 1) .
n!S(r, n) = r!cf x , x +
2!
3!
r!

130

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

Example 7.3.23.

1. In how many ways can you get Rs 2007 using denominations 1, 10, 100, 1000

only?


Ans: cf x

2007


1
,
.
(1 x)(1 x10 )(1 x100 )(1 x1000 )

2. If we use at most 9 of each denomination in (a), then this number is

"

cf x

2007

9
X
i=1

9
X

10i

i=1

9
X

100i

i=1

9
X

1000i

i=1

!#

= cf x

2007


1 x10000
= 1.
,
1x

3. Every natural number has a unique base-r representation (r 2). Note that Item (2)
corresponds to the case r = 10.

4. Consider n integers k1 < k2 < < kn with gcd(k1 , . . . , kn ) = 1. Then, the number of

natural numbers not having a partition using {k1 , . . . , kn } is finite. Since gcd(k1 , . . . , kn ) =
P
1, there exist i Z such that
i ki = 1. Let m = max{|1 |, . . . , |n |}, k = min{ki } and
P
N = km(k1 + +kn ). Notice that N, N +k, N +2k, . . . can be represented as i ki where
P
P
i km. For 1 r < k, we have N + r = km(k1 + + kn ) + r i ki = (km ri )ki .

Thus, each integer greater than N can be represented using k1 , . . . , kn . Determining the
largest such integer (Frobenius number) is the coin problem/ money changing
problem. The general problem is NP-hard. No closed form formula is known for n > 3.
Notice!
Some times we have a way to obtain a recurrence relation from the generating function.
This is important and hence study the next example carefully.
1
= a0 + a1 x + +
(1 x)(1
x100 )(1 x1000 )
n
an x + . Then, taking log and differentiating, we get

Example 7.3.24.

1. Suppose F =

F = F

x10 )(1


1
10x9
100x99
1000x999
+
+
+
.
1 x 1 x10 1 x100 1 x1000

So,


 n1 
n1
nan = cf x
, F = cf x
,F

10x9
100x99
1000x999
1
+
+
+
1 x 1 x10 1 x100 1 x1000



n
X

ank bk ,

k=1

where

bk = cf xk1 ,

10x9

100x99

1
+
+
1 x 1 x10 1 x100

2. We know that lim

n
P

n k=1

1
k



999
11
1000x
+
=
1000
111
1x

1111
n
P

1
,
n k=1 pk

= . What about lim

if 10 k
if 10|k, 100 k
if 10|k, 100|k, 1000 k
else.

where pk is the k-th prime?

7.3. GENERATING FUNCTIONS

Ans: For n > 1, let sn =

n
P

k=1

sn

1 1
1 + + +
2 4



131

1
k.

Then, note that

1 1
1 + + +
3 9

1
1
1 +
+ 2 +
pn pn

n
Y

(1 +

k=1

1
).
pk 1

Thus,
n
Y

log sn log

!
n
n
n1
X
X
X 1
1
1
1
(1 +
)
)
1+
.
log(1 +
pk 1
pk 1
pk 1
pk

k=1

k=1

n
P

1
n i pi

As sn , we see that lim

k=1

k=1

= as lim log sn = .
n

3. Let S be the set of natural numbers with only prime divisors 2, 3, 5, 7. Then,
X1
1 1
1 1
1
1
2357
1+
= (1 + + + )(1 + + + ) (1 + +
+ ) =
.
n
2 4
3 9
7 49
1246
nS

Exercise 7.3.25.

1. Let (n) =

d|n

d, for n N. Then, prove that nn =

n
P

nk (k).

k=1

2. A Durfee square is the largest square in a Ferrers diagram. Find the generating function
for the number of self conjugate partitions of n with a fixed size k of Durfee square. Hence,
2

X
xk
3
.
show that (1 + x)(1 + x ) = 1 +
(1 x2 )(1 x4 ) (1 x2k )
k=1

3. Show that the number of partitions of n into distinct terms ( each term is distinct) is the
same as the number of partitions of n into odd terms (each term is odd).
4. Find the number of r-digit binary numbers that can be formed using an even number of
0s and an even number of 1s.
5. Find the egf of the number of words of size r using A, B, C, D, E, if the word has
(a) all the letters and the letter A appears an even many times.
(b) all the letters and the first letter of the word appears an even number of times.

6. A permutation of [n] is said to be connected if there does not exist k, 1 k < n such
that takes [k] to itself. Let cn denote the number of connected permutations of [n] (put
c0 = 0), then show that

n
X
k=1

ck (n k)! = n!.

Hence, derive the relationship between the generating functions of (n!) and (cn ).
7. Let f (n, r) be the number of partitions of n where each part repeats less than r times.
Let g(n, r) be the number of partition of n where no part is divisible by r. Show that
f (n, r) = g(n, r).
8. Find the number of 9-sequences that can be formed using 0, 1, 2, 3 in each case.
(a) The sequence has an even number of 0s.
(b) The sequence has an odd number of 1s and an even number of 0s.
(c) No digit appears exactly twice.

132

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

7.4

Recurrence relation

Definition 7.4.1. [Recurrence relation] A recurrence relation is a way of recursively defining

the terms of a sequence as a function of preceding terms together with certain initial conditions.
Example 7.4.2. an = 3 + 2an1 for n 1 with the initial condition a0 = 1 is a recurrence

relation. Note that it completely determines the sequence (an ) = {1, 5, 13, 29, 61, . . .}.

Definition 7.4.3. [Difference equation] For a sequence (an ), the first difference d(an ) is
an an1 . The k-th difference dk (an ) = dk1 (an ) dk1 (an1 ). A difference equation is

an equation involving an and its differences.

Example 7.4.4.

1. an d2 (an ) = 5 is a difference equation. But, note that it doesnt give

a recurrence relation as we dont have any initial condition(s).

2. Every recurrence relation can be expressed as a difference equation. The difference equation corresponding to the recurrence relation an = 3 + 2an1 is an = 3 + 2(an d(an )).
Definition 7.4.5. [Solution] A solution of a recurrence relation is a function f (n) satisfying
the recurrence relation.
Example 7.4.6.

1. f (n) = 2n+2 3 is a solution of an = 3 + 2an1 with a0 = 1.

2. The Fibonacci sequence is given by an = an1 + an2 for n 2 with a0 = 1, a1 = 1.

Definition 7.4.7. [LNHRRCC/LHRRCC] A recurrence relation is a linear nonhomogeneous
recurrence relation with constant coefficients (LNHRRCC) of order r if, for a known function f
an = c1 an1 + + cr anr + f (n), where ci R for 1 i r, cr 6= 0.

(7.3)

If f = 0, then Equation (7.3) is homogeneous and is called the associated linear homogeneous
recurrence relation with constant coefficients (LHRRCC).
Theorem 7.4.8. For k N, let fi , 1 i k be known functions. Consider the k LNHRRCC
an = c1 an1 + + cr anr + fi (n) for i = 1, . . . , k,

(7.4)

with the same set of initial conditions. If gi is a solution of the i-th recurrence then,
an = c1 an1 + + cr anr +
under the same set of initial conditions has

k
P

k
X

i fi (n)

(7.5)

i=1

i gi (n) as its solution.

i=1

Proof. The proof is left as an exercise for the reader.

Definition 7.4.9. [Characteristic equation] Consider a LHRRCC an = c1 an1 + + cr anr
with cr 6= 0. If an = xn is a solution, then either x = 0 or x is a root of
xr c1 xr1 cr = 0.

(7.6)

7.4. RECURRENCE RELATION

133

Equation (7.6) is called the characteristic equation of the given LHRRCC. If x1 , . . . , xr are
r
P
the roots of Equation (7.6), then an = xni (and hence an =
i xni for i R) is a solution of
i=1

the given LHRRCC.

Theorem 7.4.10. [General solution: distinct roots] If the roots xi , i = 0, . . . , r 1 of Equation (7.6) are distinct, then every solution h(n) is a linear combination of xni . Moreover, the

solution is unique if we are given r consecutive initial conditions.

Proof. Let h(n) be any solution. Then, note that there exists 0 , . . . , r1 , such that

1
h(0)
x0 xr1 0

..
..
. ,

=
..
.

.
h(r 1)
r1
r1
r1
x0
xr1
as the r r matrix is an invertible matrix. That is, for every i R, h(n) =

r1
P
i=0

i xni , 0 n

r 1. Hence, we have proved the result for the first r values of h(n). So, let us assume that the

result is true for n < k. Then, by definition

h(k) =

r
X
j=1

cj h(k j) =

r
X
j=1

cj

r1
X

i xikj

i=0

r1
X

i=0

r
X

cj xikj

r1
X

i xki ,

i=0

j=1

as for n = k, xki is a solution of Equation (7.6). Thus, by PMI, h(n) =

r1
P
i=0

i xni for all n. The

uniqueness is left as an exercise for the reader.

Example 7.4.11.
1. Solve an 4an2 = 0 for n 2 with a0 = 1 and a1 = 1.
Ans: Note that 2 are the roots of the characteristic equation, x2 4 = 0. As the roots
are distinct, the general solution is an = (2)n + 2n for , R. The initial conditions

give + = 1 and 2 2 = 1. Hence, =

an = 2n2 3 + (1)n .

1
4,

3
4.

Thus, the unique solutions is

2. Solve an = 3an1 + 4an2 for n 2 with a0 = 1 and a1 = c, a constant.

Ans: Note that 1 and 4 are the roots of the characteristic equation, x2 3x 4 = 0.

As the roots are distinct, the general solution is an = (1)n + 4n for , R. Now,

1+c
the initial conditions imply = 4c
5 and = 5 . Thus, the unique general solution is
(1 + c)4n
(4 c)(1)n
+
, if c 6= 4.
(a) an =
5
5
(b) an = 4n , if c = 4.

3. Solve the Fibonacci recurrence an = an1 + an2 with initial conditions a0 = a1 = 1.

Ans: In this case, note that the roots of the characteristic equation, x2 x 1 = 0, are

n
1 5
1+ 5 n
+ 12 5 for
2 . As the roots are distinct, the general solution is an =
2

, R. Now, using the initial conditions, we get =

5+ 5
10 ,

required solution is

!n+1




1+ 5 n
1 5 n
1 1+ 5

an =
+
=
2
2
2
5

5 5
10 .

Hence, the

!n+1
1 5
.
2

134

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

Theorem 7.4.12. [General solution: multiple roots] Let t is a root of Equation (7.6) of
multiplicity s. Then, u(n) = tn (1 + n2 + + ns1 s ) is a solution (basic solution). In

general, if ti is a root of Equation (7.6) with multiplicity si ), for i = 1, . . . , k, then every solution
is a sum of the k basic solutions.

Proof. It is given that t is a zero of the polynomial F = xr c1 xr1 cr of multiplicity s.

Put G0 = xnr F = xn c1 xn1 cr xnr and G1 = xG0 , G2 = xG1 , . . ., Gs1 = xGs2 .

Then, each of G0 , G1 , . . . , Gs1 has a zero at t. That is, for i = 0, 1, . . . , s 1, we have
Gi (t) = tn ni c1 tn1 (n 1)i . . . cr tnr (n r)i = 0.
Now, take u(n) = tn P (n), where P (n) =

i=0

0 i s 1. Then,
s1
X
i=0

s1
P

ni i is a fixed polynomial, with i R for

i Gi (t) = tn P (n) c1 tn1 P (n 1) cr tnr P (n r) = 0.

Hence, for 0 i s 1 and i R, u(n) is a solution of the LHRRCC. The other part of the
proof is left for the reader.

Example 7.4.13. Suppose that a LHRRCC has roots 2, 2, 3, 3, 3. Then, the general solution is
given by 2n (1 + n2 ) + 3n (1 + n2 + n2 3 ).
Theorem 7.4.14. [LNHRRCC] Consider the LNHRRCC in Equation (7.3) and let un be a
general solution to the associated LHRRCC. If vn is a particular solution of the LNHRRCC,
then an = un + vn is a general solution of the LNHRRCC.
Proof. The proof is left for the reader.
Notice!
No general algorithm are there to solve a LNHRRCC. If f (n) = an or nk or a linear
combination of these, then a particular solution can be obtained easily.
Obtaining particular solution after knowledge of the characteristic roots.

1. If f (n) = an and a is not a root of Equation (7.3), then vn = can .

2. If f (n) = an and a is a root of Equation (7.3) of multiplicity t, then vn = cnt an .
3. If f (n) = nk and 1 is not a root of Equation (7.3), then use vn = c0 + c1 n + + ck nk .
4. If f (n) = nk and 1 is a root of Equation (7.3) of multiplicity t, then vn = nt (c0 +
c1 n + + ck nk ).
Example 7.4.15.

1. Let an = 3an1 + 2n for n 1 with a0 = 1.

Ans: Observe that 3 is the characteristic root of the associated LHRRCC (an = 3an1 ).
Thus, the general solution of LHRRCC is un = 3n . Note that 1 is not a characteristic

7.5. GENERATING FUNCTION FROM RECURRENCE RELATION

135

root and hence a particular solution is a + nb, where a and b are to be computed using
3
n
a + nb = 3(a + (n 1)b) + 2n. This gives a = 3
2 and b = 1. Hence, an = 3 n 2 .
5
Using a0 = 1, check that = .
2
2. Let an = 3an1 2an2 + 3(5)n for n 3 with a1 = 1, a2 = 2.
Ans: Observe that 1 and 2 are the characteristic roots of the associated LHRRCC (an =
3an1 2an2 ). Thus, the general solution of the LHRRCC is un = 1n + 2n . Note that

5 is not a characteristic root and thus, vn = c5n is a particular solution of LNHRRCC if

and only if c5n = 3c5n1 2c5n2 + 3(5)n . That is, if and only if c = 25/4. Hence, the
general solution of LNHRRCC equals an = + 2n + (25/4)5n , where compute and
using the initial conditions.

3. In the above take f (n) = 3(2n ). Then, we see that with c(2)n as a choice for a particular
solution, we will have 4c = 6c 2c + 12, an absurd statement. But, with the choice cn(2)n ,
we have 4nc = 6(n 1)c 2(n 2)c + 12, implying c = 6. Hence, the general solution of

LNHRRCC is an = + 2n + 6n2n , where compute and using the initial conditions.

7.5

Generating function from recurrence relation

Sometimes we can find a solution to the recurrence relation using the generating function of an .
Example 7.5.1.

1. Consider an = 2an1 + 1, a0 = 1.

Ans: Let F (x) = a0 + a1 x + be the generating function for {ai }. Then,

F =1+

ai x i = 1 +

1
(1x)(12x)

2
12x

i=0

i=0

i=1

i=1

Hence, F =

X
X
X
ai x i =
xi + 2x
(2ai1 + 1)xi =

1
1x .

1
+ 2xF.
1x

Thus, an = cf[xn , F ] = 2n+1 1.

2. Find the ogf F for the Fibonacci recurrence relation an = an1 + an2 , a0 = 0, a1 = 1.
Ans: We have
F =

X
i=0

ai x = x +

ai2 x +

i=2

ai1 x = x + x

i=2

ai x + x

ai xi = x + (x2 + x)F.

i=1

i=0

x
1 +
x
Thus, F =
, where =
=
2
1xx
(x )(x )
2

1
,=
2

. So,





x
1

xi
1 X xi

F =
=

.
=
(x )(x )
5 x x
5 i=0 i i
Hence, using = 1, an =

cf[xn , F ]

(1 +
(1)n
= ( n n ) =
5

5)n (1

2n 5

5)n

The next result follows using a small calculation and hence the proof is left for the reader.

136

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

Theorem 7.5.2. [Obtaining generating function from recurrence relation] The generating
function of the r-th order LHRRCC an = c1 an1 + + cr anr with initial conditions ai = Ai ,
i = 0, 1, . . . , r 1 is
r1
P
i=0

Ai xi c1 x

r2
P
i=0

Ai xi c2 x2

r3
P
i=0

Ai xi cr1 xr1 A0

1 c1 x cr xr
Example 7.5.3.
1. Find the ogf for the Catalan numbers Cn s.
P
Ans: Let g(x) = 1 +
Cn xn , where Cn = C(2n,n)
n+1 with C0 = 1. Then,

n1

g(x) 1 =
=

Cn xn =

n1

n1

X
4n + 4

n=1

2n! n X 2(2n 1)
1

x =
Cn1 xn
n + 1 n!n!
n
+
1
n=1

X
6
6
Cn1 x +
Cn1 xn = 4xg(x) +
tg(t)dt.
n+1
n+1
x
n

n=1

Rx
So, [g(x) 1 4xg(x)]x = 6 tg(t)dt. Now, we differentiate with respect to x to get
0

g x(1 4x) + g(1 2x) = 1. To solve the ode, we first observe that



Z 
Z
1
2
x
1 2x
=
+
= ln
.
x(1 4x)
x 1 4x
1 4x
Thus, the integrating factor of the given ode is

x
14x

and hence the ode can be re-written

as
g(x)


1
1
d 
x
x
1 2x

=
=

.
g(x)
+ g(x)
3/2
3/2
dx
(1 4x)
(1 4x)
(1 4x)3/2
1 4x
1 4x

x
Hence, g(x) 14x
=

1
2 14x

+ C, where C R. Or, equivalently

1 + 2C 1 4x
.
g(x) =
2x

(7.7)

Note that C0 = lim g(x) = 1 and hence, C = 21 . Thus,

x0

g(x) =

1 4x
.
2x

Alternate. Recall that Cn is the number of representations of the product of n + 1 square

matrices of the same size, using n pairs of brackets. From such a representation, remove
the leftmost and the rightmost brackets to obtain the product of two representations of
the form:
A1 (A2 An+1 ), (A1 A2 )(A3 An+1 ), , (A1 Ak )(Ak+1 An+1 ), , (A1 An )An+1 .
Hence, we see that
Cn = C0 Cn1 + C1 Cn2 + + Cn1 C0 .

(7.8)

7.5. GENERATING FUNCTION FROM RECURRENCE RELATION

Thus, if we define g(x) =

n=0

Cn xn , then for n 1,

X

cf xn1 , g(x)2 = cfxn1 ,
Cn xn

137

n=0

!2 n1
X
=
Ci Cn1i = Cn using Equation (7.8).
i=0


That is, cf xn , xg(x)2 = Cn . Hence, g(x) = 1 + xg(x)2 . Solving for g(x), we get
!
r

1 1 4x
4
1
1 1
=

.
g(x) =
2 x
x2 x
2x

As the function g is continuous (being a power series in the domain of convergence) and
lim g(x) = C0 = 1, it follows that

x0

g(x) =

1 4x
.
2x

2. Fix r N and let (an ) be a sequence with a0 = 1 and

n
P

ak ank = C(n + r, r), for all

k=0

n 1. Determine an .
P
Ans: Let g(x) =
an xn . Then, note that C(n + r, r) = c(n + (r + 1) 1, n). Hence,
n0

g(x) =

n
X X

n0

ak ank

k=0

xn =

C(n + r, r)xn =

n0

C(n + r, n)xn =

n0

1
.
(1 x)r+1

h
i
1
Hence, an = cf xn , (1x)(r+1)/2
. For example, for r = 2,

3 5 7 (2n + 1)
(2n + 1)!
= 2n
.
n
2 n!
2 n!n!

an = (1)n C(3/2, n) =

3. Determine the sequence {f (n, m) | n, m N0 } which satisfies f (n, 0) = 1 for all n 0,

f (0, m) = 0 for all m > 0 and

f (n, m) = f (n 1, m) + f (n 1, m 1) for (n, m) 6= (0, 0).

Ans: Define Fn (x) =

m0

Fn (x) =

m0

m0

f (n, m)xm . Then, for n 1, Equation (7.9) gives

f (n, m)xm =

m0
m

f (n 1, m)x +

(f (n 1, m) + f (n 1, m 1)) xm
X

m0

f (n 1, m 1)xm

= Fn1 (x) + xFn1 (x) = (1 + x)Fn1 (x) = = (1 + x)n F0 (x).

Now, using the initial conditions, F0 (x) = 1 and hence Fn (x) = (1 + x)n . Thus,
(
C(n, m) if 0 m n
m
n
f (n, m) = cf[x , (1 + x) ] =
0
if m > n.

(7.9)

138

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

Alternate. Define Gm (y) =

n0

Gm (y) =
=

f (n, m)y n . Then, for m 1, Equation (7.9) gives

f (n, m)y n =

n0

n0

n0

f (n 1, m)y +

(f (n 1, m) + f (n 1, m 1)) y n
X

n0

f (n 1, m 1)y n

= yGm (y) + yGm1 (y).

y
1
Gm1 (y). Using initial conditions, G0 (y) =
. Hence,
1y
1y

Therefore, Gm (y) =
Gm (y) =

ym
. Thus,
(1 y)m+1


f (n, m) = cf y n ,



 (
C(n, m)
ym
1
= cf y nm ,
=
m+1
m+1
(1 y)
(1 y)
0

if 0 m n

if m > n.

4. Determine the sequence {S(n, m) | n, m N0 } which satisfy S(0, 0) = 1, S(n, m) = 0 if

either m = 0 or n = 0 but not both and

S(n, m) = mS(n 1, m) + S(n 1, m 1), (n, m) 6= (0, 0).

Ans: Define Gm (y) =

n0

Gm (y) =

S(n, m)y n . Then, for m 1, Equation (7.10) gives

S(n, m)y n =

n0

= m

(7.10)

n0

n0

(mS(n 1, m) + S(n 1, m 1)) y n

S(n 1, m)y n +

n0

S(n 1, m 1)y n

= myGm (y) + yGm1 (y).

Therefore, Gm (y) =

y
Gm1 (y). Using initial conditions, G0 (y) = 1 and hence
1 my
m

X k
ym
Gm (y) =
= ym
,
(1 y)(1 2y) (1 my)
1 ky

(7.11)

k=1

(1)mk km
, for 1 k m. Thus,
k! (m k)!
"
#


m
m
X
X
k
k
n m
nm
S(n, m) = cf y , y
=
cf y
,
1 ky
1 ky

where k =

k=1

m
X

k knm =

k=1

k=1

m
X
(1)mk kn
k=1

k! (m k)!

k=1

Therefore, S(n, m) =

1 X
1 X
(1)mk kn C(m, k) =
(1)k (m k)n C(m, k).
m!
m!

m
1 P
(1)k (m k)n C(m, k).
m! k=1

k=1

This identity is generally known as the Stirlings Identity.

(7.12)

7.5. GENERATING FUNCTION FROM RECURRENCE RELATION

139

Observation.
P

(a) Let us consider Hn (x) =

m0

S(n, m)xm . Then, verify that Hn (x) = (x + xD)n 1 as

H0 (x) = 1. Therefore, H1 (x) = x, H2 (x) = x + x2 , . Thus, we dont have a single

expression for Hn (x) which gives the value of S(n, m)s. But, it helps in showing that
S(n, m), for fixed n N, first increase and then decrease (commonly called unimodal).
The same holds for the sequence of binomial coefficients {C(n, m), m = 0, 1, . . . , n}.
(b) As there is no restriction on n.m N0 , Equation (7.12) is also valid for n < m. But,
we know that S(n, m) = 0, whenever n < m. Hence, we get the following identity,
m
X
(1)mk kn1
= 0 whenever n < m.
(k 1)! (m k)!
k=1

5. Bell Numbers For n N, the n-th Bell number, denoted b(n), is the number of partitions
n
P
of [n]. Thus, b(n) =
S(n, m), for n 1 and b(0) = 1. Hence, for n 1,
m=1

b(n) =

n
X

S(n, m) =

k1

mk

m=1

m1

m
XX
(1)mk kn1
S(n, m) =
(k 1)! (m k)!
m1 k=1

X kn X (1)mk
1 X kn
1 X kn
=
=
=
as 0n = 0 for n 6= 0. (7.13)
k!
(m k)!
e
k!
e
k!
k1

k0

kn
Thus, Equation (7.13) is valid even for n = 0. As b(n) has terms of the form
, we
k!
n
P
x
then,
b(n)
compute its egf. Thus, if B(x) =
n!
n0
B(x) = 1 +

n1

X 1 X k n xn
xn

=1+
b(n)
n!
e
k! n!
n1

k1

xn

1 X 1 X (kx)n
1X 1 X n
k
=1+
e
k!
n!
e
k!
n!
n1
n1
k1
k1



1 X 1  kx
1 X (ex )k
1
= 1+
e 1 = 1+

e
k!
e
k!
k!
= 1+

k1

Recall that ee

x 1

k1

x
1 ex
e 1 (e 1) = ee 1 .
= 1+
e

(7.14)

is a valid formal power series (see Remark 7.3.5).Taking logarithm of

Equation (7.14), we get log B(x) = ex 1. Hence, B (x) = ex B(x), or equivalently

x

X b(n)xn1

n1

(n 1)!

= xex

n0

n
X xm
X
x

b(n)
= x
b(n) .
n!
m!
n!
xn

m0

n0

140

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

Thus,
b(n)
(n 1)!

n
X b(n)xn
X xm X
x
= cfxn1 ,
= cfxn ,

b(n)
(n 1)!
m!
n!

n1
X

m=0

Hence, we get b(n) =

n1
P

m=0

Exercise 7.5.4.

n1

n0

m0

b(m)
1

.
(n 1 m)! m!

C(n 1, m)b(m), for n 1, with b(0) = 1.

1. Find the number of binary words without having a subword 00 and 111.

2. Find the number of subsets of {1, . . . , n} not containing consecutive integers.

3. Prove that Fn divides Fnm where n, m are positive integers.
Objects

Places

Places

distinct?

distinct?

nonempty?

Onto functions

All functions

Relate

Number
r!S(n, r) =
r1
P
(1)i C(r, i)(r i)n
i=0

rn

r-partition of a

S(n, r)

set
All partitions of

b(n) =

a set

C(n 1, r 1)

solutions
Nonnegative
integer solutions

Partitions of n

S(n, i)

i=1

Positive integer

r-partition of n

r
P

C(n + r 1, r 1)
n (r) =
h
i
cf xnr , (1x)(1x12 )(1xr )
r
P

of length r

n (i)

i=1

1. Find the number of circular permutations of {A, A, B, B, C, C, C, C}.

X
P
15!
2. Let S = {(n1 , n2 , n3 ) | ni N, ni = 15}. Evaluate
.
n1 !n2 !n3 !

Exercise 7.5.5.

(n1 ,n2 ,n3 )S

3. Each of the 9 senior students said: the number of junior students I want to help is exactly
one. There were 4 junior students a, b, c, d, who wanted their help. The allocation was
done randomly. What is the probability that either a has exactly two seniors to help him
or b has exactly 3 seniors to help him or c has no seniors to help him?
4. In a particular semester 6 students took admission in our PhD programme. There were
9 professors who were willing to supervise these students. As a rule a student can have

7.5. GENERATING FUNCTION FROM RECURRENCE RELATION

141

either one or two supervisors. In how many ways can we allocate supervisors to these
students if all the willing professors are to be allocated? What if we have an additional
condition that exactly one supervisor gets to supervise two students?
5. How many lattice paths are there from (0, 0) to (9, 9) which does not cross the dotted line?
(9, 9)

(0, 0)
6. (a) Prove combinatorially that, for n 2, we have Dn = (n 1)(Dn1 + Dn2 ).
ex
.
(b) Use Part (a) to show that the exponential generating function of Dn is
1x
7. My friend says that he has n 2 subsets of [14] each of which has size 6. Give a value of
n so that we can guarantee some two of his subsets have 3 elements in common, without

seeing his collection ? What is the smallest possible value of n?

8. Find the number of words of size 12 made using letters from {A, B, C} in which BCA
does not appear (as a consecutive subword). For example: ABCABCCCCCBA has an
appearance of BCA but BCCABCCABCCA does not.
9. Find the number of 8 letter words made using alphabets from {A, B, C, D} in which 3

consecutive letters are not allowed to be the same.

i8
i1 P
i2
9 P
P
P
10. Evaluate
i29 .

i1 =1 i2 =1 i3 =1

i9 =1

11. We have 3 blue bags, 4 red bags and 5 green bags. We have many balls of each of the colors
blue, red and green. Fill in the blank with the smallest positive integer.
If we distribute

balls (without seeing the colors) into these bags, then one of the

following must happen:

(a) a blue bag contains 3 blue balls or 4 red balls or 5 green balls
(b) a red bag contains 3 blue balls or 5 red balls or 7 green balls
(c) a green bag contains 3 blue balls or 6 red balls or 9 green balls
12. We have an integer polynomial f (x). Fill in the blank with the smallest positive integer.
If f (x) = 2009 has

many distinct integer roots, then f (x) = 9002 cannot have an

integer root.
13. In how many ways can one distribute
(a) 10 identical chocolates among 10 students?

142

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

(b) 10 distinct chocolates among 10 students?
(c) 10 distinct chocolates among 10 students so that each receives one?
(d) 15 distinct chocolates among 10 students so that each receives at least one?
(e) 10 out of 15 distinct chocolates among 10 students so that each receives one?
(f ) 15 distinct chocolates among 10 students so that each receives at most three?
(g) 15 distinct chocolates among 10 students so that each receives at least one and at
most three?
(h) 15 identical chocolates among 10 students so that each receives at most three?

14. In how many ways can one carry

(a) 15 distinct objects with 10 identical bags? Answer using S(n, r).
(b) 15 distinct objects in 10 identical bags with no empty bag? Answer using S(n, r).
(c) 15 distinct objects in 10 identical bags with each bag containing at most three objects?
(d) 15 identical objects in 10 identical bags?
(e) 15 identical objects in 10 identical bags with no empty bag?
(f ) 15 identical objects in 20 identical bags with no empty bag?
15. What is the number of integer solutions of x + y + z = 10, with x 1, y 2 and
z 3?

16. Is the number of solutions of x+y+z = 10 in nonnegative multiples of

1
2

(x, y, z are allowed

to be 0, 1/2, 1, 3/2, . . .) at most four times the number of nonnegative integer solutions of
x + y + z = 10?
17. How many words of length 8 can be formed using the English alphabets, where each letter
can appear at most twice? Give answer using generating function.
18. Let p1 , . . . , pn , n 2 be distinct prime numbers. Consider the set {p1 , . . . , pn , p21 , . . . , p2n }.

In how many ways can we partition the set into subsets of size two such that no prime is

in the same subset containing its square?

15
P
19. What is the value of
(1)k C(15, k)(15 k)5 ?
k=0

20. What is the number of partitions

(a) of n with entries at most r? Give your answers using generating function.
(b) of n with most r parts? Give your answers using generating function.
(c) n (r) of n with exactly r parts? Give your answers using generating function.
(d) n (r) of n + C(r, 2) with r distinct parts? Give your answers using generating function.
(e) of n with distinct entries? Give your answers using generating function.
(f ) of n with entries odd? Give your answers using generating function.
(g) of n with distinct odd entries? Give your answers using generating function.
(h) of n which are self conjugate? Give your answers using generating function.

7.5. GENERATING FUNCTION FROM RECURRENCE RELATION

143

21. How many words of length 15 are there using the letters A,B,C,D,E such that each letter
must appear in the word and A appears an even number of times? Give your answers
using generating function.
22. The characteristic roots of a LHRRCC are 2, 2, 2, 3, 3. What is the form of the general
solution?
23. Consider the LNHRRCC an = c1 an1 + + cr anr + 5n . Give a particular solution.
24. Obtain the ogf for an , where an = 2an1 an2 + 2n , a0 = 0, a1 = 1.
25. Solve the recurrence relation an = 2an1 an2 + 2n + 5, a0 = 0, a1 = 1.
26. My class has n CSE, m MSC and r MC students. Suppose that t copies of the same book
are to be distributed so that each branch gets at least s. In how many ways can this be
done, if each student gets at most one? In how many ways can this be done, without the
previous restriction? Answer only using generating function.
Exercise 7.5.6.

1. My class has n CSE, m MSC and r MC students. Suppose that t distinct

books are to be distributed so that each branch gets at least s. In how many ways can this
be done, if each student gets at most one? In how many ways can this be done, without
the previous restriction? Answer only using generating function.
2. My class has N students. Assume that, to conduct an exam, we have M identical answer
scripts. In how many ways can we distribute the answer scripts so that each student gets
at least 2. Answer only using generating function.
3. My class has N students. Assume that, for an exam, we have M questions; each student
answers all the questions in an order decided by him/her (for example one can follow
1, 2, , M and another can follow M, M 1, , 1). In how many ways can it happen that
some three or more students have followed the same order? Answer only using generating

function.
4. When Freshers Welcome was organized 11 teachers went to attend. There were 4 types of
soft drinks available. In how many ways a total of 18 glasses of soft drinks can be served
to them, in general? Answer only using generating function.

144

CHAPTER 7. ADVANCED COUNTING PRINCIPLES

Chapter 8

Graphs
8.1

Basic Concepts

Experiment
Start from a dot. Move through each line exactly once. Draw it. Which of the following
pictures can be drawn? What if we want the starting dot to be the finishing dot ?
b

Later, we shall see a theorem by Euler addressing this question.

Definition 8.1.1. [Pseudograph, Vertex set and Edge set] A pseudograph or a general
graph G is a pair (V, E) where V is a nonempty set and E is a multiset of unordered pairs of
points of V . The set V is called the vertex set and its elements are called vertices. The set
E is called the edge set and its elements are called edges.
 

Example 8.1.2. G = [4], {1, 1}, {1, 2}, {2, 2}, {3, 4}, {3, 4}
is a pseudograph.

Discussion 8.1.3. A pseudograph can be represented in picture in the following way.

1. Put different points on the paper for vertices and label them.
2. If {u, v} appears in E some k times, draw k distinct lines joining the points u and v.
3. A loop at u is drawn if {u, u} E.
Example 8.1.4. A picture for the pseudograph in Example 8.1.2 is given in Figure 8.1.
Definition 8.1.5. [Loop, End vertex and Incident vertex/edge]
1. An edge {u, v} is sometimes denoted uv. An edge uu is called a loop. The vertices u and
v are called the end vertices of the edge uv. Let e be an edge. We say e is incident on

u to mean that u is an end vertex of e.

145

146

CHAPTER 8. GRAPHS

1
b

3
b

Figure 8.1: A pseudograph

2. [Multigraph and simple graph]

A multigraph is a pseudograph without loops. A

multigraph is a simple graph if no edge appears twice.1

3. Henceforth, all graphs in this book are simple with a finite vertex set, unless stated otherwise.
4. We use V (G) (or simply V ) and E(G) (or simply E) to denote the vertex set and the
edge set of G, respectively. The number |V (G)| is the order of the graph G. Sometimes
it is denoted |G|. By kGk we denote the number of edges of G. A graph with n vertices

and m edges is called a (n, m) graph. The (1, 0) graph is the trivial graph.
5. [Neighbor and independent set] If uv is an edge in G, then we say u and v are adjacent
in G or u is a neighbor of v. We write u v to denote that u is adjacent to v. Two

edges e1 and e2 are adjacent if they have a common end vertex. A set of vertices or edges
is independent if no two of them are adjacent.

6. [Isolated and pendant vertex] If v V (G), by N (v) or NG (v), we denote the set of
neighbors of v in G and |N (v)| is called the degree of v. It is usually denoted by dG (v) or

d(v). A vertex of degree 0 is called isolated. A vertex of degree one is called a pendant
vertex.
Discussion 8.1.6. Note that a graph is an algebraic structure, namely, a pair of sets satisfying
some conditions. However, it is easy to describe and carry out the arguments with a pictorial
representation of a graph. Henceforth, the pictorial representations are used to describe graphs
and to provide our arguments, whenever required. There is no loss of generality in doing this.
Example 8.1.7. Consider the graph G in Figure 8.2. The vertex 12 is an isolated vertex. We
have N (1) = {2, 4, 7}, d(1) = 3. The set {9, 10, 11, 2, 4, 7} is an independent vertex set. The set


{1, 2}, {8, 10}, {4, 5} is an independent edge set. The vertices 1 and 6 are not adjacent.

Definition 8.1.8. [Complete graph, path graph, cycle graph and bipartite graph] Let G =
(V, E) be a graph on n vertices, say V = {v1 , . . . , vn }. Then, G is said to be a
1. complete graph, denoted Kn , if each pair of vertices in G are adjacent.
2. path graph, denoted Pn , if E = {vi vi+1 | 1 i n 1}.
1

A simple graph is a hypergraph, (V, E), if E is a collection of nonempty subsets of V .

8.1. BASIC CONCEPTS

147

4
5
6

8
10
b

11

12

2
b

7
b

13

3
b

Figure 8.2: A graph G.

3. cycle graph, denoted Cn , if E = {vi vi+1 | 1 i n 1} {vn v1 }.
4. complete bipartite graph, denoted Kr,s and E = {vi vj | 1 i r, r + 1 j n} with
r + s = n.

The importance of the labels of the vertices depends on the context. At this point of time,
even if we interchange the labels of the vertices, we still call them a complete graph or a path
graph or a cycle or a complete bi-partite graph.

n 1
Pn

n 1
Cn

Figure 8.3: Pn and Cn .

Quiz 8.1.9. What is the maximum number of edges possible in a simple graph of order n?1
Lemma 8.1.10. [Hand shaking lemma] In any graph G,

Proof. Each edge contributes 2 to the sum

d(v). Hence,

vV

is even. So,
d(v)

d(v) =
d(v)

is odd

d(v) = 2|E|. Note that

vV

vV

d(v) = 2|E|. Thus, the number

vV

of vertices of odd degree is even.

2|E| =

d(v) +
d(v)

d(v)

is even

d(v) is even. Hence, the number of vertices of odd degree is even.

is odd

Quiz 8.1.11. In a party of 27 persons, prove that someone must have an even number of friends
(friendship is mutual).

Proposition 8.1.12. In a graph G with n = |G| 2, there are two vertices of equal degree.
1
2

C(n, 2).
P
Otherwise
d(v) is odd.

148

CHAPTER 8. GRAPHS
1

K1,1

K1,2

K2,2

K2,3

4
1

K1

K2

K3

2
K4

4
2

C3

C4

P1

P2

1
K5
4

1
3

C5

C6

P3

P4

P5

Figure 8.4: Some well known family of graphs

Proof. If G has two or more isolated vertices, we are done. So, suppose G has exactly one
isolated vertex. Then, the remaining n 1 vertices have degree between 1 and n 2 and hence

by PHP, the result follows. If G has no isolated vertex then G has n vertices whose degree lie
between 1 and n 1. Now, again apply PHP to get the required result.

Example 8.1.13. The graph in Figure 8.5 is called the Petersen graph. We shall use it as
an example in many places.
3
b

7
b

4
b

2
b

8
b

10

b
b

5
Figure 8.5: Petersen graphs
Exercise 8.1.14.

1. Let X = (V, E) be a graph with a vertex v V of odd degree. Then,

prove that there exists a vertex u V such that there is a path from v to u and deg(u) is

8.1. BASIC CONCEPTS

149

also odd.
2. Let X = (V, E) be a graph having exactly two vertices, say u and v, of odd degree. Then,
prove that there is a path in X connecting u and v.
Definition 8.1.15. [Regular graph, cubic graph] The minimum degree of a vertex in G is
denoted (G) and the maximum degree of a vertex in G is denoted (G). A graph G is called
k-regular if d(v) = k for all v V (G). A 3-regular graph is called cubic.
Example 8.1.16.

1. The graph Kn is regular.

2. The graph K4 is cubic.

3. The graph C4 is 2-regular.
4. The graph P4 is not regular.
5. The Petersen graph is cubic.
6. Consider the graph G in Figure 8.2. We have (G) = 0 and (G) = 3.
Quiz 8.1.17. Can we have a cubic graph on 5 vertices?1
Definition 8.1.18. [Subgraph, induced subgraph, spanning subgraph and k-factor] A graph
H is a subgraph of G if V (H) V (G) and E(H) E(G). If U V (G), then the subgraph

induced by U is denoted by hU i = (U, E), where the edge set E = {uv E(G) | u, v U }. A
subgraph H of G is a spanning subgraph if V (G) = V (H). A k-regular spanning subgraph is

called a k-factor.
Example 8.1.19.

1. Consider the graph G in Figure 8.2.



(a) Let H1 be the graph with V (H1 ) = {6, 7, 8, 9, 10, 12} and E(H1 ) = {6, 7}, {9, 10} .
Then, H1 is not a subgraph of G.



(b) Let H2 be the graph with V (H2 ) = {6, 7, 8, 9, 10, 12} and E(H2 ) = {6, 7}, {8, 10} .
Then, H2 is a subgraph but not an induced subgraph of G.

(c) Let H3 be the induced subgraph of G on the vertex set {6, 7, 8, 9, 10, 12}. Then, verify


that E(H3 ) = {6, 7}, {8, 9}, {8, 10} .

(d) The graph G does not have a 1-factor.

2. A complete graph has a 1-factor if and only if it has an even order.

3. The Petersen graph has many 1-factors. One of them is obtained by selecting the edges
{1, 6}, {2, 7}, {3, 8}, {4, 9}, and {5, 10}.
Quiz 8.1.20. Consider K8 on the vertex set [8]. How many 1-factors does it have?2
Definition 8.1.21. [Vertex/edge deleted graph] Let G be a graph and v be a vertex. Then,
the graph G v is obtained by deleting v and all the edges that are incident with v. If
e E(G), then the graph G e = (V, E(G) \ {e}). If u, v V (G) such that u v, then
G + uv = (V, E(G) {uv}).
1

P
No, as
d(v) = 15, not even.
4
8!/(2!) .

150

CHAPTER 8. GRAPHS

Example 8.1.22. Consider the graph G in Figure 8.2. Let H2 be the graph with V (H2 ) =


{6, 7, 8, 9, 10, 12} and E(H2 ) = {6, 7}, {8, 10} . Consider the edge e = {8, 9}. Then, H2 + e is
the induced subgraph h{6, 7, 8, 9, 10, 12}i and H2 8 = h{6, 7, 9, 10, 12}i.

Definition 8.1.23. [Complement graph] The complement G of a graph G is defined as

(V (G), E), where E = {uv | u 6= v, uv
/ E(G)}.
Example 8.1.24.
4

1. See the graphs in Figure 8.6.

4

4
1

C4

C4

C5

C5 = C5

Figure 8.6: Complement graphs

2. The complement of K3 contains 3 isolated points.
3. For any graph G, kGk + kGk = C(|G|, 2).
4. In any graph G of order n, dG (v) + dG (v) = n 1. Thus, (G) + (G) n 1.
1. Characterize graphs G such that (G) + (G) = n 1.1

Quiz 8.1.25.

2. Can we have a graph G such that (G) + (G) = n?

3. Show that a k-regular simple graph on n vertices exists if and only if kn is even and
n k + 1.
Definition 8.1.26. [Intersection, union and disjoint union] The intersection of two graphs
G and H, denoted G H, is defined as (V (G) V (H), E(G) E(H)). The union of two graphs

G and H, denoted G H, is defined as (V (G) V (H), E(G) E(H)). A disjoint union of two
graphs is the union while treating the vertex sets as disjoint sets.

Example 8.1.27. Two graphs G and H are shown below. The graphs G H and G H are

also shown below.

2
4
b

3
G

1
H

GH

b
b

1
GH

The disjoint union of G and G H is G1 in Figure 8.7.

1

If dG (u) < dG (v), then dG (u) = n 1 dG (u). Hence, (G) + (G) dG (v) + n 1 dG (u) > dG (v) + n

1 dG (v) n. Thus, the answer is regular graphs.

8.2. CONNECTEDNESS

151

2
b
b

K2
b

K3

K2 + K3

G1

K2 + K2

Figure 8.7: Disjoint union and join of graphs

Definition 8.1.28. [Join of two graphs] If V (G) V (G ) = , then the join G + G is defined

as G G + {vv : v V, v V }. The first + means the join of two graphs and the second

+ means adding a set of edges to a given graph.

Example 8.1.29. (a) K2 + K3 = K5 .
(b) K2 + K2 = C4 .

Quiz 8.1.30. What is the complement of the disjoint union of G and H?1
Definition 8.1.31. [Cartesian product of two graphs] Let G = (V, E) and G = (V , E ) be
two graphs. Then, the cartesian product of G and G , denoted G G = (V1 , E1 ), is a graph
having V1 = V V and whose edge set consists of all elements {(u1 , u2 ), (v1 , v2 )}, where either
u1 = v1 and {u2 , v2 } E or u2 = v2 and {u1 , v1 } E.

Example 8.1.32. See the graphs in Figure 8.8.

13
1b

2b

3b
12

a
X

1a

2a

3a

X Y

11

23
22
21

33
32
31

Y Y

Figure 8.8: Cartesian product of graphs

8.2

Connectedness

Definition 8.2.1. [Walk, trail, path, cycle, circuit, length and internal vertex] An u-v
walk in G is a finite sequence of vertices [u = v1 , v2 , , vk = v] such that vi vi+1 E, for all

i = 1, , k 1. The length of a walk is the number of edges on it. A walk is called a trail if

edges on the walk are not repeated. A v-u walk is a called a path if the vertices involved are all

distinct, except that v and u may be the same. A path can have length 0. A walk (trail, path)
1

G + H.

152

CHAPTER 8. GRAPHS

is called closed if u = v. A closed path is called a cycle/circuit. Thus, in a simple graph a

cycle has length at least 3. A cycle (walk, path) of length k is also written as a k-cycle (k-walk,
k-path). If P is an u-v path with u 6= v, then we sometimes call u and v as the end vertices

of P and the remaining vertices on P as the internal vertices.

Example 8.2.2.
(a) Take G = K5 with vertex set [5].

Then, [1, 2, 3, 2, 1, 2, 5, 4, 3] is a 8-walk in G and [1, 2, 2, 1] is not a walk.

The walk [1, 2, 3, 4, 5, 2, 4, 1] is a closed trail.

The walk [1, 2, 3, 5, 4, 1] is a closed path, that is, it is a 5-cycle.

The maximum length of a cycle in G is 5 and the minimum length of a cycle in G is 3.

There are 10 = C(5, 3) many 3-cycles in G.

Verify that the number of 4-cycles in G is not C(5, 4).

(b) Let G be the Petersen graph.

There is a 9-cycle in G, namely, [6, 8, 10, 5, 4, 3, 2, 7, 9, 6].

There are no 10-cycles in G. We shall see this when we discuss the Eulerian graphs.

Proposition 8.2.3 (Technique). Let G be a graph and u, v V (G), u 6= v. Let W = [u =

u1 , . . . , uk = v] be a walk. Then, W contains an u-v-path.

Proof. If no vertex on W repeats, then W is itself a path. So, let ui = uj for some i < j. Now,
consider the walk W1 = [u1 , . . . , ui1 , uj , uj+1 , . . . uk ]. This is also an u-v walk but of shorter
length. Thus, using induction on the length of the walk, the desired result follows.
Definition 8.2.4. [Distance, diameter, radius, center and girth] The distance d(u, v) of two
vertices in G is the shortest length of an u-v path in G. If no such path exists, the distance
is taken to be . The greatest distance between any two vertices in a graph G is called the

diameter of G. We shall use diam(G) to denote the diameter of G. Let distv = max d(v, u).
uG

The radius is the min distv and the center consists of all vertices v for which distv is the radius.
vG

The girth, denoted g(G), of a graph G is the minimum length of a cycle contained in G. If G
has no cycle, then we put g(G) = .
Example 8.2.5. Let G be the Petersen graph. It has diameter 2. The radius is 2. Each vertex
is in the center. Its girth is 5.
Practice 8.2.6. Determine the diameter, radius, center and girth of the following graphs: Pn ,
Cn , Kn and Kn,m = K n + K m .
Exercise 8.2.7. Let G be a graph. Then, show that the distance function d(u, v) is a metric
on V (G). That is, it satisfies
1. d(u, v) 0 for all u, v V (G) and d(u, v) = 0 if and only if u = v,
2. d(u, v) = d(v, u) for all u, v V (G) and
3. d(u, v) d(u, w) + d(w, u) for all u, v, w V (G).

8.2. CONNECTEDNESS

153

Proposition 8.2.8 (Technique). Let G be a graph with kGk 1 and d(v) 2, for each vertex

except one, say v1 . Then, G has a cycle.

Proof. Consider a longest path [v1 , . . . , vk ] in G (as V (G) is finite, such a path exists). As
d(vk ) 2, it must be adjacent to some vertex from v2 , . . . , vk2 , otherwise, we can extend it to

a longer path. Let i 2 be the smallest such that vi is adjacent to vk . Then, [vi , vi+1 , . . . , vk , vi ]
is a cycle.

Proposition 8.2.9 (Technique). Let P and Q be two different u-v paths in G. Then, P Q
contains a cycle.

Proof. Imagine a signal was sent from u to v via P and was returned back from v to u via Q.
Call an edge dead if signal has passed through it twice. Notice that each vertex receives the
signal as many times as it sends the signal.
Is E(P ) = E(Q)? No, otherwise both P and Q are the same graphs.

So, there are some alive edges. Get an alive edge

v
1 v2 . There must be an alive edge

v
v .1 Similarly get
v
v and so on. Stop at the first instance of repetition of a vertex:
2 3

3 4

[v1 , v2 , , vi , vi+1 , vj = vi ]. Then, [vi , vi+1 , vj = vi ] is a cycle.

Alternate. Consider the graph H = V (P ) V (Q), E(P )E(Q) , where is the symmetric

difference. Notice that E(H) 6= , otherwise P = Q. As the degree of each vertex in the

multigraph P Q is even and H is obtained after deleting pairs of multiple edges, each vertex
in H has even degree. Hence, by Proposition 8.2.8, H has a cycle.

Proposition 8.2.10. Every graph G containing a cycle satisfies g(G) 2 diam(G) + 1.

Proof. Let C = [v1 , v2 , . . . , vk , v1 ] be the shortest cycle and diam(G) = r. If k 2r + 2, then

consider the path P = [v1 , v2 , . . . , vr+2 ]. Since the length of P is r + 1 and diam(G) = r, there

is a vr+2 -v1 path R of length at most r. Note that P and R are different v1 -vr+2 paths. By
Proposition 8.2.9, the closed walk P R of length at most 2r + 1 contains a cycle. Hence, the

length of this cycle is at most 2r + 1, a contradiction to C having the smallest length k 2r + 2.

Definition 8.2.11. [Chord, chordal and acyclic graphs] Let C = [v1 , . . . , vk = v1 ] be a cycle.
An edge vi vj is called a chord of C if it is not an edge of C. A graph is called chordal if each
cycle of length at least 4 has a chord. A graph is acyclic if it has no cycles.
Example 8.2.12. Complete graphs are chordal, so are the acyclic graphs. The Petersen graph
is not chordal.
Quiz 8.2.13.

1. How many acyclic graphs are there on the vertex set [3]?2

2. How many chordal graphs are there on the vertex set [4]?3
1

Otherwise, v2 is incident to just one alive edge and some dead edges. This means v2 has received more signal

than it has sent.

2
7: 3 edges can be put in 23 ways. One of them is a cycle.
3
61: 6 edges can be put in 26 ways. There are three 4-cycles.

154

CHAPTER 8. GRAPHS

Definition 8.2.14.

1. [Maximal and minimal graph] A graph G is said to be maximal

with respect to a property P if G has property P and no proper supergraph of G has the
property P . We similarly define the term minimal.
Notice!
The class of all graphs with that property is the POSET here. So, the maximality and
the minimality are defined naturally.
2. [Clique, clique number and connected graph] A complete subgraph of G is called a
clique. The maximum order of a clique is called the clique number of G. It is denoted
(G). A graph G is called connected if there is an u-v path, for each u, v V (G).
3. [Disconnected graph and component of a graph] A graph which is not connected is
called disconnected. If G is a disconnected graph, then a maximal connected subgraph
is called a component or sometimes a connected component.
Example 8.2.15. Consider the graph G shown in Figure 8.2. Then,
1. some cliques in G are h{8, 10}i, h{2}i, h{1, 2, 4}i. The first and the last are maximal cliques.
Notice that every vertex is a clique. Similarly each edge is a clique. Here (G) = 3.

2. the graph G is not connected. It has four connected components, namely, h{8, 9, 10, 11}i,
h{1, 2, 3, 4, 5, 6, 7}i, h{12}i and h{13}i.

Quiz 8.2.16. What is (G) for the Petersen graph?1

Proposition 8.2.17. If (G) 2, then G has a path of length (G) and a cycle of length at
least (G) + 1.

Proof. Let [v1 , , vk ] be a longest path in G. As d(vk ) 2, vk is adjacent to some vertex

v 6= vk1 . If v is not on the path, then we have a path that is longer than [v1 , , vk ] path. A

contradiction. Let i be the smallest positive integer such that vi is adjacent to vk . Thus,
(G) d(vk ) |{vi , vi+1 , , vk1 }|.

Hence, the cycle C = [vi , vi+1 , , vk , vi ] has length at least (G) + 1 and the length of the path

P = [vi , vi+1 , , vk ] is at least (G).

Definition 8.2.18. [Edge density] The edge density, denoted (G), is defined to be the
number

|E(G)|
|V (G)| .

Quiz 8.2.19.
2. Is
1
2

(G)
2

Observe that (G) is also a graph invariant.

1. When does deletion of a vertex reduce edge density?2

a lower bound for (G)?3

2.
Put H = G v. Then, kHk = (G)n d(v), so that (H) =

choose a vertex v with degree more that (G).

3
Yes.

(G)nd(v)
n1

= (G) +

(G)d(v)
.
n1

So, we should

8.3. ISOMORPHISM IN GRAPHS

155

3. Suppose that (G) (G). Should we have a vertex v with (G) d(v)?1
Proposition 8.2.20. Let G be a graph with kGk 1. Then, G has a subgraph H with (H) >

(H) (G).

Proof. If (G) < (G), then we take H = G. Otherwise, there is a vertex v with (G) d(v).
Put G1 = G v. Then, it can be easily verified that (G1 ) (G).

If (G1 ) < (G1 ), then we take H = G1 . Otherwise, there is a vertex v G1 with (G1 ) d(v).

Put G2 = G1 v. Then, we again have (G2 ) (G1 ) (G).

Continuing as above, we note that Initially (G) > 0. At the i-th stage, we obtained the

subgraph Gi satisfying |V (Gi )| = |G| i, (Gi ) (Gi1 ). That is, we have been reducing the
number of vertices and the corresponding edge densities have been nondecreasing. Hence, this
process must stop before we reach a single vertex, as its edge density is 0.
So, let us assume that the process stops at H. Then, (H) < (H) must be true, or else, the
process would not stop at H and hence the required result follows.

8.3

Isomorphism in graphs

Definition 8.3.1. [Isomorphic graphs] Two graphs G = (V, E) and G = (V , E ) are said to
be isomorphic if there is a bijection f : V V such that u v is G if and only if f (u) f (v)

in G , for each u, v V . In other words, an isomorphism is a bijection between the vertex sets
which preserves adjacency. We write G
= G to mean that G is isomorphic to G .
Example 8.3.2. Consider the graphs in Figure 8.9. Then, note that
4
b

5
6

1
F

2
b

1
b

5
b

6
b

2
1
b

G
H
Figure 8.9: F is isomorphic to G but F is not isomorphic to H

1. the graph F is not isomorphic to H as the independence number, denoted (F ), of F

(the maximum size of an independent vertex set) is 3 whereas (H) = 2. Alternately, H
has a 3-cycle, whereas F does not.
2. the graph F is isomorphic to G as the map f : V (F ) V (G) defined by f (1) = 1,
f (2) = 5, f (3) = 3, f (4) = 4, f (5) = 2 and f (6) = 6 gives an isomorphism.

Yes. Otherwise, we have (G) < d(v), for each v. In particular (G) < (G), a contradiction.

156

CHAPTER 8. GRAPHS
Check the adjacency
F

1 2, 4, 6

f (1) = 1 f (2) = 5, f (4) = 4, f (6) = 6

3 2, 4, 6

f (3) = 3 f (2) = 5, f (4) = 4, f (6) = 6

5 2, 4, 6 f (5) = 2 f (2) = 5, f (4) = 4, f (6) = 6

All edges are covered, no need to check any further.
Thus, f is an isomorphism.
Discussion 8.3.3. [Isomorphism] Let F and G be isomorphic under f : V (F ) V (G). Take

F . Relabel each vertex v F as f (v). Call the new graph F . Then, F = G. This is so, as
V (F ) = V (G) and E(F ) = E(G) due to the isomorphic nature of the function f .

Practice 8.3.4. Take the graphs F and G of Figure 8.9. Take the isomorphism f (1) = 1,
f (2) = 5, f (3) = 3, f (4) = 4, f (5) = 2 and f (6) = 6. Obtain the F as described in Discussion
8.3.3. List V (F ) and E(F ). List V (G) and E(G). Notice that they are the same.
Definition 8.3.5. [Self-complementary] A graph G is called self-complementary if G
= G.
Example 8.3.6.

1. Note that the cycle C5 = [0, 1, 2, 3, 4, 0] is self complimentary. An iso-

morphism from G to G is described by f (i) = 2i (mod 5).

2. If |G| = n and G
= G then kGk = n(n 1)/4. Thus, n = 4k or n = 4k + 1.
Exercise 8.3.7.

1. Construct a self-complementary graph of order 4k.

2. Construct a self-complementary graph of order 4k + 1.

Definition 8.3.8. A graph invariant is a function which assigns the same value (output) to
isomorphic graphs.
Example 8.3.9. Observe that some of the graph invariants are: |G|, kGk, (G), (G), the

multiset {d(v) : v V (G)}, (G) and (G).

Exercise 8.3.10. How many graphs are there with vertex set {1, 2, . . . , n}? Do you find it easy
if we ask for nonisomorphic graphs (try for n = 4)?

Proposition 8.3.11 (Technique). Let f : G H be an isomorphism and v V (G). Then,

Gv
= H f (v).
Proof. Consider the bijection g : V (G v) V (H f (v)) described by g = fV (Gv) .
Definition 8.3.12. An isomorphism of G to G is called an automorphism.
Example 8.3.13.

1. Identity map is always an automorphism on any graph.

2. Any permutation in Sn is an automorphism of Kn .

3. There are only two automorphisms of a path P8 .
Proposition 8.3.14. Let G be a graph and let (G) denote the set of all automorphisms of G.
Then, (G) forms a group under composition of functions.

8.4. TREES

157

Proof. Let V (G) = [n] and , (G) be two automorphisms. Then,

ij E(G) (i)(j) E(G) ( )(i)( )(j) E(G).
Thus, is an automorphism. Moreover, 1 , 1 are indeed automorphisms.
Example 8.3.15. Determine (C5 ).
Ans: Consider C5 = [1, . . . , 5, 1]. Note that = (2, 3, 4, 5, 1) is an automorphism. Hence,
{e, , 2 , . . . , 4 } (C5 ) as 5 = e.

Now, let be an automorphism with (1) = i. Put = 6i . Then, is an automorphism

with (1) = 1. If (2) = 2, then the adjacency structure implies that (j) = j for j = 3, 4, 5.
Hence, in this case, 6i = e and thus, = i6 = i1 .
If (2) 6= 2, then (2) = 5, (3) = 4 and so = (2, 5)(3, 4) is the reflection which fixes 1. Let

us denote the permutation (2, 5)(3, 4) by . Then, (C5 ) is the group generated by and and
hence (C5 ) has 10 elements.
Example 8.3.16. Notice that (C5 ) has a subgroup 1 = {e, , 2 , . . . , 4 }, with 5 = e, of

order 5. Let G be a subgraph of C5 obtained by deleting some (zero allowed) edges. If kGk = 5,

then |(G)| = 10. If kGk = 0, then |(G)| = |S5 | = 5!. If kGk = 4, then |(G)| = 2. If kGk = 3,

then |(G)| = 2 or 4. If kGk = 2, then |(G)| = 4 or 8. If kGk = 1, then |(G)| = 2 3!. Thus,

there is no subgraph of G whose automorphism group is 1 .

Exercise 8.3.17.

1. Determine the graphs G for which (G) = Sn , the group of all permu-

tations of 1, . . . , n.
2. Compute (G) for some graphs of small order.
3. Let G be a subgraph of H of the same order. Explore more about the relationship between
(G) and (H).

8.4

Trees

Definition 8.4.1. [Tree and forest] A connected acyclic graph is called a tree. A forest is a
graph whose components are trees.
Proposition 8.4.2. Let T be a tree and u, v V (T ). Then, there is a unique u-v-path in T .
Proof. On the contrary, assume that there are two u-v-paths in T . Then, by Proposition 8.2.9,
T has a cycle, a contradiction.
Proposition 8.4.3. Let G be a graph with the property that between each pair of vertices there
is a unique path. Then, G is a tree.
Proof. Clearly, G is connected. If G has a cycle [v1 , v2 , , vk = v1 ], then [v1 , v2 , . . . , vk1 ] and

[v1 , vk1 ] are two v1 -vk1 paths. A contradiction.

Definition 8.4.4. [Cut vertex] Let G be a connected graph. A vertex v of G is called a cut
vertex if G v is disconnected. Thus, G v is connected if and only if v is not a cut vertex.

158

CHAPTER 8. GRAPHS

Proposition 8.4.5. Let G be a connected graph with |G| 2. If v V (G) with d(v) = 1, then
G v is connected. That is, a vertex of degree 1 is never a cut vertex.

Proof. Let u, w V (G v), u 6= w. As G is connected, there is an u-w path P in G. The vertex

v cannot be an internal vertex of P , as each internal vertex has degree at least 2. Hence, the
path P is available in G v. So, G v is connected.
Proposition 8.4.6 (Technique). Let G be a connected graph with |G| 2 and let v V (G). If
G v is connected, then either d(v) = 1 or v is on a cycle.

Proof. Assume that G v is connected. If dG (v) = 1, then there is nothing to show. So, assume

that d(v) 2. We need to show that v is on a cycle in G.

Let u and w be two distinct neighbors of v in G. As G v is connected there is a path, say

[u = u1 , . . . , uk = w], in G v. Then, [u = u1 , . . . , uk = w, v, u] is a cycle in G containing v.

Quiz 8.4.7. Let G be a graph and v be a vertex on a cycle. Can G v be disconnected?1

Definition 8.4.8. [Cut edge] Let G be a graph. An edge e in G is called a cut edge or a
bridge if G e has more connected components than that of G.
Proposition 8.4.9 (Technique). Let G be connected and e = [u, v] be a cut edge. Then, G e
has two components, one containing u and the other containing v.

Proof. If G e is not disconnected, then by definition, e cannot be a cut edge. So, G e has

at least two components. Let Gu (respectively, Gv ) be the component containing the vertex u
(respectively, v). We claim that these are the only components.
Let w V (G). Then, G is a connected graph and hence there is a path, say P , from w to

u. Moreover, either P contains v as its internal vertex or P doesnt contain v. In the first case,

w V (Gv ) and in the latter case, w V (Gu ). Thus, every vertex of G is either in V (Gv ) or in
V (Gu ) and hence the required result follows.

Proposition 8.4.10 (Technique). Let G be a graph and e be an edge. Then, e is a cut edge if
and only if e is not on a cycle.
Proof. Suppose that e = [u, v] is a cut edge of G. Let F be the component of G that contains
e. Then, by Proposition 8.4.9, F e has two components, namely, Fu that contains u and Fv
that contains v.

Let if possible, C = [u, v = v1 , . . . , vk = u] be a cycle containing e = [u, v]. Then, [v =

v1 , . . . , vk = u] is an u-v path in F e. Hence, F e is still connected. A contradiction. Hence,
e cannot be on any cycle.

Conversely, let e = [u, v] be an edge which is not on any cycle. Now, suppose that F is the
component of G that contains e. We need to show that F e is disconnected.

Let if possible, there is an u-v-path, say [u = u1 , . . . , uk = v], in F e. Then, [v, u =

u1 , . . . , uk = v] is a cycle containing e. A contradiction to e not lying on any cycle.

Hence, e is a cut edge of F . Consequently, e is a cut edge of G.
1

Yes. Take G = ([4], {{1, 2}, {1, 3}, {1, 4}, {3, 4}}) and v = 1.

8.4. TREES

159

Proposition 8.4.11. The center of a tree always consists of a set of at most two vertices.
Proof. Let T be a tree of radius k. Since the center contains at least one vertex, let u be a vertex
in the center of T . Now, let v be another vertex in the center. We claim that u is adjacent to v.
Suppose u v. Then, there exists a path from u to v, denoted P (u, v), with at least one
internal vertex, say w. Let x be any pendant (d(x) = 1) vertex of T . Then, either v P (x, w)
or v
/ P (x, w). In the latter case, check that kP (x, w)k < kP (x, v)k k.
u

w
b

v
b

b
b

If v P (x, w), then u

/ P (x, w) and kP (x, w)k < kP (x, u)k k. That is, the distance from w

to any pendant vertex is less than k. Hence, k is not the radius, a contradiction. Thus, uv T .
We cannot have another vertex in the center, or else, we will have a C3 in T , a contradiction.

Theorem 8.4.12. Let G be a graph with V (G) = [n]. Then, the following are equivalent.
1. G is a tree.
2. G is a minimal connected graph on n vertices.
3. G is a maximal acyclic graph on n vertices.
Proof. (a)(b). Suppose that G is a tree. If it is not a minimal connected graph on n vertices,
then there is an edge [u, v] such that G [u, v] is connected. But then, by Theorem 8.4.10, [u, v]
is on a cycle in G. A contradiction.

(b)(c). Suppose G is a minimal connected graph on n vertices. If G has a cycle, say , then
select an edge e . Thus, by Theorem 8.4.10, G e is still connected graph on n vertices, a

contradiction to the fact that G is a minimal connected graph on n vertices. Hence, G is acyclic.

Since G is connected, for any new edge e, the graph G + e contains a cycle and hence, G is
maximal acyclic graph.
(c)(a). Suppose G is maximal acyclic graph on n vertices. If G is not connected, let G1 and
G2 be two components of G. Select v1 G1 and v2 G2 and note that G + [v1 , v2 ] is acyclic

graph on n vertices. This contradicts that G is a maximal acyclic graph on n vertices. Thus, G
is connected and acyclic and hence is a tree.
Exercise 8.4.13.

1. Show that a graph G is a tree if and only if between each pair of vertices

of G there is a unique path.

2. Draw a tree on 8 vertices. Label V (T ) as 1, . . . , 8 so that each vertex i 2 is adjacent to
exactly one element of [i 1].

Proposition 8.4.14. Let T be a tree. Then, any graph G with (G) |T | 1 has a subgraph
H
= T.
Proof. We prove the result by induction on n = |T |. The result is trivially true if n = 1 or 2.

So, let the result be true for every tree on n 1 vertices and take a tree T on n vertices. Also,
suppose that G is any graph with (G) |T | 1.

160

CHAPTER 8. GRAPHS

Let v V (T ) with d(v) = 1. Take u V (T ) such that uv E(T ). Now, consider the tree

T1 = T v. Then, (G) |T | 1 = n 1 > n 2. Hence, by induction hypothesis, G has a

subgraph H such that H
= T1 under a map, say . Let h V (H) such that (h) = u. Since
(G) |T | 1, h has a neighbor, say h1 , such that h1 is not a vertex in H but is a vertex in G.

Now, map this vertex to v to get the required result.

Proposition 8.4.15. Let T be a tree on n vertices. Then, T has n 1 edges.

Proof. We proceed by induction. Take a tree on n 2 vertices and delete an edge e. Then, we get

two subtrees T1 , T2 of order n1 , n2 , respectively, where n1 + n2 = n. So, E(T ) = E(T1 ) E(T2 )

{e}. By induction hypothesis kT k = kT1 k + kT2 k + 1 = n1 1 + n2 1 + 1 = n1 + n2 1 = n 1.

Proposition 8.4.16. Let G be a connected graph with n vertices and n 1 edges. Then, G is
acyclic.

Proof. On the contrary, assume that G has a cycle, say . Now, select an edge e and note

that G e is connected. We go on selecting edges from G that lie on cycles and keep removing

them, until we get an acyclic graph H. Since the edges that are being removed lie on some
cycle, the graph H is still connected. So, by definition, H is a tree on n vertices. Thus, by
Proposition 8.4.15, |E(H)| = n 1. But, in the above argument, we have deleted at least one

edge and hence, |E(G)| n. This gives a contradiction to |E(G)| = n 1.

Proposition 8.4.17. Let G be an acyclic graph with n vertices and n 1 edges. Then, G is
connected.

Proof. Let if possible, G be disconnected with components G1 , . . . , Gk , k 2. As G is acyclic,

P
by definition, each Gi is a tree on, say ni 1 vertices, with
i = 1k ni = n. Thus, kGk =
k
P
(ni 1) = n k < n 1 = kGk, as k 2. A contradiction.

i=1

Theorem 8.4.18. The following are Equivalent for a graph of order n.

(a) G is a tree.
(b) G is minimal connected.
(c) G is maximal acyclic.
(d) G is acyclic with kGk = n 1.
(e) G is connected with kGk = n 1.
Proof. Left as an exercise.

Exercise 8.4.19. Let G be a graph on n > 2 vertices. If kGk > C(n 1, 2), is G necessarily

connected? Give an if and only if condition for the connectedness of a graph with exactly
C(n 1, 2) edges.
Proposition 8.4.20. A tree on n 2 vertices has at least two pendant vertices.

8.4. TREES

161

Proof. Let T be any tree on n vertices. Then,

vV (T )

d(v) = 2kE(T )k = 2(n 1) = 2n 2.

Hence, by PHP, T has at least two vertices of degree 1.

Definition 8.4.21. Let T be a tree with vertices labeled by n integers, say [n]. The Pr
ufer
code PT of T is a sequence X of size n 2 created in the following way.
1. Find the largest pendant vertex, say v1 . Let u1 be the neighbor of v1 . Put X(1) = u1 .
2. Let T1 = T v1 and find X(2).
3. Repeat the procedure to obtain X(3), . . . , X(n 2).
Example 8.4.22. For example, Consider the tree T in Figure 8.10.
5
b

4
b

Figure 8.10: A tree T on 6 vertices

Then, the above process proceeds as follows.
Step Pendant vi Neighbor ui PT = X(1), X(2), . . .

Ti = T vi
b

3
b

2
b

3
b

2,2

2,2,2

2,2,2,6

Exercise 8.4.23. In the above process, prove that uj = i, for some j, if and only if d(i) 2.
Example 8.4.24. Can I get back the original tree T from the sequence 2, 2, 2, 6? Ans: Yes.
The process of getting back the original tree is as follows.
1. Plot points 1, 2, . . . , 6.
2. Since ui is either 2 or 6, it implies that 2 and 6 are not the pendant vertices. Hence, the
pendant vertices in T must be {1, 3, 4, 5}. Thus, the algorithm implies that the largest
pendant 5 must be adjacent to (the first element of the sequence) 2.

3. At step 1, the vertex 5 was deleted. Hence, V (T1 ) = {1, 2, 3, 4, 6} with the given sequence
2, 2, 6. So, the pendants in T1 are {1, 3, 4} and the vertex 4 (largest pendant) is adjacent
to 2.

162

CHAPTER 8. GRAPHS

4. Now, V (T2 ) = {1, 2, 3, 6} with the sequence as 2, 6. So, 3 is adjacent to 2.

5. Now, V (T3 ) = {1, 2, 6} with the sequence as 6. So, the pendants in the current T are
{1, 2} and 2 is adjacent to 6.

6. Lastly, V (T4 ) = {1, 6}. As the process ends with K2 and we have only two vertices left,
they must be adjacent.

The corresponding set of figures are as follows.

1

Proposition 8.4.25. Let T be a tree on the vertex set [n]. Then, d(v) 2 if and only if v
appears in the Pr
ufer code PT . Thus, {v : v
/ PT } are precisely the pendant vertices in T .

Proof. Let d(v) 2. Since the process ends with an edge, there is a stage, say i, where d(v)

decreases strictly. Thus, till at the (i 1)-th stage, v was adjacent to a pendant vertex w and

at the i-th stage v was deleted and thus, v appears in the sequence.

Conversely, let v appear in the sequence at k-th stage for the first time. Then, the tree Tk had
a pendant vertex w of highest label that was adjacent to v. Note that Tk w is a tree with at
least two vertices. Thus, d(v) dTk (v) 2.

Exercise 8.4.26. Prove that in the Pr

ufer code of T a vertex v appears exactly d(v) 1 times.

[Hint: If v is the largest pendant adjacent to w and T = T v, then PT = w, PT .]

Proposition 8.4.27. Let T and T be two trees on the same vertex set of integers. If PT = PT ,
then T = T .
Proof. The statement is trivially true for |T | = 3. Assume that the statement holds for 3 <

|T | < n. Now, let T and T be two trees with vertex set [n] and PT = PT . As PT = PT , T

and T have the same set of pendants. Further, the largest labeled pendant w is adjacent to
the vertex X(1) in both the trees. Thus, PT w = PT w and hence, by induction hypothesis

T w = T w. Thus, by PMI, T = T .

Proposition 8.4.28. Let S be a set of n 3 integers and X be a sequence of length n 2 of

elements form S. Then, there is a tree T with V (T ) = S and PT = X.

Proof. Verify the statement for |T | = 3. Now, let the statement hold for all trees T on n > 3
vertices and consider a set S of n + 1 integers and a sequence X of length (n 1) of elements
of S.

8.4. TREES

163

Let v = max{x S : x
/ X}, S = S v and X = X(2), . . . , X(n 1). Note that as

v 6= X(i), for any i, X is a sequence of elements of S of length n 2. As |S | = n, by induction

hypothesis, there is a tree T with PT = X .

Let T be the tree obtained by adding a new pendant v at the vertex X(1) of T . In T , the
vertices X(i), for i 2, were not available as pendants and now in T the vertex X(1) is also

not available as a pendant. (Here some X(i)s may be the same). Let R = {x S : x
/ X }

be the pendants in T . Then, the set of pendants in T is (R {v}) \ {X(1)} which equals

{x S : x
/ X}. Thus, v is the pendant of T of maximum label. Hence, PT = X.

Theorem 8.4.29. [A. Cayley, 1889, Quart. J. Math] Let n 3. Then, there are nn2
different trees with vertex set [n].

Proof. Let F be the class of trees on the vertex set [n] and let G be the class of n 2-sequences

of [n]. Note that the function f : F G defined as f (T ) = PT , the Pr

ufer code, is a one-one

and onto mapping. As |G| = nn2 , the required result follows.

Exercise 8.4.30.

1. Find out all nonisomorphic trees of order 7 or less.

2. Show that every automorphism of a tree fixes a vertex or an edge.

3. Give a class of trees T with |(T )| = 6.
4. Let T be a tree, (T ), u V (T ) such that 2 (u) 6= u. Can we have an edge [u, v] T
such that (u) = v?

5. Let T be a tree with center {u} and radius r. Let v satisfy d(u, v) = r. Show that r is a
pendant.

6. Let T be a tree with |T | > 2. Let T be obtained from T by deleting the pendant vertices of
T . Show that the center of T is the same as the center of T .

7. Let T be a tree with center {u} and (T ). Show that (u) = u.

8. Is it possible to have a tree such that |(T )| = 7?
9. Construct a tree T on vertices S = {1, 2, 3, 6, 7, 8, 9} for which PT = 6, 3, 7, 1, 2.
10. Practice with examples: get the Pr
ufer code from a tree; get the tree from a given code and
a vertex set.
11. How many trees of the following forms are there on the vertex set [100]?
b

b
b

..
.

b
b

..
.

..
.

b
b

12. Show that any tree has at least (T ) leaves (pendant edges).
13. Let T be a tree and T1 , T2 , T3 be subtrees of T such that T1 T3 6= , T2 T3 6= and
T1 T2 T3 = . Show that T1 T2 = .

164

CHAPTER 8. GRAPHS

14. Let T be a set of subtrees of a tree T . Assume that the trees in T have nonempty pairwise

intersection. Show that their overall intersection is nonempty. Is this true, if we replace
T by a graph G?

15. Recall that a connected graph G is said to be unicyclic if G has exactly one cycle as its
subgraph. Prove that if |G| = kGk, then G is a unicyclic graph.

8.5

Connectivity

Proposition 8.5.1. Let G be a connected graph on vertex set [n]. Then, its vertices can be
labeled in such a way that the induced subgraph on the set [i] is connected for 1 i n.
Proof. If n = 1, there is nothing to prove. Assume that the statement is true if n < k and let G
be a connected graph on the vertex set [k]. If G is a tree, pick any pendant vertex and label it
k. If G has a cycle, pick a vertex on a cycle and label it k. In both the case G k is connected.

Now, use the induction hypothesis to get the required result.

Definition 8.5.2. [Separating set] Let G be a graph. Then, a set X V (G) E(G) is called

a separating set if G X has more connected components than that of G.

Let X be a separating set of G. Then, there exists u, v V (G) that lie in the same component

of G but lie in different components of G X. If {u} V (G) is a separating set of G, then u

is a cut vertex. If {e} E(G) is a separating set of G, then it is a bridge/cut edge.

Example 8.5.3.

1. In a tree, each edge is a bridge and each non-pendant vertex is a cut

vertex. Is it true for a forest?

2. The graph K7 does not have a separating set of vertices. In K7 , a separating set of edges
must contain at least 6 edges.
Definition 8.5.4. [Vertex connectivity] A graph G is said to be k-connected if |G| > k and

G is connected even after deletion of any k 1 vertices. The vertex connectivity (G) of a

non trivial graph G is the largest k such that G is k-connected. Convention: (K1 ) = 0.
Example 8.5.5.

1. Each connected graph of order more than one is 1-connected.

2. A 2-connected graph is also a 1-connected graph.

3. For a disconnected graph, (G) = 0 and for n > 1, (Kn ) = n 1.
4. The graph G in Figure 8.11 is 2-connected but not 3-connected. Thus, (G) = 2.
b

Figure 8.11: graph with vertex connectivity 2

5. The Petersen graph is 3-connected.

8.5. CONNECTIVITY

165

Definition 8.5.6. [Edge connectivity] A graph G is called l-edge connected if |G| > 1 and
G F is connected for every F E(G) with |F | < l. The greatest integer l such that G is
l-edge connected is the edge connectivity of G, denoted (G). Convention: (K1 ) = 0.
Example 8.5.7.

1. Note that (Pn ) = 1, (Cn ) = 2 and (Kn ) = n 1, whenever n > 1.

2. Let T be a tree on n vertices. Then, (T ) = 1.

3. For the graph G in Figure 8.11, (G) = 3.
4. For the Petersen graph G, (G) = 3.
Exercise 8.5.8. Let |G| > 1. Show that (G) = |G| 1 if and only if G = Kn . Can we say

the same for (G)?

Theorem 8.5.9. [H. Whitney, 1932] For any graph G, (G) (G) (G).
Proof. If G is disconnected or |G| = 1, then we have nothing to prove. So, let G be connected

graph and |G| 2. Then, there is a vertex v with d(v) = (G). If we delete all edges incident

on v, then the graph is disconnected. Thus, (G) (G).

Suppose that (G) = 1 and G uv is disconnected with components Cu and Cv . If |Cu | =

|Cv | = 1, then G = K2 and (G) = 1. If |Cu | > 1, then we delete u to see that (G) = 1.

If (G) = k 2, then there is a set of edges, say e1 , . . . , ek , whose removal disconnects G.

Notice that G {e1 , . . . , ek1 } is a connected graph with a bridge, say ek = uv. For each of

e1 , . . . , ek1 select an end vertex other than u or v. Deletion of these vertices from G results
in a graph H with uv as a bridge of a connected component. Note that (H) 1. Hence,

(G) (G).

Exercise 8.5.10. Give a lower bound on the number of edges of a graph G on n vertices with
vertex connectivity (G) = k.
Theorem 8.5.11. [Chartrand and Harary, 1968] For all integers a, b, c such that 0 < a b c,
there exists a graph with (G) = a, (G) = b and (G) = c.
Proof. Omitted, as it is out of the scope of this book.
Theorem 8.5.12. [Mader, 1972] Every graph G of average degree at least 4k has a k-connected
subgraph.
Proof. For k = 1, the assertion is trivial. So, let k 2. Note that
n = |G| (G) 4k 2k 1 and

(8.1)

m = kGk = 2kn (2k 3)(n k + 1) + 1.

(8.2)

We shall use induction to show that if G satisfies Equations (8.1) and (8.2), then G has a kconnected subgraph. If n = 2k1, then m (2k3)(nk+1)+1 = (n2) (n+1)
2 +1 =

G is a graph on n vertices with at least

n(n1)
2

n(n1)
.
2

So,

edges and hence G = Kn . Thus, Kk+1 Kn = G.

Assume n 2k. If v is a vertex with d(v) 2k 3, then we apply induction hypothesis to

G v to get the result. So, let d(v) 2k 2, for each vertex v. If G is k-connected then, we

166

CHAPTER 8. GRAPHS

have nothing to prove. Assume, if possible that G is not k-connected. Then, G = G1 G2 with

|G1 G2 | < k and |G1 |, |G2 | < n. Thus, both G1 V (G2 ) and G2 V (G1 ) have at least one

vertex and there is no edge between them. As the degree of these vertices is at least 2k 2, we
have |G1 |, |G2 | 2k 1.

If G1 or G2 satisfies Equation (8.2), using induction hypothesis, the result follows. Otherwise,

kGi k (2k 3)(|Gi | k + 1), for i = 1, 2 and hence

m = kGk kG1 k + kG2 k (2k 3)(|G1 | + |G2 | 2k + 2) (2k 3)(n k + 1),
a contradiction to Equation (8.2) and hence the required result follows.
Theorem 8.5.13. [Menger] A graph is k-edge-connected if and only if there are k edge disjoint
paths between each pairs of vertices. A graph is k-connected if and only if there are k internally
vertex disjoint paths between each pairs of vertices.
Proof. Omitted.

8.6

Eulerian Graphs

Definition 8.6.1. [Eulerian graph] An Eulerian tour in a graph G is a closed walk [v0 , v1 , . . . , vk , v0 ]
such that each edge of the graph appears exactly once in the walk. The graph G is said to be
Eulerian if it has an Eulerian tour.
Note that by definition, a disconnected graph is not Eulerian. In this section, the graphs can
have loops and multiple edges. The graphs that have a closed walk traversing each edge exactly
once have been named Eulerian graphs due to the solution of the famous K
onigsberg bridge
problem by Euler in 1736. The problem is as follows: The city K
onigsberg (the present day
Kaliningrad) is divided into 4 land masses by the river Pregel. These land masses are joined by
7 bridges (see Figure 8.12). The question required one to answer is there a way to start from
a land mass that passes through all the seven bridges in Figure 8.12 and return back to the
starting land mass? Euler, rephrased the problem along the following lines: Let the four land
masses be denoted by the vertices A, B, C and D of a graph and let the 7 bridges correspond to
7 edges of the graph. Then, he asked does this graph have a closed walk that traverses each
edge exactly once? He gave a necessary and sufficient condition for a graph to have such a
closed walk and thus giving a negative answer to K
onigsberg bridge problem.
One can also relate the above problem to the problem of starting from a certain point, draw a
given figure with pencil such that neither the pencil is lifted from the paper nor a line is repeated
such that the drawing ends at the initial point.
Theorem 8.6.2. [Euler, 1736] A connected graph G is Eulerian if and only if d(v) is even, for
each v V (G).
Proof. Let G have an Eulerian tour, say [v0 , v1 , . . . , vk , v0 ]. Then, d(v) = 2r, if v 6= v0 and v

appears r times in the tour. Also, d(v0 ) = 2(r 1), if v0 appears r times in the tour. Hence,
d(v) is always even.

8.6. EULERIAN GRAPHS

167

C
B

A
Figure 8.12: K
onigsberg bridge problem
Conversely, let G be a connected graph with each vertex having even degree. Let W =
v0 v1 vk be a longest walk in G without repeating any edge in it. As vk has an even degree

it follows that vk = v0 , otherwise W can be extended. If W is not an Eulerian tour then there

exists an edge, say e = vi w, with w 6= vi1 , vi+1 . In this case, wvi vk (= v0 )v1 vi1 vi is a

longer walk, a contradiction. Thus, there is no edge lying outside W and hence W is an Eulerian
tour.

Proposition 8.6.3. Let G be a connected graph with exactly two vertices of odd degree. Then,
there is an Eulerian walk starting at one of those vertices and ending at the other.
Proof. Let x and y be the two vertices of odd degree and let v be a symbol such that v
/ V (G).

Then, the graph H with V (H) = V (G) {v} and E(H) = E(G) {xv, yv} has each vertex of
even degree and hence by Theorem 8.6.2, H is Eulerian. Let = [v, v1 = x, . . . , vk = y, v] be an
Eulerian tour. Then, v is an Eulerian walk with the required properties.

Exercise 8.6.4. Let G be a connected Eulerian graph and e be any edge. Show that G e is
connected.

How to find an Eulerian tour (algorithm)?

Start from a vertex v0 , move via edge that has not been taken and go on deleting them.
Do not take an edge whose deletion creates a non trivial component not containing v0 .
Exercise 8.6.5. Find Eulerian tours for the following graphs.
11
12
10
3

7
b

8
b

13
b

5
6

16

b
b

15

14

13

10

11

12

Theorem 8.6.6. [Finding Eulerian tour] The previous algorithm correctly gives an Eulerian
tour whenever, the given graph is Eulerian.
Proof. Let the algorithm start at a vertex, say v0 . Now, assume that we are at u with H as the
current graph and C as the only non trivial component of H. Thus, dH (u) > 0. Assume that

168

CHAPTER 8. GRAPHS

the deletion of the edge uv creates a non trivial component not containing v0 . Let Cu and Cv
be the components of C uv, containing u and v, respectively.

We first claim that u 6= v0 . In fact, if u = v0 , then H must have all vertices of even degree

and dH (v0 ) 2. So, C is Eulerian. Hence, C uv cannot be disconnected, a contradiction to

C uv having two components Cu and Cv . Thus, u 6= v0 . Moreover, note that the only vertices
of odd degree in C is u and v0 .

Now, we claim that Cu is a non trivial component. Suppose Cu is trivial. Then, v0 Cv , a

contradiction to the assumption that the deletion of the edge uv creates a nontrivial component
not containing v0 . So, Cu is non trivial.
Finally, we claim that v0 Cv . If possible, let v0 Cu . Then, the only vertices in C uv of

odd degree are v Cv and v0 Cu . Hence, C uv + v0 v is a connected graph with each vertex

of even degree. So, by Theorem 8.6.2, the graph C uv + v0 v is Eulerian. But, this cannot be

true as vv0 is a bridge. Thus, v0 Cv .

Hence, Cu is the newly created non trivial component not containing v0 . Also, each vertex of

Cu has even degree and hence by Theorem 8.6.2, Cu is Eulerian. This means, we can take an
edge e incident on u and complete an Eulerian tour in Cu . So, at u if we take the edge e in
place of the edge e, then we will not create a non trivial component not containing v0 .
Thus, at each stage of the algorithm either u = v0 or there is a path from u to v0 . Moreover,
this is the only non trivial connected component. When the algorithm ends, we must have
u = v0 . Because, as seen above, the condition u 6= v0 gives the existence of an edge that is
incident on u and that can be traversed (as dH (u) is odd). Hence, if u 6= v0 , the algorithm

cannot stop. Thus, when algorithm stops u = v0 and all components are trivial.

Exercise 8.6.7. Apply the algorithm to graphs of Exercise 8.6.5. Also, create connected graphs
such that each of its vertex has even degree and apply the above algorithm.
Definition 8.6.8. [bipartite graph] A graph G = (V, E) is said to be bipartite if V = V1 V2

such that |V1 |, |V2 | 1, V1 V2 = and khV1 ik = 0 = khV2 ik. The complete bipartite graph

Km,n is shown below. Notice that Km,n = K m + K n .

all possible edges

Km

Exercise 8.6.9.

Kn

1. Give a necessary and sufficient condition on m and n so that Km,n is

Eulerian.
2. Each of the 8 persons in a room has to shake hands with every other person as per the
following rules:
(a) The handshakes should take place sequentially.

8.7. HAMILTONIAN GRAPHS

169

(b) Each handshake (except the first) should involve someone from the previous handshake.
(c) No person should be involved in 3 consecutive handshakes.
Is there a way to sequence the handshakes so that these conditions are all met?
3. Let G be a connected graph. Then, G is an Eulerian graph if and only if the edge set of G
can be partitioned into cycles.

8.7

Hamiltonian Graphs

Definition 8.7.1. [Hamiltonian] A cycle in G is said to be Hamiltonian if it contains all

vertices of G. If G has a Hamiltonian cycle, then G is called a Hamiltonian graph. Finding a
nice characterization of a Hamiltonian graph is an unsolved problem.
1. For each positive integer n 3, the cycle Cn is Hamiltonian.

Example 8.7.2.

b
b
b

b
b
b

b
b

b
b

b
b

b
b
b

b
b

The Petersen graph

The dodecahedron graph

Figure 8.13: A Hamiltonian and a non-Hamiltonian graph

2. The graphs corresponding to all platonic solids are Hamiltonian.
3. The Petersen graph is a non-Hamiltonian Graph (the proof appears below).
Proposition 8.7.3. The Petersen graph is not Hamiltonian.
Proof. Suppose that the Petersen graph, say G, is Hamiltonian. Also, each vertex of G has
degree 3 and hence, G = C10 + M , where M is a set of 5 chords in which each vertex appears
as an endpoint. We assume that C10 = [1, 2, . . . , 10, 1]. Now, consider the vertices 1, 2 and 3.
3
b

2
b

10

170

CHAPTER 8. GRAPHS

Since, g(G) = 5, the vertex 1 can only be adjacent to one of the vertices 5, 6 or 7. Hence,
if 1 is adjacent to 5, then the third vertex that is adjacent to 10 creates cycles of length 3
or 4. Similarly, if 1 is adjacent to 7, then there is no choice for the third vertex that can be
adjacent to 2. So, let 1 be adjacent to 6. Then, 2 must be adjacent to 8. In this case, note that
there is no choice for the third vertex that can be adjacent to 3. Thus, the Petersen graph is
non-Hamiltonian.
Theorem 8.7.4. Let G be a Hamiltonian graph. Then, for S V (G) with S 6= , the graph

G S has at most |S| components.

Proof. Note that by removing k vertices from a cycle, one can create at most k connected
components. Hence, the required result follows.
Theorem 8.7.5. [Dirac, 1952] Let G be a graph with |G| = n 3 and d(v) n/2, for each
v V (G). Then, G is Hamiltonian.

Proof. Let is possible, G be disconnected. Then, G has a component, say H, with |V (H)| = k

n/2. Hence, d(v) k 1 < n/2, for each v V (H). A contradiction to d(v) n/2, for each

v V (G). Now, let P = [v1 , v2 , , vk ] be a longest path in G. Since P is the longest path, all

neighbors of v1 and vk are in P .

We claim that there exists an i such that v1 vi and vi1 vk . Otherwise, for each vi v1 ,

we must have vi1 vk . Then, |N (vk )| k 1 |N (v1 )|. Hence, |N (v1 )| + |N (vk )| k 1 < n,

a contradiction to d(v) n/2, for each v V (G). So, the claim is valid and hence, we have a
cycle P := v1 vi vi+1 vk vi1 v1 of length k.
We now prove that P gives a Hamiltonian cycle. Suppose not. Then, there exists v V (G)

such that v is outside P and v is adjacent to some vj . Note that in this case, P cannot be the
path of longest length, a contradiction. Thus, the required result follows.
Theorem 8.7.6. [Ore, 1960] Let G be a graph on n 3 vertices such that d(u) + d(v) n, for

every pair of nonadjacent vertices u and v. Then, G is Hamiltonian.

Proof. Exercise.

Exercise 8.7.7. Let u and v be two vertices such that d(u) + d(v) |G|, whenever uv
/ E(G).
Prove that G is Hamiltonian if and only if G + uv is Hamiltonian.

Definition 8.7.8. [closure of a graph] The closure of a graph G, denoted C(G), is obtained
by repeatedly choosing pairs of nonadjacent vertices u, v such that d(u) + d(v) n and adding
edges between them.

Proposition 8.7.9. The closure of G is unique.

Proof. Let K be a closure obtained by adding edges e1 = u1 v1 , . . . , ek = uk vk sequentially and F
be a closure obtained by adding edges f1 = x1 y1 , . . . , fr = xr yr sequentially. Let ei be the first
edge in the e-sequence which does not appear in the f -sequence. Put H = G + e1 + + ei1 .

Then, ei = ui vi implies that ei

/ E(H) and dH (ui )+dH (vi ) n. Also, H is a subgraph of F and

8.7. HAMILTONIAN GRAPHS

171

hence, dF (ui ) + dF (vi ) n. Moreover, ei = ui vi

/ F as ei does not appear in the f -sequence.
Thus, F cannot be a closure and therefore the required result follows.

Exercise 8.7.10. Let G be a graph on n 3 vertices.

1. If G has a cut vertex, then prove that C(G) 6= Kn .
2. Then, prove a generalization of Diracs theorem: If the closure C(G)
= Kn , then G is
Hamiltonian.
Theorem 8.7.11. Let d1 dn be the vertex degrees of G. Suppose that, for each k < n/2

with dk k, the condition dnk n k holds. Then, prove that G is Hamiltonian.

Proof. We show that under the above condition H = C(G)

= Kn . On the contrary, assume
that there exist a pair of vertices u, v V (G) such that uv
/ E(G) and dH (u) + dH (v) n 1.

Among the above pairs, choose a pair u, v V (G) such that uv

/ E(H) and dH (u) + dH (v) is

maximum. Assume that dH (v) dH (u) = k (say). Clearly, k < n/2.

Now, let Sv = {x V (H) | xv

/ E(H), x 6= v} and Su = {w V (H) | wu
/ E(H), w 6= u}.

Therefore, the assumption that dH (u) + dH (v) is the maximum among each pair of vertices u, v

with uv
/ E(H) and dH (u) + dH (v) n 1 implies that |Sv | = n 1 dH (v) dH (u) = k

and for each x Sv , dH (x) dH (u) = k. So, there are at least k vertices in H (elements of Sv )
with degrees at most k.

Also, for any w Su , note that the choice of the pair u, v implies that dH (w) dH (v)

n 1 dH (u) = n 1 k < n k. Moreover, |Su | = n 1 k. Further, the condition

dH (u) + dH (v) n 1, dH (v) dH (u) = k and u

/ Su implies that dH (u) n 1 dH (v)
n 1 k < n k. So, there are n k vertices in H with degrees less than n k.

Therefore, if d1 dn are the vertex degrees of H, then we observe that there exists a

k < n/2 for which dk k and dnk < n k. As k < n/2 and di di , we get a contradiction.

Exercise 8.7.12. Complete an alternate proof of Theorem 8.7.11. Let R denote the property:
R : If dk k then dnk n k, for each k < n/2.

We know that G has this property.

1. Let e be an edge not in G. Show that G + e also has the property. What about the closure
H := C(G) of G?
2. Assume that max{d(u) + d(v) : u, v H are not adjacent} n 2. Let e be an edge not
in H. Does H + e have property R? Is C(H + e) = H + e?

3. In view of the previous observations assume that G is an edge maximal graph with property
R which is not Hamiltonian. Do you have C(G) = G? Show that there are some k vertices
having degree at most k and some n k vertices having degree less than n k. Does that
contradict R?

Definition 8.7.13. [Line graph] The line graph H of a graph G is a graph with V (H) = E(G)
and e1 , e2 V (H) are adjacent in H if e1 and e2 share a common vertex/endpoint.

172

CHAPTER 8. GRAPHS

Example 8.7.14. Verify the following:

1. Line graph of C5 is C5 .
2. Line graph of P5 is P4 .
3. Line graph of any graph G contains a complete subgraph of size (G).
Exercise 8.7.15.

1. Let G be a connected Eulerian graph. Then, show that the line graph

of G is Hamiltonian. Is the converse true?

2. What can you say about the clique number of a line graph?
Theorem 8.7.16. A connected graph G is isomorphic to its line graph if and only if G = Cn ,
for some n 3.
Proof. If G is isomorphic to its line graph, then |G| = kGk. Thus, G is a unicyclic graph.

Let [v1 , v2 , . . . , vk , vk+1 = v1 ] form the cycle in G. Then, the line graph of G contains a cycle
P = [v1 v2 , v2 v3 , . . . , vk v1 ]. We now claim that dG (vi ) = 2.

Suppose not and let dG (v1 ) 3. So, there exists a vertex u

/ {v2 , vk } such that u v1 . In

that case, the line graph of G contains the triangle T = [v1 v2 , v1 vk , v1 u] and P 6= T . Thus, the

line graph is not unicyclic, a contradiction.

Exercise 8.7.17.

1. Determine the closure of G.

2. Show that H is not Hamiltonian.

b

b
b

b
b

3. Give a necessary and sufficient condition on m, n N so that Km,n is Hamiltonian.

4. Show that any graph G with |G| 3 and kGk C(n 1, 2) + 2 is Hamiltonian.
5. Show that for any n 3 there is a graph H with kGk = C(n 1, 2) + 1 that is not Hamiltonian. But, prove that all such graphs H admit a Hamiltonian path (a path containing
all vertices of H).

8.8

Bipartite Graphs

Definition 8.8.1. [2-colorable graphs] A graph is 2-colorable if its vertices can be colored with
two colors in a way that adjacent vertices get different colors.
Lemma 8.8.2. Let P and Q be two v-w-paths in G such that length of P is odd and length of
Q is even. Then, G contains an odd cycle.

8.9. MATCHING

173

Proof. Suppose P, Q have an inner vertex x (a vertex other than v, w) in common. Then, one
of P (v, x), P (x, w) has odd length and the other is even, say, P (v, x) is odd. If the lengths of
Q(v, x) and Q(x, w) are both odd then we consider the x-w-paths P (x, w) and Q(x, w), otherwise
we consider the paths P (v, x) and Q(v, x).
In view of the above argument, we may assume that P, Q have no inner vertex in common. In
that case it is clear that P Q is an odd cycle.
Theorem 8.8.3. Let G be a connected graph with at least two vertices. Then, the following
statements are equivalent.
1. G is 2 colorable.
2. G is bipartite.
3. G does not have an odd cycle.
Proof. Part 1 Part 2. Let G be 2-colorable. Let V1 be the set of red vertices and V2 be the

set of blue vertices. Clearly, G is bipartite with partition V1 , V2 .

Part 2 Part 1. Color the vertices in V1 with red color and that of V2 with blue color to get

the required 2 colorability of G.

Part 2 Part 3. Let G be bipartite with partition V1 , V2 . Let v0 V1 and suppose =

v0 v1 v2 vk = v0 is a cycle. It follows that v1 , v3 , v5 V2 . Since, vk V1 , we see that k is

even. Thus, has an even length.

Part 3 Part 2. Suppose that G does not have an odd cycle. Pick any vertex v. Define

V1 = {w | there is a path of even length from v to w} and V2 = {w | there is a path of odd length

from v to w}. Clearly, v V1 . Also, G does not have an odd cycle implies that V1 V2 = . As

G is connected each w is either in V1 or in V2 .

Let x V1 . Then, there is an even path P (v, x) from v to x. If xy E(G), then we have

a v-y-walk of odd length. Deleting all cycles from this walk, we have an odd v-y-path. Thus,
y V2 . Similarly, if x V2 and xy E, then y V1 . Thus, G is bipartite with parts V1 , V2 .
Exercise 8.8.4.

1. There are 15 women and some men in a room. Each man shook hands

of exactly 6 women and each woman shook hands of exactly 8 men. How many men are
there in the room?
2. How do you test whether a graph is bipartite or not?

8.9

Matching

Definition 8.9.1. [Matching in a graph] A matching in a graph G is an independent set of

edges. A maximum matching is a matching with maximum number of edges. A vertex v is
saturated by a matching M if there is an edge e M incident on v. A matching is a perfect

matching if every vertex is saturated.

Example 8.9.2.

1. In Figure 8.14, M1 = {u1 u2 } is a matching. So, is M2 = {e}, where e is

any edge. The set M3 = {u3 u2 , u4 u7 } is also a matching. The set M4 = {u1 u2 , u4 u5 , u6 u7 }

174

CHAPTER 8. GRAPHS
u1
b

u7

u2
b

u6
b

u3

u4

u5

Figure 8.14: A graph

is also a matching and it is maximum (why?). Can you give another maximum matching?

2. Any non trivial graph G has a maximum matching.

3. Vertices that are saturated for M3 are {u2 , u3 , u4 , u7 }.
4. Any graph with a perfect matching must have even order. The Figure 8.14 cannot have a
perfect matching.
Definition 8.9.3. [Alternating path] Let M be a matching in G. A path P is called M alternating if its edges are alternately from M and from G M . An M -alternating path with

two unmatched vertices as end points is called M -augmenting. Convention: Each path of
length 1 in M is M -alternating.
Example 8.9.4. Consider Figure 8.14.
1. The path [u1 , u2 ] is M1 -alternating. The only path of length 2 which is M1 -alternating is
[u1 , u2 , u3 ].
2. The path [u1 , u2 , u4 , u7 ] is not M3 -alternating. But, [u2 , u3 , u4 , u7 ] is M3 -alternating.
3. The path P = [u1 , u2 , u3 , u4 , u7 , u6 ] is M3 -alternating and M3 -augmenting. This gives us
a way to get a larger (in size) matching M5 using M3 : throw away the even edges of P
from M3 and add the odd edges; i.e., M5 = M3 {u2 u3 , u4 u7 } + {u1 u2 , u3 u4 , u7 u6 }.
Theorem 8.9.5. [Berge, 1957] A matching M is maximum if and only if there is no M augmenting path in G.
Proof. Let M = {u1 v1 , . . . , uk vk } be a maximum matching. If there is an M -augmenting path

P , then (P \ M ) M \ P is a larger matching, a contradiction. Conversely, suppose that M is

not maximum. Let M be a maximum matching. Consider the graph H = (V, M M ). Note

that dH (v) 2, for each vertex in H. Thus, H is a collection of isolated vertices, paths and
cycles. Since a cycle contains equal number of edges of M and M , there is a path P which

contains more number of edges of M than that of M . Then, P is an M -augmenting path. A

contradiction.
Exercise 8.9.6. How do we find a maximum matching in a graph G.
Example 8.9.7. Can we find a matching that saturates all vertices in the graph given below?

8.9. MATCHING

175

1
2
3

Ans: No. Let X be the given graph and take S = {1, 2, 3}. If there is such a matching then

|N (S)| should at least be |S|. But this is not the case with this graph.
Question: What if |N (S)| were at least |S|, for each S X?

Theorem 8.9.8. [Hall, 1935] Let G = (X Y, E) be a bipartite graph. Then, there is a matching

that saturates all vertices in X if and only if for all S X, |N (S)| |S|.

Proof. If there is such a matching, then obviously |S| |N (S)|, for each subset S of X.

Conversely, suppose that |N (S)| |S|, for each S X. Let if possible, M be a maximum
matching that does not saturate x X.

As |N ({x})| |{x}|, there is a y Y such that xy

/ M . Since M cannot be extended, y

must have been matched to some x1 X.

Now consider N ({x, x1 }). It has a vertex y1 which is adjacent to either x or x1 or both by an

edge not in M . Again the condition that M cannot be extended implies that y1 must have

been matched to some x2 X. Continuing as above, we see that this process never stops and
thus, G has infinitely many vertices, which is not true. Hence, M saturates each x X.

Corollary 8.9.9. Let G be a k-regular (k 1) bipartite graph. Then, G has a perfect matching.
Proof. Let X and Y be the two parts. Since G is k-regular |X| = |Y |. Let S X and E be the
P
set of edges with an end vertex in S. Then k|S| = |E|
d(v) = k|N (S)|. Hence, we see
vN (S)

that for each S X, |S| |N (S)| and thus, by Halls theorem the required result follows.

Definition 8.9.10. [Covering of a graph] Let G be a graph. Then, S V (G) is called a

covering of G if each edge has at least one end vertex in S. A minimum covering of G is a
covering of G that has minimum cardinality.
Exercise 8.9.11.

1. Show that for any graph G the size of a minimum covering is n (G).

2. Characterize G in terms of its girth if the size of a minimum covering is |G| 2.

Proposition 8.9.12. Let G be a graph. If M is a matching and K is a covering of G, then
|M | |K|. If |M | = |K|, then M is a maximum matching and K is a minimum covering.

176

CHAPTER 8. GRAPHS

Proof. By definition, the proof of the first statement is trivial. To prove the second statement,
suppose that |M | = |K| and M is not a maximum matching. Let M be a matching of G with

|M | |M |. Then, using the first statement, we have |K| |M | > |M |. Thus, M is maximum.
As each covering must have at least |M | elements, we see that K is a minimum covering.

Exercise 8.9.13. Let G = Kn , n 3. Then, determine

1. the cardinality of a maximum matching?
2. the cardinality of a minimum covering?
Is the converse of Proposition 8.9.12 necessarily true? Can you guess the class of graphs for
which the converse of Proposition 8.9.12 is true?
Theorem 8.9.14. [Konig, 1931] Let M be a maximum matching in a bipartite graph G and
let K be a minimum covering. Then, |M | = |K |.
Proof. Let V = X Y be the bipartition of V and let M be a maximum matching. Let U be
the vertices in X that are not saturated by M and let Z be the set of vertices reachable from U

by an M -alternating path.
Put S = Z X, T = Z Y and K = T (X \ S). Then, U Z X Y . Also, every

vertex in T is saturated by M (as M is a maximum matching) and N (S) = T . Further, a vertex

v X \ S is matched to some vertex y

/ T . Thus, |K| = |T (X \ S)| |M |. If K is not a
covering, then there is an edge xy G with x S and y
/ T , a contradiction to N (S) = T .

Thus, K is a covering and hence, using |K| |M | and Proposition 8.9.12, we get |K| = |M |.
Furthermore, by Proposition 8.9.12, we also see that K is a minimum covering.
Exercise 8.9.15. How many perfect matchings are there in a labeled K2n ?

8.10

Ramsey Numbers

Recall that in any group of 6 or more persons either we see 3 mutual friends or we see 3 mutual
strangers. Expressed using graphs it reads let G = (V, E) be a graph with |V | 6. Then, either

K3 G or K 3 G.

Definition 8.10.1. [Ramsey number] The Ramsey number r(m, n) is the smallest natural
number k such that any graph G on k vertices either has a Km or a K n as its subgraph.
Example 8.10.2. As C5 does not have K3 or K 3 as its subgraph, r(3, 3) > 5. But, using
the first paragraph of this section, we get r(3, 3) 6 and hence, r(3, 3) = 6. It is known that

r(3, 4) = 9 (see the text by Harary for a table).

Proposition 8.10.3. Let G be a graph on 9 vertices. Then, either K4 G or K 3 G.

Proof. Assume that |V | = 9. Then, we need to consider three cases.

Case I. There is a vertex a with d(v) 4. Then, |N (a)c | 4. If all vertices in N (a)c are

pairwise adjacent, then K4 G. Otherwise, there are two nonadjacent vertices, say b, c N (a)c .

In that case a, b, c induces the graph K 3 .

8.11. DEGREE SEQUENCE

177

Case II. There is a vertex a with d(a) 6. If hN (a)i has a K 3 , we are done. Otherwise,

r(3, 3) = 6 implies that hN (a)i has a K3 with vertices, say, b, c, d. In that case a, b, c, d induces

the graph K4 .

Case III. Each vertex has degree 5. This case is not possible as

d(v) should be even.

Exercise 8.10.4. Can you draw a graph on 8 vertices which does not have K3 , K 4 in it?
Theorem 8.10.5. [Erdos & Szekeres, 1935] Let m, n N. Then, r(m, n) r(m 1, n) +
r(m, n 1).

Proof. Let p = r(m 1, n) and q = r(m, n 1). Now, take any graph G on p + q vertices and

take a vertex a. If d(a) p, then hN (a)i has either a subgraph Km1 (and Km1 together

with a gives Km ) or a subgraph Kn . Otherwise, |N (a)c | q. In this case, hN (a)c i has either a
subgraph Km or a subgraph K n1 (K n1 together with a gives K n ).

8.11

Degree Sequence

Definition 8.11.1. [Degree sequence] The degree sequence of a graph of order n is the tuple
(d1 , . . . , dn ) where d1 dn . A nondecreasing sequence d = (d1 , . . . , dn ) of nonnegative

integers is graphic if there is a graph whose degree sequence is d.

Exercise 8.11.2. Show that (1, 1, 3, 3) is not graphic.

Theorem 8.11.3. Fix n 1 and the natural numbers d1 dn . Then, d = (d1 , . . . , dn ) is

P
the degree sequence of a tree on n vertices if and only if
di = 2n 2.

Proof. If d = (d1 , . . . , dn ) is the degree sequence of a tree on n vertices then

2(n 1) = 2n 2.

Conversely, let d1 dn be a sequence of natural numbers with

di = 2|E(T )| =

di = 2n 2. We

use induction to show that d = (d1 , . . . , dn ) is the degree sequence of a tree on n vertices. For

n = 1, 2, the result is trivial. Let the result be true for all n < k and let d1 dk , k > 2, be
P
P
natural numbers with
di = 2k 2. Since,
di = 2k 2, we must have d1 = 1 and dk > 1.
Then, we note that d2 = d2 , , dk1 = dk1 and dk = dk 1 are natural numbers such that
P
di = 2(k 1) 2. Hence, by induction hypothesis, there is a tree T on vertices 2, , k 1, k

with degrees di s. Now, introduce a new vertex 1 and add the edge {1, k} to get a tree T that

has the required degree sequence.

Theorem 8.11.4. [Havel-Hakimi, 1962] The degree sequence d = (d1 , . . . , dn ) is graphic if and
only if the sequence d1 , d2 , . . . , dndn 1 , dndn 1, . . . , dn1 1 is graphic.
Proof. If the later sequence is graphic then we introduce a new vertex and make it adjacent to
the vertices whose degrees are dndn 1, . . . , dn1 1. Hence, the sequence d = (d1 , . . . , dn )

is graphic. Now, assume that d is graphic and G is a graph with degree sequence d. Let n be
a vertex with dn = k and suppose that the vertices 1, . . . , k are adjacent to n in G. Also, let
deg(1) be the minimum among deg(1), . . . , deg(k). If deg(1) deg(k + 1), . . . , deg(n 1), then

178

CHAPTER 8. GRAPHS

G n is the required graph. So, let deg(k + 1) > deg(1). Then, k + 1 has a neighbor v 6= 1, n

with v 6 1. Now, consider the graph G = G {k + 1, v} + {n, k + 1} + {1, v} {1, n}. Then, G

a also has d as its degree sequence with a

average increases by at least

1
k.

better degree of neighbors. Note that the

Obviously the average cannot go beyond n 1. Thus, repeating

the above procedure a finite number of times, the required result follows.
Exercise 8.11.5.

1. How many different degree sequences are possible on a graph with 5

vertices? List all the degree sequences and draw a graph for each one. (Include connected
and disconnected graphs.)
2. Which of the sequences below are graphic? Draw the graph or supply an argument.
(a) (2, 2, 3, 4, 4, 5)
(b) (1, 2, 2, 3, 3, 4)
(c) (22 , 36 , 42 ) = (2, 2, 3, 3, 3, 3, 3, 3, 4, 4)
3. If two graphs have the same degree sequence, are they necessarily isomorphic?
4. If two graphs are isomorphic, is it necessary that they have the same degree sequence?

8.12

Planar Graphs

Definition 8.12.1. [Embedding, Planar graph] A graph is said to be embedded on a surface

S when it is drawn on S so that no two edges intersect. A graph is said to be planar if it can
be embedded on the plane. A plane graph is a graph which is embedded on the plane.

K5 -Non-planar

K3,3 -Non-planar

K4

K4 - Planar embedding

Figure 8.15: Planar and non-planar graphs

Example 8.12.2.

1. A tree is embed-able on a plane and when it is embedded we have only

one face, the exterior face.

2. Any cycle Cn , n 3 is planar and any plane representation of Cn has two faces.
3. The planar embedding of K4 is given in Figure 8.15.
4. Draw a planar embedding of K2,3 .
5. Draw a planar embedding of the three dimensional cube.
Definition 8.12.3. [Face of a planar embedding] Consider a planar embedding of a graph
G. The regions on the plane defined by this embedding are called faces/regions of G. The
unbounded face/region is called the exterior face (see Figure 8.16).

8.12. PLANAR GRAPHS

9

14

13
f2

8
15 f4
1

179

10

f1

11

f3

12

f1
f2
10
2

5
6

f6

f4
3

11
12

f3
4

Planar Graph X1

f5

Planar Graph X2

Figure 8.16: Planar graphs with labeled faces to understand the Eulers theorem
The faces of the planar graph X1 and their corresponding edges are listed below.
Face

Corresponding Edges

f1

{9, 8}, {8, 9}, {8, 2}, {2, 1}, {1, 2}, {2, 7}, {7, 2}, {2, 3}, {3, 4}, {4, 6}, {6, 4}, {4, 5},

f2
f3
f4

{5, 4}, {4, 12}, {12, 4}, {4, 11}, {11, 10}, {10, 13}, {13, 14}, {14, 10}, {10, 8}, {8, 9}
{10, 13}, {13, 14}, {14, 10}
{4, 11}, {11, 10}, {10, 4}

{2, 3}, {3, 4}, {4, 10}, {10, 8}, {8, 2}, {2, 15}, {15, 2}

From the table, we observe that each edge of X1 appears in two faces. This can be easily
observed for the faces that dont have pendant vertices (see the faces f2 and f3 ). In faces f1
and f4 , there are a few edges which are incident with a pendant vertex. Observe that the edges
that are incident with a pendant vertex, e.g., the edges {2, 15}, {8, 9} and {1, 2} etc., appear
twice when traversing a particular face. This observation leads to the proof of Eulers theorem
for planar graphs which is the next result.
Theorem 8.12.4. [Euler formula] Let G be a connected plane graph with f as the number of
faces. Then,
|G| kGk + f = 2.

(8.3)

Proof. We use induction on f . Let f = 1. Then, G cannot have a subgraph isomorphic to a

cycle. For if, G has a subgraph isomorphic to a cycle then in any planar embedding of G, f 2.
Therefore, G is a tree and hence |G| kGk + f = n (n 1) + 1 = 2.

So, assume that Equation (8.3) is true for all plane connected graphs having 2 f < n. Now,

let G be a connected planar graph with f = n. Now, choose an edge that is not a cut-edge, say
e. Then, G e is still a connected graph. Also, the edge e is incident with two separate faces
and hence its removal will combine the two faces and thus G e has only n 1 faces. Thus,
|G| kGk + f = |G e| (kG ek + 1) + n = |G e| kG ek + (n 1) = 2
using the induction hypothesis. Hence, the required result follows.
Lemma 8.12.5. Let G be a plane bridgeless graph with kGk 2. Then, 2kGk 3f . Further,
if G has no cycle of length 3 then, 2kGk 4f .

180

CHAPTER 8. GRAPHS

Proof. For each edge put two dots on either side of the edge. The total number of dots is 2kGk.
If G has a cycle then each face has at least three edges. So, the total number of dots is at least
3f . Further, if G does not have a cycle of length 3, then 2kGk 4f .
Theorem 8.12.6. The complete graph K5 and the complete bipartite graph K3,3 are not planar.
Proof. If K5 is planar, then consider a plane representation of it. By Equation (8.3), f = 7.
But, by Lemma 8.12.5, one has 20 = 2kGk 3f = 21, a contradiction.

If K3,3 is planar, then consider a plane representation of it. Note that it does not have a C3 .

Also, by Eulers formula, f = 5. Hence, by Lemma 8.12.5, one has 18 = 2kGk 4f = 20, a
contradiction.

Definition 8.12.7. [Subdivision, homeomorphic] Let G be a graph. Then, a subdivision of

an edge uv in G is obtained by replacing the edge by two edges uw and wv, where w is a new
vertex. Two graphs are said to be homeomorphic if they can be obtained from the same graph
by a sequence of subdivisions.
For example, for each m, n N, the paths Pn and Pm are homeomorphic. Similarly, all the

cyclic graphs are homeomorphic to the cycle C3 if our study is over simple graphs. In general,

one can say that all cyclic graphs are homeomorphic to the graph G = (V, E), where V = {v}

and E = {e, e} (i.e., a graph having exactly one vertex and a loop). Also, note that if two graphs

are isomorphic then they are also homeomorphic. Figure 8.17 gives examples of homeomorphic
graphs that are different from a path or a cycle.

Figure 8.17: Homeomorphic graphs

Theorem 8.12.8. [Kuratowski, 1930] A graph is planar if and only if it has no subgraph
homeomorphic K5 or K3,3 .
Proof. Omitted.
We have the following observations that directly follow from Kuratowski theorem.
Remark 8.12.9.

1. Among all simple connected non-planar graphs

(a) the complete graph K5 has minimum number of vertices.

(b) the complete bipartite graph K3,3 has minimum number of edges.
2. If Y is a non-planar subgraph of a graph X then X is also non-planar.
Definition 8.12.10. [Blocks of a graph] Let G be a graph. Define a relation on the edges of G
by e1 e2 if either e1 = e2 or there is a cycle containing both these edges. Note that this is an

8.12. PLANAR GRAPHS

181

equivalence relation. Let Ei be the equivalence class containing the edge ei . Also, let Vi denote
the endpoints of the edges in Ei . Then, the induced subgraphs hVi i are called the blocks of G.
Proposition 8.12.11. A graph G is planar if and only if each of its blocks are planar.
Proof. Omitted.
Definition 8.12.12. [Maximal planar] A graph is called maximal planar if it is planar and
addition of any more edges results in a non-planar graph. A maximal plane graph is necessarily
connected.
Proposition 8.12.13. If G is a maximal planar graph with m edges and n 3 vertices, then
every face is a triangle and m = 3n 6.

Proof. Suppose there is a face, say f , described by the cycle [u1 , . . . , uk , u1 ], k 4. Then, we

can take a curve joining the vertices u1 and u3 lying totally inside the region f , so that G + u1 u3

is planar. This contradicts the fact that G is maximal planar. Thus, each face is a triangle. It
follows that 2m = 3f . As n m + f = 2, we have 2m = 3f = 3(2 n + m) or m = 3n 6.
Exercise 8.12.14.

1. Suppose that G is a plane graph with n vertices such that each face is

a 4-cycle. What is the number of edges in G?

2. Show that the Petersen graph has a subgraph homeomorphic to K3,3 .
3. Show that a plane graph on n 3 vertices can have at most 2n 5 bounded faces.
4. Is it necessary that a plane graph G should contain a vertex of degree less than 5?
5. Let G be a plane graph on n vertices, m edges, f faces and k components. Prove by
induction that n m + f = k + 1.
6. If G is a plane graph without 3-cycles, then show that (G) 3.
7. Show that any plane graph on n 4 vertices has at least four vertices of degree at most
five.

8. Produce a planar embedding of the graph G that appears in Figure 8.18.

7

2
4

Figure 8.18: A graph on 8 vertices

182

8.13

CHAPTER 8. GRAPHS

Vertex Coloring

Definition 8.13.1. [k-colorable] A graph G is said to be k-colorable if the vertices can be

assigned k colors in such a way that adjacent vertices get different colors. The chromatic
number of G, denoted (G), is the minimum k such that G is k-colorable.
Theorem 8.13.2. For every graph G, (G) (G) + 1.
Proof. If |G| = 1, the statement is trivial. Assume that the result is true for |G| = n and let G
be a graph on n + 1 vertices. Let H = G 1. As H is ((G) + 1)-colorable and d(1) (G),
the vertex 1 can be given a color other than its neighbors.

Theorem 8.13.3. [Brooks, 1941] Every non complete graph which is not an odd cycle has
(G) (G).
Theorem 8.13.4. [5-color Theorem] Every Planar graph is 5-colorable.
Proof. Let G be a minimal planar graph on n 6 vertices and m edges, such that G is not
5-colorable. Then, by Proposition 8.12.13, m 3n 6. So, n(G) 2m 6n 12 and hence,

(G) 2m/n 5. Let v be a vertex of degree 5. Note that by the minimality of G, G v is

5-colorable. If neighbors of v use at most 4 colors, then v can be colored with the 5-th color
to get a 5-coloring of G. Else, take a planar embedding in which the neighbors v1 , . . . , v5 of v
appear in clockwise order.

Let H = G[Vi Vj ] be the graph spanned by the vertices colored i or j. If vi and vj are in

different connected components of H, then we can swap colors i and j in a component that

contains vi , so that the vertices v1 , . . . , v5 use only 4 colors. Thus, as above, in this case the
graph G is 5-colorable. Otherwise, there is a 1, 3-colored path between v1 and v3 and similarly,
a 2, 4-colored path between v2 and v4 . But this is not possible as the graph G is planar. Hence,
every planar graph is 5-colorable.

8.14

Adjacency Matrix

Definition 8.14.1. [adjacency matrix] Let G = (V, E) be a simple (undirected) graph on

vertices 1, . . . , n. Then, the adjacency matrix A(G) of G (or simply A) is described by
aij =

1 if {i, j} E,
0 otherwise.

Let H be the graph obtained by relabeling the vertices of G. Then, note that A(H) =
S 1 A(G)S, for some permutation matrix S (recall that for a permutation matrix S t = S 1 ).
Hence, we talk of the adjacency matrix of a graph and do not worry about the labeling of the
vertices of G.
will give an adjacency matrix, say B. But, the

8.15. MORE EXERCISES

183

Example 8.14.2. The adjacency matrices of the 4-cycle C4 and the path P4 on 4 vertices are
given below.

0 1 0 1

1 0 1 0
A(C4 ) =
0 1 0 1

1 0 1 0

0 1 0 0

, A(P4 ) = 1 0 1 0

0 1 0 1

0 0 1 0

Exercise 8.14.3. A"graph G is# not connected if and only if there exists a permutation matrix
A11 0
, for some matrices A11 and A22 .
P such that A(G) =
0 A22
Theorem 8.14.4. The (i, j) entry of B = A(G)k is the number of i-j-walks of length k.
Proof. Note that by the definition of matrix product
bij =

i1 ,...,ik1

aii1 ai1 i2 aik1 ik .

Thus, bij = r if and only if we have r sequences i1 , . . . , ik1 with aii1 = = aik1 ik = 1. That
is, bij = r if and only if we have r walks of length k between i and j.


Theorem 8.14.5. Let G be a graph G of order n. Then, G is connected if and only if I +
n1
A(G)
is entrywise positive.
n1
Proof. Put B = I + A and let G be connected. If P is an i-j-path of length n 1, then Bij

n1
Aij
1. If P = [i, i1 , . . . , ik = j] is an i-j-path of length k < n1, then bii . . . bii bii1 . . . bik1 j =

n1
> 0.
1, where bii is used n 1 k times. Thus, Bij

n1
Conversely, let Bij
> 0. Then, the corresponding summand bii1 . . . bin1 j is positive. By

throwing out entries of the form bii , for 1 i n, from this expression, we have an expression

which corresponds to an i-j-path of length at most n 1. As B n1 is entrywise positive, it

follows that G is connected.

Do you want to put Vertex-edge incidence matrix or the {1, 0, 1}incidence matrix? If so, which results other than Matrix Tree
Theorem should be stated. We need not give the proofs.
Definition 8.14.6. [Vertex-edge incidence matrix] The vertex-edge incidence matrix M
of G is a |G| kGk matrix whose (i, j)-entry is described by
(
1 if edge ej is incident on vi ,
mij =
0 else.

8.15

More Exercises

Exercise 8.15.1.

1. Can there be a graph in which the size of a minimum covering is |G|?

2. Characterize G if the size of a minimum covering is |G| 1.

3. What relationship is there between the size of a minimum covering and (G)?

184

CHAPTER 8. GRAPHS

4. Is it necessary that a plane graph G should contain a vertex of degree at most 5?

5. Is K5 e planar, where e is any edge?
6. Is K3,3 e planar, where e is any edge?
7. Is it true that any group of 7 persons there are 3 mutual friends or 4 mutual strangers?
8. Prove/disprove: A two colorable graph is necessarily planar.
9. Draw the tree on the vertex set [12] whose Pr
ufer code is 9954449795.
10. A game of thief is played in the following way: There is a coin. There are n participants.
One participant takes the coin and passes it to whoever he/she wishes to. Whoever has
the coin must pass to somebody (other than the person from whom he/she received it) as
quickly as possible. When the music stops, the person with the coin is the thief .
My class students (there are 115 in total) were playing it. There were 2009 passes when
the music stopped. I guarantee that the person who started it is not the thief. How?

Not able to understand

11. How many chordal graphs are there on the vertex set [4]?
12. Count with diameter: how many nonisomorphic trees are there of order 7?
13. List the automorphisms of the following graph.
4
b

5
b

6
b

Bibliography
[1] G. Agnarson and R. Greenlaw, Graph Theory: Modelling, Applications and Algorithm,
Pearson Education.
[2] R. B. Bapat, Graphs and Matrices, Hindustan Book Agency, New Delhi, 2010.
[3] J. Cofman, Catalan Numbers for the Classroom?, Elem. Math., 52 (1997), 108 - 117.
[4] D. M. Cvetkovic, Michael Doob and Horst Sachs, Spectra of Graphs: theory and applications,
Academic Press, New York, 1980.
[5] D. I. A. Cohen, Basic Techniques of Combinatorial Theory, John Wiley and Sons, New
York, 1978.
[6] William Dunham, Euler: The Master of Us All, Published and Distributed by The Mathematical Association of America, 1999.
[7] F. Harary, Graph Theory, Addison-Wesley Publishing Company, 1969.
[8] G. E. Martin, Counting: The Art of Enumerative Combinatorics, Undergraduate Texts in
Mathematics, Springer, 2001.
[9] R. Merris, Combinatorics, 2th edition, Wiley-Interscience, 2003.
[10] J. Riordan, Introduction to Combinatorial Analysis, John Wiley and Sons, New York, 1958.
[11] R. P. Stanley, Enumerative Combinatorics, vol. 2, Cambridge University Press, 1999.
[12] H. S. Wilf, Generatingfunctionology, Academic Press, 1990.

185

Index
(n, m) graph, 146

d(v), 146

C(n, r), 93

dG (v), 146

Cn , 111, 147

fA , 8

E, 146

k-colorable, 182

E(G), 146

k-connected, 164

G v, 149

k-factor, 149

M -alternating, 174

k-place predicate, 71

M -augmenting, 174

k-regular, 149

N (v), 146

k-th difference, 132

NG (v), 146

l-edge connected, 165

P (n, r), 91

n-set, 90

P (n; n1 , . . . , nk ), 93

p implies q, 59

Pn , 146

r-combination, 93

S(n, r), 106

r-permutation, 90

V , 146

r-sequence, 90

V (G), 146

r(m, n), 176

(G), 149

s(n, k), 109

(G), 156

u-v walk, 151

B30DGB30D, 146

x-bound part, 72

hAU B, 149

xn, 109

(G), 149

acyclic, 153

dom f , 7

addition rule, 89

B , 29

adequate, 63

(G), 164

adjacency matrix, 182

(G), 165

adjacent, 146

A, 44

algebraic, 45

n , 109

algebraic expansion, 96

n (k), 109

anti-chain, 46

rng f , 7

antisymmetric, 46

B , 29

argument, 57

{Ax }xA , 29

assignment, 58

{f1 , . . . , fn } g, 67

atom, 85

a + b, 55

atomic formula, 58, 72

ab, 55

atomic variable, 58
186

INDEX

187

automorphism, 156

composition, 31

axiom of choice, 51

conclusion, 67
congruent, 21

basic solution, 134

conjugate, 128

Bell Numbers, 139

conjunction/and, 58

bijection, 9

conjunctive normal form, 65

bipartite, 168

connected, 154

blocks, 181

connected component, 154

Boolean algebra, 83

connected permutation, 131

Boolean homomorphism, 86

connectives, 60

Boolean isomorphism, 86

contradiction (F ), 61

bound, 72

contrapositive, 60

bounded lattice, 81

converse, 60

bridge, 158

countable, 43

Cardinal numbers, 44
cartesian product, 5, 151
Catalan number, 111
Cauchy product, 122
center, 152
chain, 46

covering, 175
cubic, 149
cut edge, 158
cut vertex, 157
cycle graph, 147
cycle/circuit, 152

characteristic equation, 133

degree, 146

chord, 153

degree sequence, 177

chordal, 153

derangement, 120

chromatic number, 182

diameter, 152

circular permutation, 97

dictionary, 47

clique, 154

difference equation, 132

clique number, 154

direct product, 79

closed, 152

disconnected, 154

closure, 170

disjoint, 6

co-prime, 17

disjoint union, 150

codomain set, 7

disjunction/or, 58

coin problem/ money changing problem, 130

disjunctive normal form, 65

commutative ring with unity, 25

distance, 152

comparable, 46

distributive lattice, 77

complement, 5, 81, 150

divide, 17

complemented, 81

divisor, 17

complete, 81, 146

domain of f , 7

complete bipartite graph, 147

domain set, 7

component, 154

dual, 66

composite, 19

Durfee square, 131

188

INDEX
edge connectivity, 165

functionally complete, 63

edge density, 154

Fundamental theorem of arithmetic, 20

edge set, 145

edges, 145
egf, 123
embedded, 178
end vertices, 145
end vertices of P , 152
enumeration, 43

general graph, 145

girth, 152
graph invariant, 156
graphic, 177
greatest common divisor, 17
greatest lower bound (glb), 49

equivalence class, 33

Hamel basis, 53

equivalence relation, 31

Hamiltonian, 169

equivalent, 9, 61, 73

Hasse diagram, 48

Euclids Algorithm, 17

height, 46

Euclids algorithm, 18

homeomorphic, 180

Euclids lemma, 19

homogeneous recurrence relation, 132

Eulers totient function, 120

hypergraph, 146

Eulerian, 166

hypothesis/premise, 67

Eulerian tour, 166

exponential generating function, 123
exterior face, 178

identity relation, 31
image, 7
incident, 145

faces/regions, 178

independence number, 155

falling factorial, 91, 109

independent, 146

family of finite character, 52

induced, 149

family of sets, 29

induced partial order, 52, 85

Ferrers Diagram, 128

infinite, 10

Fibonacci sequence, 132

infinite cardinal numbers, 45

field, 23

initial condition, 132

finite, 10

initial segment of a, 50

first difference, 132

injection, 8

forest, 157

internal vertices, 152

Formal power series, 122

interpretation, 72

differentiation, 125

interpretation of f , 72

integration, 125

intersection, 6, 29, 150

formal power series, 122

inverse, 7, 83

free, 72

isolated, 146

free Boolean algebra, 84

isomorphic, 155

Frobenius number, 130

Function
Identity, 36
function, 7

join, 151
lattice, 77
lattice homomorphism, 80

INDEX
lattice isomorphism, 80
lattice of n-tuples of 0 and 1, 79
lattice path, 110
least common multiple, 20
least upper bound (lub), 49
length, 151
lexicographic, 47
LHRRCC, 132
line graph, 171
linear/complete/total order, 46
linearly ordered set, 46
literal, 65
LNHRRCC, 132
logical conclusion, 67
loop, 145
lower bound, 49

189
ordinary generating function, 123
partial order, 46
partially ordered set, 46
partition, 105
partition of n into k parts, 109
partition of S, 105
path, 151
path graph, 146
pendant, 146, 159
perfect matching, 173
permutation, 90
Petersen graph, 148
Pigeonhole principle (PHP), 115
planar, 178
plane graph, 178
Polish notation, 67

matching, 173

poset, 46

maximal, 49, 154

Pr
ufer code, 161

maximal planar, 181

pre-image, 7

maximum, 49

prime, 19

maximum matching, 173

principal connective, 66

minimal, 49, 154

principle of duality, 84

minimum, 49

principle of transfinite induction, 50

minimum covering, 175

product, 51

multigraph, 146

propositional function, 71

multiplication rule, 89

pseudograph, 145

multiplicative, 121
multiset, 92
negation, 58
neighbor, 146
Newtons Identity, 95
nonhomogeneous recurrence relation, 132

quantifiers, 71
quotient, 16
radius, 152
Ramsey number, 176
range of f , 7
reciprocal, 124

ogf, 123

recurrence relation, 132

one-one, 8

reflexive, 31

onto, 9

relation, 7

orbit, 98

relation on X, 31

orbit size, 98

relatively prime, 17

order, 146

remainder, 16

order of operations, 67

residue, 21

190

INDEX
restriction of f to A, 8
rising factorial, 109
rotation, 98

symmetric difference, 6
tautology (T ), 61
trail, 151

satisfiable, 67

transcendental, 45

saturated by a matching, 173

transitive, 31

scope, 72

tree, 157

self conjugate, 128

trivial graph, 146

self-complementary, 156

truth function, 61

separating set, 164

truth table, 59

Set

truth value, 58
Cartesian Product, 5
Complement, 5
Composition of Relations, 31
Difference, 6
Equality, 5
Equivalence Class, 33
Equivalence Relation, 31
Identity Relation, 31

uncountable, 43
union, 6, 29, 150
unit, 19
unity, 19
universe of discourses, 71
upper bound, 49
usual partial order, 46

Intersection, 6

valid, 73

Power Set, 6

vertex connectivity, 164

Subset, 5

vertex set, 145

Symmetric Difference, 6

vertex-edge incidence matrix, 183

Union, 6

vertices, 145

set difference, 6
simple graph, 146
single valued, 7
solution, 132
spanning subgraph, 149
Squarefree, 26
squarefree, 88
standard representative, 21
Stirling number of the first kind, 109
Stirling numbers of the second kind, 106
Stirlings Identity, 138
subdivision, 180
subgraph, 149
subset, 5
substitution instance, 62
surjection, 9
symmetric, 31

well formed formulae (wff), 60

well order, 50
width, 46
Wilsons Theorem, 24
word expansion, 96