21MAB302T -Discrete Mathematics

UNIT -5
Graph Theory
 Birth of Graph Theory
The Konigsberg Bridge Problem

o The city of Konigsberg was located on the Pregel river

in Prussia.

o The city occupied 2 islands plus areas on both banks.

o These regions were linked by 7 bridges.

o The citizens wondered whether they could leave

home, cross every bridge exactly once, and return home.

Konigsberg Bridge Problem
 Euler represented the seven bridges as seven edges and the four

places as four vertices.

 He concluded that it is not possible to traverse all the bridges

exactly once from start to end of any town which made Euler to

introduce the concept of Eulerian path and Eulerian circuit

which we will discuss in the later section.

 The following are the definitions and theorems in graph theory.

Graph: A graph G=(V, E) consists of a nonempty set V called the set of vertices (or nodes or
points) and a set E of ordered or unordered pairs of elements of V, called edges (or lines)
such that there is a mapping from E to V.
Example: Draw a graph G=(V, E) given by set with V={a,b,c,d} and E={(a, b), (a, c), (b, c),
(c, d)}.
Directed Graph and undirected Graph: If is graph G=(V, E), each
edge e∈E is associated with an ordered pair of vertices, then G is
called a directed graph or digraph.

If each edge is associated with an unordered pair of vertices then G

is called an undirected graph.
Isolated vertex: A vertex of a graph which is not adjacent to any other vertex.
Example: Let G= (V, E) with V={v 1 , v 2 , v 3 , v 4 , v 5 } and E={(v 1 , v 2 ), (v 2 , v 3 ), (v 3 , v 4 ), (v4,
v 1 )}. Here v 5 is an isolated vertex.

Null graph: A graph containing only isolated vertices is called null graph.

Example:

Here v 1 , v 2 , v 3 , v 4 and v 5 are all isolated vertex.

Loop: An edge of a graph that joins a vertex to itself is called a loop.

Here v2 is having a loop

Parallel edges: If in a directed graph or undirected graph certain pairs of vertices are joined
by more than one edge, such edges are called paralled edge. A graph which contains some
parallel edges is called a multigraph.

Here B and C are having parallel edges

Simple Graph: A graph in which there is only one edge between a pair of vertices.

Pseudo graph: A graph in which loops and parallel edges are allowed.
W e i g h t e d grap h : Graphs in whi ch a n u m b e r (weight) is assigned to each edge.

De g r e e of a Vertex : T h e degree of a vertex in an undirect ed graph is the n u m b e r o f edges

incident wit h it, wi t h the ex ception that a l oop at a vertex contributes twice to the degree of
that vertex. T h e degree of a vertex v is denot ed b y d e g (v).Clearl y the degree o f an isolated
vertex is zero.

Here d e g (v 1 )=3, deg (v 2 )=3, deg (v 3 )=2, d e g (v 4 )=2, deg (v 5 )=2 and d e g (v 6 )=2.
Pendant vertex: If the degree of a vertex is one,
it is called a pendant vertex.

Here v7,v8 and v9 are called pendant vertices.

Theorem: (The Handshaking Theorem)

If G= (V, E) is an undirected graph with e edges, then ∑𝑖 deg(𝑣𝑖 ) = 2𝑒. [ The sum of
the degrees of all vertices of an undirected graph is twice the number of edges of the graph and
hence even].

Proof: Since every edge is incident with exactly two vertices, every edge contributes 2 to

the sum of the degree of the vertices.

Therefore all the e edges contribute (2e) to the sum of the degrees of the vertices.

Hence the proof.

Theorem: The number of vertices of odd degree in an undirected graph

is even.

Proof:

Let G=(V,E) be the undirected graph. Let V1 and V2 be the set

of vertices of G of even and odd degrees respectively.

Then by handshaking theorem,

2e   deg(v i )   deg(vj ) - (1)

v i V1 v j V 2
Since each deg(vi) is even, deg(v )i is even
vi V1

As left hand side of the equation – (1) is even,

we get  deg(v ) is even.

v j V2
j

 the number of vertices of odd degree is even.

The out degree and in degree of a vertex:

In a directed graph G, the out degree of a vertex v of G, denoted by 𝑑𝑒𝑔+𝐺 𝑣 is the

number of edges beginning at v and the in degree of v, denoted by 𝑑𝑒𝑔−𝐺 𝑣 is the
number of edges ending at v.

𝑑𝑒𝑔+
𝐺 𝑎 =1 𝑑𝑒𝑔𝐺− 𝑎 =2

𝑑𝑒𝑔+
𝐺 𝑏 =5 𝑑𝑒𝑔𝐺− 𝑏 =1

𝑑𝑒𝑔+
𝐺 𝑐 =1 𝑑𝑒𝑔𝐺− 𝑐 =2

𝑑𝑒𝑔+
𝐺 𝑑 =1 𝑑𝑒𝑔𝐺− 𝑑 =3

Theorem :

If G=(V,E) be a directed graph with 𝑒 edges then ∑𝑣∈𝑉 𝑑𝑒𝑔+𝐺 𝑣 = ∑𝑣∈𝑉 𝑑𝑒𝑔−
𝐺 𝑣 =e.
Verify handshaking theorem for the following digraph

Sum of the in-degrees + sum of the out-degrees= 2e

Sum of the in-degrees=8
Sum of the out-degrees=8
The number of edges (e) =8
5.1.2 Special Simple Graphs:

Complete Graph: A simple graph, in which there is exactly one edge

between each pair of distinct vertices, is called a complete graph.The
complete graph on n vertices is denoted by Kn.

𝑛 (𝑛−1)
Note: The number of edges in Kn is nC2 or . Hence, the maximum
2
number of edges in a simple graph with n vertices is 𝑛(𝑛−1)
2
.
Regular graph: If every vertex of a simple graph has the same
degree, then the graph is called a regular graph. If every vertex in a
regular graph has degree n, then the graph is called n-regular.
Bipartite graph: If the vertex set V of a simple graph G= (V, E)
can be partitioned into two subsets V 1 and V2 such that every
edge of G connects a vertex in V1 and a vertex in V2, then G is
called bipartite graph.
Complete bipartite graph: If each vertex of V 1 is
connected with every vertex of V 2 by an edge, then G is
called a complete bipartite graph. If V 1 contains m vertices
and V 2 contains n vertices, then the complete bipartite
graph is denoted by K m,n .
 𝑛2
Theorem: The number of edges in a bipartite graph with n vertices is at most ( ).
4

Proof: Let the vertex set be partitioned into the two subsets V1 and V2. Let V1 contain x
vertices. Then V2 contains (n-x) vertices. The largest number of edges of the graph can be
obtained, when each of the x vertices in V1 is connected to each of the (n-x) vertices in V2.
Therefore, the largest number of edges f(x)= x(n-x), is a function of x. Now we have to find
the value of x for which f(x) is maximum. By calculus, 𝑓′ (x)=n-2x and 𝑓" (x)=-2.
𝑛 𝑛
whenx=( ) and 𝑓"( ) < 0.
2 2
𝑛
Hence, f(x) is maximum, when x= .
2
𝑛
Therefore the maximumnumber of edges required= 𝑓 = 𝑛2/4 .
2
Isomorphism: The simple graphs G1=(V1,E1) and G2=(V2,E2) are

isomorphic if there exists a one to one and onto function f from V1 to

V2 with the property that a and b are adjacent in G1, if and only if f(a)

and f(b) are adjacent in G2 for all a, b V1, such a function f is called

an isomorphism.

Two simple graphs that are not isomorphic are called

nonisomorphic
Example: Show that the graphs G and H are not isomorphic.

In G, |V|=5 and |E|=6. In H, |V|=5 and |E|=7. The number of edges in both the graphs G and
H are not equal. Therefore the graphs G and H are not isomorphic.
Problems for practice:
1. Show that the graphs G and H are not isomorphic.
Problem 2: Show that the following graphs G and H are
isomorphic.

The graphs G and H are isomorphic

5.3.1 Paths and Cycles:
Walk: A walk in G is a finite non-null sequence W=v0e1v1e2v2…ekvk, whose terms are
alternatively vertices and edges, such that for 1≤i≤k, the ends of ei are vi-1 and vi. W is a walk
from v0 to vk or a (vo,vk)-walk.The vertices v0 and vk are called the origin and terminus of W
respectively and v1,v2,…,vk-1 its internal vertices. The integer k is the length of W.
If the edges e1,e2,e3,…,ek of a walk W are distinct, W is called a trail.
If the vertices v0,v1,v2,…,vk are distinct, W is called a path.

Walk –uavfyfvgyhwbv; Trail-wcxdyhwbvgy and Path-xcwhyeuav

Cycle: A walk is closed if it has positive length and its
origin and terminus are same. A closed trail whose origin
and internal vertices are distinct is a cycle.

Closed trail: ucvhxgwfwdvbu; Cycle: xaubvhx

A graph is said to be connected if there is a path between
every pair of vertices.
Theorem: If a graph G (either connected or not) has exactly two vertices of odd degree then
there is a path joining these two vertices.
Proof:

Case (i) Let G be connected.

Let v1 and v2 be the only vertices of G which are of odd degree. But we have already
proved that the number of odd vertices is even. Clearly there is a path connecting v1 and v2,
since G is connected.

Case (ii) Let G be disconnected.

Then the components of G are connected. Hence, v1 and v2 should belong to the same
component of G. Hence, there is a path between v1 and v2.
Theorem: The maximum number of edges in a simple connected

graph G with n vertices and k components is (n  k)(n  k 1) .

2
Proof:

Let the number of vertices in the ith component of G be ni (i 1).

k
Then n1+n2+…+nk= n or n i  n - (1)
i1
k
Hence
 (n 1)  (n
i 1 1)  (n2 1) ... (nk 1)
i1
k
 ni  k
i1

nk
Squaring on both sides we get,

(n 1) 2  (n 1) 2  ...  (n 1) 2  (n  k) 2

1 2 k

n 2  n 2 ... n 2  2n  2n ... 2n  k  (n  k)2

1 2 k 1 2 k

k
  ni  2n  k  (n  k)2
2

i1
k
  ni  (n  k) 2  2n  k
2
- (2)
i1

1
Now the maximum number of edges in the ith component of G  ni (ni 1)
2
 maximum number of edges of G, k


1

2 i1
ni (ni 1)
 1  k 2
  n i  n 
2  i1 


1
2

(n k) 2  2n  k  n 

1
2
(n k) 2  (n  k)
(n  k)(n  k 1)

2

 The maximum number of edges of G

(n  k)(n  k 1)
2
Eulerian Path: A path of graph G is called an Eulerian path, if it includes

each edge of G exactly once

Eulerian Circuit: A circuit of a graph G is called an Eulerian circuit, if it

include each edge of G exactly once.

Eulerian Graph: A graph containing an Eulerian circuit is called an Euler

graph
 EULERIAN CIRCUIT
deg(A)=2

deg(B)=2

deg(C) =4

deg(D)=2

deg(E)=2

A-B-C-E-D-C-A

Theorem: A connected graph contains an Eulerian circuit if and only

if each of its vertices is of even degree.
 EULERIAN PATH

A-B-C-D-E-F

Theorem: A connected graph contains an Euler path, if and only if it

has exactly two vertices of odd degree.
Find an Euler path or an Euler circuit, if it exists for the following graphs:

G1 G2
Hamiltonian Path: A path of a graph G is called a Hamiltonian path,

if it includes each vertex of G exactly once.

Hamiltonian Circuit: A circuit of a graph G is called a Hamiltonian

circuit, if it includes each vertex of G exactly once, except the starting

and end vertices which appear twice.

Hamiltonian Graph: A graph containing a Hamiltonian circuit is

called a Hamiltonian graph.

Hamiltonian Circuit

a-b-c-d-a
Hamiltonian Path

a-b-c-d
Practice problem:

Give an example of a graph which contains:

i) an Eulerian circuit and a Hamiltonian circuit.

ii) an Eulerian circuit but not a Hamiltonian circuit.

iii) a Hamiltonian circuit and but not an Eulerian circuit.

iv) neither an Eulerian circuit nor a Hamiltonian circuit.

Matrix Representation of Graphs

Adjacency Matrix Representation: If an Undirected Graph G consists of n vertices

then the adjacency matrix of a graph is an n x n matrix A = [aij] and defined by

If there exists an edge between vertex vi and vj, where i is a row and j is a column
then the value of aij=1.

If there is no edge between vertex vi and vj, then value of aij=0.

 Find the adjacency matrix MA of graph G shown in Fig:

Since graph G consist of four vertices. Therefore, the adjacency matrix will be 4 x 4
matrix.

𝑀𝐴 =
 Incidence Matrix Representation: If an Undirected Graph G consists of n
vertices and m edges, then the incidence matrix is an n x m matrix C = [cij] and

defined by
by

Consider the undirected graph G as shown in fig. Find its incidence matrix
MI.

:
 Graph colouring

 A graph has been coloured if colour is assigned to each vertex in such a way that
adjacent vertices have different colours.
 Vertex colouring of a graph G is a mapping from f:V(G) to S.

 The elements of S are called colors.

 The vertices of one color form a color class.

 If |S| = k then it is k-coloring.

 A graph is k colorable if it has proper k-coloring.

 In proper coloring each color class is a stable set.

 Therefore, one may consider k-coloring as a partition of the vertex set of G

into k stable sets which are disjoint.
 Chromatic number

The chromatic number of a graph is the smallest number of colours

with which it can be coloured.

What is the chromatic number of the graph Cn n≥3. Cn denotes a

cycle with n vertices.

In general 2 colors are needed when n is even.

Example: C6 cycle with six vertices only two colors needed red
and blue.
Example: C 5 cycle with five vertices three colors needed.

Here pick a vertex color it green color the next vertex with red,
proceed the same way till you reach the fourth vertex. The fifth
vertex cannot be red or green so we need a third color say purple as
shown above.
What is the chromatic number of Kn?

No two vertices can be assigned the same color because every two
vertices of this graph are adjacent. Therefore the chromatic
number of Kn is n.

Example: K5

Five distinct colors are needed so chromatic number is 5.

Example: Coloring of K3,2 bipartite graph.

Only two colors are needed to color a complete bipartite graph.

Applications of graph coloring
1. Scheduling final exams:
 Vertices represent the courses
 There is an edge between the courses if there is a common
student in the courses the vertices represent.
 Each time slot for the exams may be represented by
different color.
 Scheduling of exams corresponds to coloring of the
associated graph.
Edge colouring:

An edge colouring of G is a mapping from the edge set E(G) to the elements of S.
The edges of one colour form a colour class.

If |S|=k then it is called k-edge colouring (objective is to use minimum number of

colours).

Weisstein, Eric W. "Edge Coloring." From MathWorld--A Wolfram Web Resource. http://mathworld.wolfram.com/EdgeColoring.html
Four color theorem:

The four color theorem, or the four color map theorem, states
that, given any separation of a plane into contigious regions,
producing a figure called a map, no more than four colors are
required to color the regions of the map so that no two adjacent
regions have the same color.

Adjacent means that two regions share a common boundary

curve segment, not merely a corner where three or more regions
meet.
 TREES

A connected graph without any circuits is called a tree. It is a simple graph and has
no loops and no parallel edges.

G1 and G4, G2are trees G3contain circuit so not a tree.

Definition of tree:

A connected graph without any circuits is called a tree. It is a simple

graph and has no loops and no parallel edges.

Definition of forest:

A forest is a graph whose components are all trees.

 PROPERTIES OF TREES

 An undirected graph is a tree, if and only if, there is a unique simple path between
every pair of vertices.

 A tree with n vertices has n-1 edges.

 Any connected graph with n vertices and n-1 edges is called a tree.

 Any circuitless graph with n vertices and n-1 edges is a tree.

Property 1:
An undirected graph is a tree if and only if , there is a unique simple path between every pair of
vertices.
Proof:
(i) Consider a undirected graph T to be the tree under consideration.
Then by the definition of tree it is connected. Hence there is a simple path between
any pair of vertices say, vi and vj.
If possible, let there be two paths between vi and vj and the other from vj to vi. Union
of these two paths would contain a circuit.
But T cannot have a circuit as it is a tree.
Hence there is a unique simple path between every pair of vertices in T.
(ii) Converse:
Given that there is a unique simple path between every pair of vertices in T. We need
to prove that T is a tree.
By the assumption T is connected. If possible let T contain a circuit. That is there is a
pair of vertices vi and vj between which two paths exist which is against our
assumption. Hence T cannot have a circuit and so T is a tree.
Property 2:
If G is a tree with n vertices then G has n-1 edges.

Proof: Proof is by induction.

Basis step:If n=1, the number of vertices is one and the number of edges is zero.
Hence the theorem is true for n=1.

Induction step: For n=2, the number of vertices is 2 and edges is 1. Hence also true for n=2
Assume that the theorem is true for .

Consider a tree with k+1 vertices. Let eij be an edge of the tree connecting vi and vj. Since G
is a tree eij is the only edge of the tree connecting vi and vj.

Hence deleting eij will disconnect G into two components.

Each such component has fewer than k+1 vertices
and each of them are trees. Let the two components
have k1 and k2 vertices respectively.
Then the total number of edges in G will be
k1-1+k2-1+eij = k1-1+k2-1+1
=k1-1+k2
=k+1-1=k.
Illustration using a graph of 5 vertices.

Assume that the theorem is true for k≤5

K1 has 4 vertices and hence no of edges is 3.
K2 has 2 vertices and hence the number of edges is one.
The number of edges in k= k1+k2-1-1+eij= 3+1= 4 that is nothing but number of
vertices 5 minus 1 hence the proof.
Property3:
Any connected graph with n vertices and n-1 edges is a tree.

Proof: Assume that G has a cycle of length p. Then there

are p points and p lines on the cycle and for each of the n-p
points not on the cycle there is an incident lineon a geodesic
to a point of the cycle. Since each such line is different, the
number of edges n-p+p=n>n-1 which is a contradiction.
 SPANNING TREES

If the subgraph T of a connected

graph G is a tree containing all
the vertices of G then T is called
the spanning tree of G.

Example:
Since G has three vertices any
spanning tree of G will also have
three vertices and hence two
edges.
This can be done in 3C2 ways
that is 3 ways as shown in the
figure.
Definition of spanning set: Let G=(V, E) be a connected
undirected graph. A spanning set for G is a subset F of E such that
(V, F) is connected.

Definition of spanning tree: If the subgraph T of a connected

graph G is a tree containing all the vertices of G then T is called the
spanning tree of G.

Definition of branch of tree: Any edge in a spanning tree T of G is

called a branch of T.

Definition of chord: an edge in G that is not in a specific given

spanning tree T is called a chord.
Facts:
1. A graph can have more than one spanning tree.
2. Number of F=number of V-1, obviously.
3. Every connected graph has atleast one spanning tree.
4. With respect to any of the spanning trees, a
connected graph of n vertices and e edges has n-1
tree branches and e-n+1 chords.
5. Rank of G = number of branches in any spanning
tree of G.
6. Nulllity of G=number of chords in G.
 MINIMUM SPANNING TREE
Definition:
If G is a connected weighted graph, the spanning tree of G with the smallest total
weight is called minimum spanning tree of G.
Two popular algorithms for constructing minimum spanning trees is
 Prim’s Algorithm
 Kruskal’s Algorithm
KRUSKAL’S ALGORITHM
1. The edges of the graph are arranged in the order of increasing weights
2. An edge with minimum weight is selected as an edge of the required spanning tree.
3. Those edges with minimum weight that do not form a circuit with already selected
edges are successively added.
4. The procedure terminates once n-1 edges have been selected.
NOTE:
The weight of a minimum spanning tree is unique, different minimum spanning trees
are possible if 2 or more edges have same weight
Problem: 1
Find the minimum spanning tree for the weighted graph shown in the figure by
using Kruskal’s algorithm.
Solution:
To find the minimum spanning tree for the weighted graph shown in the figure
by using Kruskal’s algorithm.

Number of vertices is 6 .
Hence the number of edges to be considered is 5.

Edge Weight Included or not If not included circuit

formed

1-2 1 Yes -
6-3 2 Yes -
5-4 6 Yes -
1-3 9 Yes -
5-6 9 Yes -
 The other spanning trees possible for the same graph are

There are two spanning trees for the above problem.

Their minimum spanning weight is 27.(unique)
Problem: 2
Find the minimum spanning tree for the weighted graph shown in the figure by
using Kruskal’s algorithm.
Solution:
We first arrange the edges in the increasing order of the edges and proceed
as per Kruskal’s algorithm. The number of vertices of the graph is 5. Therefore,
the number of edges to be considered is 4.
. Edge Weight Included If not included
or not circuit formed
AE 2 Yes --
CD 3 Yes --
AC 4 Yes --
CE 4 No A–E–C–A
AB 6 Yes --
BC 6 No A–B–C–A
BE 6 -- --
DE 7 -- --
AD 8 -- --
BD 8 -- --

The edges of the minimum spanning tree are 𝐴𝐸, 𝐶𝐷, 𝐴𝐶, and 𝐴𝐵,
whose length is 15.
Problem: 3
Find the minimum spanning tree for the weighted graph shown in the figure by
using Kruskal’s algorithm.
Solution:
The number of vertices of the graph is 9. Therefore, the number of edges to be
considered is 8.
Edge Weight Included If not included circuit
or not formed
.
EF 1 Yes --
AD 2 Yes --
HI 2 Yes --
BD 3 Yes --
CF 3 Yes --
EH 3 Yes --
BC 4 Yes --
FH 4 No E – F –H – E The required minimum
FI 4 No E–F–I–H–E spanning tree consists of
the edges EF, AD, HI,
GH 4 Yes --
BD, CF, EH, BC, and GH.
AB 5 -- --
BE 5 -- --
The total length of the
BF 6 -- -- minimum spanning
. DG 6 -- -- tree=22
DE 7 -- --
DH 8 -- --
Concepts based on minimum spanning trees (MST):

 The number of edges in MST with n nodes is (n-1).

 The weight of MST of a graph is always unique. However
there may be different ways to get this weight (if there edges
with same weights).
 The weight of MST is sum of weights of edges in MST.
 Maximum path length between two vertices is (n-1) for MST
with n vertices.
 There exists only one path from one vertex to another in
MST.
 Removal of any edge from MST disconnects the graph.
 For a graph having edges with distinct weights, MST is
unique.
Problem: How many minimum spanning trees
are possible using Kruskal’s algorithm for a
given graph?

Solution: If all edges weight are distinct,

minimum spanning tree is unique.
If two edges have same weight, then we have to
consider both possibilities and find possible
minimum spanning trees.