Graph theory and Combinatorics 1

Module - I
Notes

e
Unit- I: Introduction to Graph Theory

in
Introduction:

nl
In the field of mathematics, graph theory plays a vital role in engineering and
computer science. It deals with the study of graphs that concerns with the relationship
among vertices and edges. It is an important subject having many more applications
in computer science, information technology, biomathematics, etc. It is also used

O
an extensively in the circuit connections. Also, it is used to facilitate the study of
algorithms like Kruskal’s algorithm, Prim’s algorithm, Dijkstra’s algorithm etc. There
is a main relationship between topology and graph theory, it is applied in the different

ity
types of topologies like star, bridge, series, parallel and network topologies etc. The
interconnected computers in the network follows from the relationship among the
principles of graph theory. The characterizations of graphs are depending on their
structures.

Definition of a Graph: 1.1

rs
A Graph is said to be a pair of vertices and edges which consisting a set of objects
(V, E) where V={v1 , v2, v3,…., vn} are called vertices and E={e1, e2, e3,….,en }are
ve
called edges . The vertex v1 is called the initial vertex and the vertex vn is called the
terminal vertex. Each edge is associated with an unordered pair of vertices. Let us see
an example of the graph in the figure (1.1)
ni

v1 e1 v2
U

e2 e1
ity

v3 e3 v4

Fig (1.1)
m

In the above figure (1.1), there are four vertices and four edges namely V={v1, v2, v3, v4}
and E={ e1, e2, e3, e4}
)A

Definition: 1.2
A graph having without self loop and parallel edges is said to be a simple graph.
[Equivalently a graph having neither self loop nor parallel edges is said to be a simple
graph]. Let us see an example of simple graph
(c

Amity Directorate of Distance & Online Education

 2 Graph theory and Combinatorics

Notes

e
in
nl
O
Fig (1.2)

The above Fig (1.2) shows the graph is a simple graph which has no self loop and paral-
lel edges.

ity
Definition: 1.3

rs
Any two edges having the same initial vertex and the same end vertex is said to be
Parallel edges. An edge having the same vertex as an initial vertex and end vertex then
it is said be a Self loop. The below example having the graph with self loop and parallel
edges.
ve
e2
v1 e1 v2
v3
e3 e2
ni

e4
e3
v4 v3
v2
U

e4
e1
Fig (1.3)
ity

In the Fig (1.3) the graph consists of five vertices and 7 edges.
v4
1. In this graph V= {v1, v2, v3, v4, v5} and E= {e1, e2, e3, e4, e5, e6, e7}. Here the two edges
3 and 4 are parallel edges , the edges 3 and 4 having the same initial vertex ‘c’ and
end vertex ‘d’.
m

2. An edge 7 having same initial and end vertex namely ‘e’ and it is said be a self loop.
)A

Definition: 1.4
A graph with finite number of vertices is called finite graph otherwise is called
infinite graph.
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 3

Notes

e
in
nl
Fig (1.4a) Fig (1.4b)

O
The above figure 1.4a is an example of finite graph which has three vertices and
three edges and the figure 1.4b is an example of infinite graph which has an infinite
number of vertices and infinite number of vertices.

ity
Definition: 1.5
Let the two vertices are said to be adjacent if they are connecting by an edge. If
a vertex is incident to an edge then the vertex is one of the two vertices that connects

rs
the edges .In the above figure (1.1) the vertex ‘a’ is adjacent to the vertex ‘b’ and also is
adjacent to ‘d’.
ve
Definition: 1.6
The degree of the vertex is said to be number of edges connecting in it. Let us see
the below example
e2
ni

v1 e1 v2
v3
e2
U

e3
e4
e3
v4 v3

ity

e4
d(v1) = 2

d(v2) = 2
m

d(v3) = 4

d(v4) = 3
)A

d(v5) = 3

Note:
●● In the self loop the degree of the edges must be counted twice.
(c

●● The number of edges in the graph must be twice the sum of the degree of the
vertices.

Amity Directorate of Distance & Online Education

 4 Graph theory and Combinatorics

Theorem: 1.8 (Handshaking Lemma)

Notes

e
In any graph G(V,E) the sum of degrees of vertex is twice the number of edges.
(i.e.)

in
n

∑ d (V ) = 2e
i =1
i

nl
Proof:

O
Let us consider the graph G with ‘V’ vertices and ‘E’ edges. Now we can prove this
theorem with a clear example. For that we consider the above graph Fig (1.3)

The number of vertices in the graph is 5 and the number of edges in the graph is 7

ity
d(v1) = 2, d(v2) = 2, d(v3) = 4, d(v4) = 3, d(v5) = 3

d(v1) + d(v2) + d(v3) +4, d(v4) +d(v5) = 2+2+4+3+3

=14 = twice the number of edges

rs
n
Therefore
∑ d (V ) = 2e
i =1
i
ve
Definition: 1.9
1. A vertex with degree one is called pendent vertex.
ni

2. A vertex with degree zero is called isolated vertex.

3. A graph without any edges is called null graph

U

a b a b
Fig (1.5a) Fig (1.5b)
ity
m

Fig (1.5c)
)A

Fig (1.5a) represents the pendent vertex with degree one.

Fig (1.5b) represents the isolated vertex with degree zero.

(c

Fig (1.5c) represents the null graph without any edges.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 5

Theorem: 1.10
Notes

e
Prove that the number of vertices of odd degree in a graph is always even.

in
Proof:
Let G be graph with ‘n’ vertices and ‘e’ edges.

nl
Let us split the vertices into two of odd degree and an even degree of vertices

Let ‘k’ be the odd degree vertices and ‘j’ be the even degree vertices

O
n n n

∑ d (Vi ) =∑ d (Vk ) + ∑ d (V j ) →(1)

=i 1 =k 1 =j 1

We know that by theorem (1.8) the sum of number of all the edges is twice the sum of

ity
degree of all the vertices.
n
That is
∑ d (V ) = 2e
i =1
i

rs
n n
2e = ∑ d (Vk ) + ∑ d (V j ) →(2)
=k 1 =j 1
ve
Therefore the R.H.S of (2) is an even number

Each d(vj)is even (since the sum of an even numbers is an even number )
ni

Therefore eqn (2) is an even number

The total number of terms in the sum must be even

U

Hence the number of vertices of odd degree in a graph is always even.

Definition: 1.11
ity

1. In a graph a walk is an alternating sequence of vertices and edges and the element
which are incident, that starts and ends with a same vertex
1

e a
m

5 2
)A

d b

4 c 3
(c

Fig (1.6)

Amity Directorate of Distance & Online Education

 6 Graph theory and Combinatorics

In the above fig (1.6) 1a2b3c4d5e1 is a walk.

Notes

e
2. A trail is a walk without repeating any edges.

in
3. A path is a walk without repeating any vertices. In the above Fig (1.6) 1a2b3c4d4e
is a path.

nl
Definition: 1.12
1. A walk is said to be open if the initial and terminal vertices are different.

O
2. A walk is said to be closed if the initial and terminal vertices are same.

ity
5 2

4
rs 3
ve
Fig (1.7)

In the above figure (1.7) 1 → 2 → 3 → 4 → 1 → 5 is an open walk.

ni

In the above figure (1.6) 1a2b3c4d5e1 is a closed walk

Definition: 1.13
U

In a graph a path is said to be a finite or infinite sequence of edges which joins a

sequence of vertices and the vertices are all distinct.

a b
ity

c
m
)A

e d

Fig (1.8)
(c

The above figure (1.8) shows that a,b,c,d,e is a path

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 7

Notes

e
Definition: 1.14

in
A graph is said to be a circuit such that no one edge is repeated and it is also
closed
2

nl
1 4

O
3

ity
6 5
Fig (1.9)

Remark: rs
In the above figure (1.9) 1 → 2 → 4 → 3 → 6 → 5 → 3 → 1 is a circuit.
ve
●● In a circuit vertex can be repeated and edge not repeated.
●● In a cycle the vertex and edge are not repeated.
ni

Definition: 1.15
U

A graph is said to be connected if every pair of vertices in the graph is connected.

[Equivalently, there is a path between every pair of vertex and should be some
path traverse. It is also called the connectivity of a graph]. Otherwise it is said to be
disconnected.
ity

2 a b
1
4
7

3 e
m

6 5
)A

d c

Fig (1.10a) Fig (1.10b)

In the above figure (1.10a) represents the graph is connected.

(c

In the above figure (1.10b) represents the graph is disconnected.

Amity Directorate of Distance & Online Education

 8 Graph theory and Combinatorics

Definition: 1.16
Notes

e
A directed graph is a graph which has a set of vertices connected by the edges and
the edges have a direction with them. The below figure represents the directed graph

in
nl
O
Fig (1.11)

ity
Definition: 1.17
A graph ‘g’ is said to be a subgraph of a graph ‘G’, if all the vertices and all the
edges of ‘g’ are in ‘G’ and each edge of ‘g’ has the same end vertices in ‘g’ as in ‘G’.

rs
a
1 2
6
8 d 6
ve
b 5 e 5 8 d e
7
7
4 3 3
c
ni

G g
U

Fig (1.12)

In the above figure ‘G’ is a graph and ‘g’ is a subgraph of a graph ‘G’ where ‘g’ contains
all the edges in the graph ‘G’
ity

Note:
1. Every graph has its own subgraph
2. A subgraph of a graph ‘G’ is a subgraph of ‘G’
m

3. A single vertex in a graph ‘G’ is also a subgraph of ‘G’

4. A single edge in ‘G’ together which its end vertices is also a subgraph of ‘G’
)A

Definition: 1.18
If two or more subgraphs ‘g1’ and ‘g2’ of G are said to be edge disjoint of ‘G’, if both
‘g1’ and ‘g2’ do not have any edges in common.
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 9

Notes

e
in
nl
O
Fig (1.13a) [G] Fig (1.13b) [g1]

ity
rs
ve
Fig (1.13c) [g2]
ni

The above figure represents the edge disjoint subgraphs

Remark:
U

1. Edge disjoint graphs do not have any edge in common, but they may have vertices
in common

2. Subgraphs that do not have vertices in common is said to be vertex disjoint subgraphs.
ity

The below figure (1.14) represents the vertex disjoint subgraphs where H1 and H2 do
not have any vertices in common.
m
)A

G H1 H2
(c

Fig (1.14)

Amity Directorate of Distance & Online Education

 10 Graph theory and Combinatorics

Theorem: 1.18
Notes

e
Prove that a graph G is disconnected if and only if its vertes set can be partitioned
into two non-empty disjoint subsets,v1, and v2 such that there exist no edge in G whose

in
one end vertex is in subset v1 and the other in subset v2.

Proof:

nl
Let G be a disconnected graph and its vertex set can be partitioned into two non-
empty subsets v1, and v2. Suppose that such a partition exist,

To prove that G is disconnected, Consider two arbitrary vertices ‘a’ and ‘b’of ‘G’

O
such that ‘a’ in v1 and ‘b’ in v2 , no path can exist between vertices ‘a’ and ‘b’, otherwise
their would be at least one edge whose one end vertex would be in v1 and the other in
v2. Hence if partition exist ’G’ is connected.

ity
Conversely,
Let G be a disconnected graph.

Consider a vertex ‘a’ in G. Let v1 be the set of all vertices that are joined by path

vertices will form a set v2.

rs
to ‘a’ .Since G is disconnected ‘v’ does not include all vertices of G. Then the remaining

Therefore no vertices in v1 is joined to only in v2 by an edge.

ve
Theorem: 1.19
Prove that a graph G is (connected or disconnected) has exactly two vertices of
ni

odd degree, there must be a path joining two vertices.

Proof:
U

Case (i)
Let G be a connected graph with exactly two vertices of odd degree say v1 and v2.
There is at least one path joining v1 and v2 because G is connected.
ity

Case (ii)
Let G be a disconnected graph with exactly two vertices of odd degree say v1 and
v2. Since G is disconnected, their must be at least two components, each of which is a
m

connected subgraph.

If v1 and v2 are in two different components say g1 and g2 respectively

Therefore the number of odd degree vertices in g1 is one and also in g2, but g1 is
)A

a graph which is connected.

We know that, in a graph the number of odd degree is always even, which is a
contradiction
(c

Therefore v1 and v2 must be the same component.

Hence there is a path between v1 and v2.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 11

Remark:
Notes

e
The maximum number of edges in a simple graph with ‘n’ vertices is n(n-1)/2

in
Theorem: 1.20
Prove that a simple graph that is a graph without parallel edges or self loop with ‘n’
vertices and ‘k’ components can have atmost ( n − k )( n − k + 1) edges.

nl
2
Proof:

O
Let G be a graph with ‘n’ vertices and ‘k’ components

To Prove:
G has atmost (n − k )(n − k + 1) edges

ity
2
Let n1, n2, n3, ……nk be the number of vertices in 1st , 2nd,…..kth components
respectively.

n1 +n2 +n3, +……+nk = n

rs
k
⇒ ∑ ni − n, ni ≥ 1
i =1
n (n − 1)
i i
ve
The maximum number of edges in the ith component of G is 2

Therefore the maximum number of possible edges in a graph G is

k
ni (ni − 1)
⇒∑
ni

i =1 2 → (1)

(ni2 − 1)k
⇒∑
U

i =1 2
k k
⇒ ∑ n −∑ (ni2 − ni )
2
i
ity

=i 1 =i 1
k k
Now consider i
=i 1 =i 1
∑=
n n,∑ ni2=
−k n
(n1 +n2 +n3, +……+nk ) – (1+1+1+……+1) = n-k
m

(n1 -1) +(n2 -1)+(n3-1)+……+(nk -1) = n-k

k

∑n
)A

i =(ni − 1) = n − k
i =1

∑n i =(ni − 1) 2 =(n − k ) 2
(c

i =1

Amity Directorate of Distance & Online Education

 12 Graph theory and Combinatorics

(n1 -1)2 +(n2 -1)2+(n3-1)2+……+(nk -1)2 + some cross terms = (n-k)2

Notes

e
n12 -2n1 + n22 -2n2 +1+…… +nk2 -2nk +1+ some cross terms = (n-k)2

in
Then (n12 +n22 +….+ nk2) - 2(n1+n2 +…… +nk )+(1+1+1+….+1)

+ some cross terms = (n-k)2

nl
Therefore (n12 +n22 +….+ nk2) - 2(n1+n2 +…… +nk )+k

+ some cross terms = (n-k)2

O
k k
⇒ ∑ n − 2∑ (ni + k ) ≤ (n − k ) 2
2
i
=i 1 =j 1

ity
k
⇒ ∑ ni2 − 2n + k ≤ (n − k ) 2
i =1
k
⇒ ∑ ni2 ≤ (n − k ) 2 + 2n − k

rs
i =1

⇒ n 2 + k 2 ( k − 2n) − ( k − 2n)
ve
k
⇒ ∑ ni2 ≤ n 2 + (k − 2n)(k − 1)
i =1
k
⇒ ∑ ni2 = n 2 − (−2n − k )(k − 1)
ni

i =1
k
ni (ni − 1)
Eqn (1) ⇒∑
U

i =1 2
1
≤ [n 2 − (2n − k )(k − 1) − n]
2
ity

1 2
[n − 2nk + 2n + k 2 − k − n]
2
1 2
m

[ n − 2n − k + k 2 + 2n − k − n]
2
k
ni (ni − 1) 1
⇒∑
)A

≤ [(n − k ) 2 + (n − k )]
i =1 2 2
1
⇒ (n − k )(n − k + 1)
2
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 13

Notes

e
1
Therefore the maximum number of edges is ⇒ (n − k )(n − k + 1)
2

in
Remark:
1. A path is also called simple path or an elementary path

nl
2. A path does not intersect itself.

3. A self loop can be include in a walk but not in a path.

O
4. The number of edges in a path is called the length of the path

Definition: 1.21

ity
A disconnected graph consists of two or more subgroups. Each of the connected
subgraph is called a component.

Definition: 1.22

rs
If some closed walk in a graph contains all the edges of the graph then the walk
is called an Euler line and this is called as Euler graph
ve

ni

U
ity

)A

The above diagram represents the Euler graph

Definition: 1.23
(c

An open walk that includes or traces or covers all the edges of a graph without
retracting any edge is called an unicursal line it is also said to be an open euler line. A
connected graph that has a unicursal line is called unicursal graph

Amity Directorate of Distance & Online Education

 14 Graph theory and Combinatorics

Notes a 4 b

e
6

in
1 3 5 e

nl
c 2 d

Definition: 1.24

O
If two graphs G and G’ are said to be isomorphic to each other, if there is a
one-to-one correspondence between the vertices and between their edges such that the
incidence relationship is preserved.

ity
v5
a e1
v4 e3 v3
e
5 2

rs
4
e4
1 e2
c
6 3
ve
b d
v1 e5 v2
ni

●● The above two graphs are isomorphic.

●● In the above graph the vertices a,b,c,d and e correspond to v1, v2,v3,v4, and v5
respectively. The edges 1,2,3,4,5 and 6 correspond to e1, e2,e3,e4,e5 and e6
U

respectively . Therefore G And G’ are isomorphism

Note:
Without the labels of their vertices and edges the isomorphic graphs are the same
ity

graph that perhaps drawn differently

m
)A
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 15

Remark:
Notes

e
1. The isomorphic graphs must have the same number of vertices and the same number
of edges.

in
2. An equal number of vertices with a given degree

nl
O
ity
rs
ve
ni
U

Isomorphic Graphs
ity
m

Non-Isomorphic graphs
)A

Relationship among subgraphs, walks, paths, and circuits:

(c

Amity Directorate of Distance & Online Education

 16 Graph theory and Combinatorics

Notes

e
Any collection of
Subgraphs of G
edges in G

in
nl
A non edge
Walk in G retracting

O
sequence of
edges of G

ity
Path in G Circuit in G

Non-intersecting rs Non-intersecting
ve
open walk in G closed walk in G

Theorem: 1.25
ni

A given connected graph G is an euler graph if and only if all vertices of a graph
G are of even degree
U
ity
m

Proof:

Necessary condition:
)A

Suppose that G is an euler graph

To Prove:

All vertices in G are of even degree

(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 17

G is an euler graph. It contains an euler line (which is a closed walk)

Notes

e
In tracing this closed walk, we know that every time the walk meets a vertex ‘v’
it goes through two new incident on v which one we entered and with the other existed

in
Thus is true not only of all intermediate vertices of the walk but also of the
terminal vertex because we existed and entered to the same vertex at the beginning and

nl
end of the walk respectively.

Thus G is an euler graph, the degree of every vertex is even.

O
Sufficient condition:

Suppose that all vertices of a graph G are of even degree.

Now we construct a walk starting at an arbitrary vertex ‘v’ and passing through the edges

ity
of G such that non edges is repeated or more than once. We continue tracing as far as
possible.

Since every vertex is of even degree, we can exit from every vertex , we enter , the tracing
cannot stop at any vertex ‘v’

rs
Since ‘v’ is also of even degree we shall eventually reach ‘v’ when the tracing come to an end

Case: (i)
ve
If this closed wallk ‘h’ , we just traced includes all the edges of G is an euler
graph.
ni

Case: (ii)

If ‘h does not contains all the edges ‘G’ , we remove from G all the edges in ‘h’
and obtain a subgraph ‘h’ of G formed by the remaining edges
U

Since both G and ‘h’ have all their vertices of even degree, the degrees of the
vertices of ‘h’ are also even.
ity

Moreover ‘h’ must touch ‘h’ atleast at one vertex ‘a’ because G is connected.

Starting from ‘a’ we can again construct a new walk in ‘h’. Since all the vertices
of ‘h’ are all even degree . This walk in ‘h’ must terminate at vertex ‘a’. But this walk in
‘h’ can be combained with ‘h’ to form a new walk, which starts and ends at vertex ‘v’ and
m

has more edges than ‘h’. This process can be repeated until we obtain a closed walk that
traverses all the edges of G .

Thus G is an euler graph.

)A

Theorem: 1.26
In a connected graph ‘G with exactly 2k odd vertices, there exists ‘k’ edge disjoint
subgraph such that they together contain all edges of ‘G’ and that each is unicursal graph
(c

Amity Directorate of Distance & Online Education

 18 Graph theory and Combinatorics

Proof:
Notes

e
Let the odd vertices of the given graph G be named v1,v2,……vk ; w1,w2,…..wk.

in
In any arbitrary order, Add ‘k’ edges to G between the vertex pairs (v1,w1), (v2,w2), (v3,w3),…..
(vk,wk) to form a new graph G’

Since every vertex of G’ is of even degree g” consists of all and the euler line.

nl
Now if we remove from the euler line the ‘k’-edges, we just added, the euler line and will
be split into ‘k’ walks, each of which is a unicursal line.

O
The first removal will live a single unicursal line. The second removal will split that into the
unicursal lines and each successive removal will split a unicursal line into two unicursal
lines until there are ‘k’ edges of them.

ity
Definition: 1.27
A graph is said to be reach ability, if a directed graph is the pair of objects G=(V,E) where
each node v in V represents a reachable marking and each edge e in E represents a tran-

rs
sition markings. The set of reachable markings can be infinite, even for a finite petri net .

Questions:
ve
1. A graph which has no self loop and no parallel edge graph is__________
a) simple
b) finite
ni

c) infinite
d) none
2. A graph is __________both its vertex set and the edge set are finite
U

a) simple
b) finite
ity

c) infinite
d) none
3. A digraph that has no parallel edges is________ graph
a) simple
m

b) finite
c) infinite
)A

d) none
4. A digraph that has atmost one directed edge between a pair of vertices is ______.
a) asymmetric
b) symmetric
(c

c) a or b

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 19

d) a and b
Notes

e
5. A digraph which is both simple and symmetric is___________
a) simple symmetric graph

in
b) simple asymmetric graph
c) a or b

nl
d) a and b
6. A digraph which is both simple and asymmetric is____________

O
a) simple symmetric graph
b) simple asymmetric graph
c) a or b

ity
d) a and b
7. A digraph is simple antisymmetric if and only if its underlying undirected graph
is___________ .

rs
a) simple
b) finite
c) infinite
ve
d) none
8. Two vertices u and v in V(G) are said to be__________ if there is an edges e in
E(G) such that (e) = (u; v).
ni

a) adjacent
b) not a adjacent
c) vertex
U

d) number
9. Let v be a vertex in a graph G. The degree dV of the vertex v in G is the
___________of G that are incident with v.
ity

a) number of vertex
b) number of edges
c) a or b
m

d) a and b
10. Let G be a graph. Then (G) = min{d(v)/v V (G)} is __________
)A

a) minimum degree
b) maximum degree
c) in degree
d) out degree
(c

Amity Directorate of Distance & Online Education

 20 Graph theory and Combinatorics

Answer:
Notes

e
1. a
2. b

in
3. a
4. b

nl
5. a
6. b

O
7. a
8. a
9. b

ity
10. a

Exercises:
1. Define Graph and give an example
2.
3.
rs
Explain finite and infinite graphs with a suitable example
Draw a graph of your own and find the degree of that graph
ve
4. Explain incidence and degree of a graph.
5. Prove that, the number of vertices of odd degree in a graph is always even.
6. Prove that, if a graph has exactly two of odd degree there must be a path joining
these two vertices.
ni

7. Define Isolated vertex, Pendant vertex and Null graph.

8. Draw all simple graphs of one, two, three and four vertices.
U

9. Define Isomorphism and explain it by comparing two graphs.

10. Define Edge-Disjoint Subgraphs and Vertex - Disjoint Subgraphs.
11. Define Connected graphs, Disconnected graphs and Components
ity

Sample Questions with Answers:

1. Prove that the sum of the degrees of the vertices of any finite graph is even.
m

Proof:
Each edge ends at two vertices. If we begin with just the vertices and no edges,
every vertex has degree zero, so the sum of those degrees is zero, an even number.
)A

Now add edges one at a time, each of which connects one vertex to another, or
connects a vertex to itself (if you allow that). Either the degree of two vertices is
increased by one (for a total of two) or one vertex’s degree is increased by two. In either
case, the sum of the degrees is increased by two, so the sum remains even.
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 21

2. Show that every simple fifinite graph has two vertices of the same degree.
Notes

e
Proof:

in
This can be shown using the pigeon hole principle. Assume that the graph has n
vertices. Each of those vertices is connected to either 0, 1, 2, . . . , n-1 other vertices.
If any of the vertices is connected to n-1 vertices, then it is connected to all the others,

nl
so there cannot be a vertex connected to 0 others. Thus it is impossible to have a
graph with n vertices where one is vertex has degree 0 and another has degree
n-1. Thus the vertices can have at most n-1 different degrees, but since there are n

O
vertices, at least two must have the same degree.
3. Prove that a complete graph with n vertices contains n(n-1)/2 edges.

Proof:

ity
This is easy to prove by induction. If n = 1, zero edges are required, and 1(1-1-0)/2
= 0. Assume that a complete graph with k vertices has k(k -1)/2. When we add the
(k + 1)st vertex, we need to connect it to the k original vertices, requiring k additional
edges. We will then have k(k -1)/2 + k = (k + 1)((k + 1) ) 1)/2 vertices,

rs
ve
ni
U
ity
m
)A
(c

Amity Directorate of Distance & Online Education

 22 Graph theory and Combinatorics

Module - II
Notes

e
Unit- II: Operations on Graph

in
Definition: 2.1

nl
The union of two graphs G1=(v1, e1) and G2=(v2, e2) is another graph G3 written as
G3 = G1  G2 whose vertex set V3 = V1  V2 and the edges are E3 = E1  E 2

O
Definition: 2.2
●● The intersection of graphs G1 and G2 is a graph G4 consisting only of those
vertices and edges that are in both G1 and G2

ity
●● The ring sum of two graphs G1 and G2 (written as G1 ⊕ G 2 is a graph) consisting
of the verex set v1 union v2 and edges that are either in G1 or G2 but not in both

Definition: 2.3

rs
If G1 and G2 are said to be commutative then
G1  G2 = G2  G1
G1  G2 = G2  G1
ve
G1 ⊕ G2 = G2 ⊕ G1

●● If G1 and G2 are edge disjoint then G1 intersection G2 is a null graph

G1 ⊕ G2 = G1  G2
ni

●● If G1 and G2 are vertex disjoint then G1 intersection G2 is empty

⇒ G1  G2 = φ
U

●● For any graph ,

GG = GG = G
G ⊕ G = null graph
ity

G ⊕ g = G − g whenever g ⊆ G

Definition: 2.4
m

A graph G is said to have been decomposed into two subgraphs, if

g1  g 2 = G
g1  g 2 = a null graph
)A

The below diagram shows the union, intersection, and ring sum of the graphs
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 23

Notes

e
in
nl
O
ity
rs
ve
ni
U
ity
m

Theorem: 2.5
)A

Prove that a connected graph G is an euler graph if and only if it can be

decomposed into circuits.
(c

Amity Directorate of Distance & Online Education

 24 Graph theory and Combinatorics

Notes

e
in
nl
Proof:

O
Suppose graph G can be decomposed into circuits (i.e.) G is a union of edge
disjoint circuits

Since the degree of every vertex in a circuit is two, the degree of every vertex in G

ity
is even

Hence G is an euler graph.

Conversely,

rs
Coinsider a vertex v1,. There are atleast two edges incident at v1

Let one of these edges between v1 and v2

ve
Since vertex v2 is also of even degree, it must have atleast another edge say
between v2 and v3

Proceeding in this way, we eventually arrive at a vertex that has previously been
traversed thus forming a circuit. Let us remove the circuit form G. All the vertices in
ni

the remaining graph must also be of even degree. From the remaining graph remove
another circuit in exactly the same way as be removed the circuit from G. Continuing
this process until no edges are left.
U

Hence a connected graph G is an euler graph if and only if it can be decomposed

into circuits.
ity

Definition: 2.6
In an euler graph , it is an euler line is obtained when follows any closed walk
from a vertex V such that one arrives at a vertex can select any edges which has not
be previous traversed. Such a graph is called an arbitrarily traceable graph from the
m

vertex
)A
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 25

Notes

e
in
nl
O
Definition: 2.7

ity
Hamiltonian circuit in a connected graph is defined as a closed walk that
traverses every vertex of G exactly once, except starting and final vertex

A circuit in a connected graph G is said to be Hamiltonian it includes every vertex

of G. Hence a Hamiltonian circuit in a graph of n vertices of exactly n edges

rs
ve
ni
U

Hamiltonian circuit Graph of decahedron with Hamiltonian circuits

ity
m

Graph without Hamiltonian circuits

)A

Definition: 2.8
If we remove any one edge from a Hamiltonian circuit, we are left with a path. This
path is called a Hamiltonian path.
(c

Amity Directorate of Distance & Online Education

 26 Graph theory and Combinatorics

Definition: 2.9
Notes

e
A simple graph in which there exist an edge between every pair of vertices is called
a complete graph

in
nl
O
ity
rs
ve
Theorem: 2.10
In a complete graph with ‘n’ vertices there are n-1/2 edges disjoint Hamiltonian
circuits if ‘n’ is an odd number greater than or equal to 3
ni

Proof:
A complete graph G of ‘n’ vertices has n(n-1)/2 edges, and a Hamiltonian circuit in
G consist of ‘n’ edges
U

Therefore the number of edge disjoint Hamiltonian circuit, when n is odd it can be
shown as follows.
ity

The subgraph (of complete graph of ‘n’ vertices) in figure is a Hamiltonian circuit

Keeping the vertices on a circle rotate the polygonal pattern clockwise by

360 360 360 n − 3 360

,2. ,3. ,...... , degrees
n −1 n −1 n −1 2 n −1
m

Observe that each rotation produces a Hamiltonian circuit that has no edges in
common with any of the previous one .
)A

Thus n-3/2 new Hamiltonian circuits, all edge disjoint from the one in figure and
also edge disjoint among themselves.

Hence there are n-1/2 edge disjoint Hamiltonian circuit.

(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 27

Notes

e
in
nl
O
ity
Problem: 1
Find the 4 edge disjoint Hamiltonian cycles in K9 .

Solution:

rs
ve
ni
U

(i) (ii)
ity
m
)A

(iii) (iv)

Edge disjoint hamiltonian cycles in K9

(c

Amity Directorate of Distance & Online Education

 28 Graph theory and Combinatorics

Dirac’s Theorem:
Notes

e
A vertex in G be at least sufficient (but by no means necessary condition) for a
simple graph G to have a Hamiltonian circuit is that the degrees of every n/2, where ‘n’

in
is the number of vertices in G

Proof:

nl
Assume that G is non-hamiltonian circuit

Let G be the maximal non-hamiltonian with n≥3

n

O
d (v i ) ≥ , i=1 to n
2
We know that,

Any complete graph is Hamiltonian.

ity
Therefore G cannot be complete.

There exist a pair of vertices u and v in G, which are not adjacent

Consider G’=G+uv by choice of G. G’ becomes Hamiltonian

rs
As G is non-hamiltonian. this mean that every hamiltonian cycle of G must contain
the edge uv (i.e.) Removal of an edge uv from any Hamiltonian cycle of G, results in
a(u–v) Hamiltonian path in G
ve
Let p=v1,v2,v3,…….,vn with v1 = u as original and vn =v as terminous

By definition,
=A {v i / uv i ∈ E (G )}
ni

=B {v i / v i u ∈ E (G )}

Now, uv ∉ E (G )
U

v ∉ A u ∉ B and v ∉ B , v ∉ A  B and | A  B |< n

⇒ vk ∈ A  B
ity

v1,v2,v3,…….,vk, vn,vk-1, v1 is a Hamiltonian cycle of G

Which is a contradiction to the fact that G is non-hamiltonian

Therefore A  B = φ
m

(i.e.) | A  B |= 0

Now the definition of A and B,

)A

|A|=dG(u) and |B|=dG(V)

Therefore dG(u) +dG(V)=|A| + |B|

≤ | A  B | + | A  B |

< n+0
(c

dG(u) +dG(V) < n

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 29

By the hypothesis,
Notes

e
n
⇒ dg (u ) ≥
2
n

in
⇒ dg (v) ≥
2
⇒ dG(u) +dG(V) ≥ n + n
2 2

nl
⇒ dG(u) +dG(V) ≥ n

Which is a contradiction

O
Therefore our assumption is wrong

Therefore G must be Hamiltonian.

ity
Shortest Path Problem:
Let us given a weighted network (V,E,C) with vertex set V, edge set E, and the weight
set C specifying weight scij for the edges (i,j) ∈E. Weare also given a starting nodes ∈V. The
one-to-all shortest path problem is the problem of determining the shortest path from

rs
nodes to all the other nodes in the network.
ve
ni
U

Algorithms Solving the Problem

ity

Dijkstra’s algorithm:
Solves only the problems with non-negative costs, i.e.,
m

cij≥ 0 for all (i,j)∈E

Properties of Dijkstra’s algorithm

)A

Many more problems than you might at first think can be cast as shortest path
problems, making Dijkstra’s algorithm a powerful and general tool.

For example:

●● Dijkstra’s algorithm is applied to automatically find directions between physical

(c

locations, such as driving directions on websites like Map quest or Google

Maps.

Amity Directorate of Distance & Online Education

 30 Graph theory and Combinatorics

●● In a networking or telecommunication applications, Dijkstra’s algorithm has

Notes

e
been used for solving the min-delay path problem (which is the shortest path
problem).

in
●● It is also used for solving a variety of shortest path problems arising in plant
and facility layout, robotics and transportation.
●● Dijkstra’s algorithm solves such a problem and finds the shortest path from a

nl
given nodes to all other nodes in the network where node s is called a starting
node or an initial node
●● Dijkstra’s algorithm starts by assigning some initial values for the distances

O
from nodes and to every other node in the network.
●● It operates in steps, where a teach step the algorithm improves the distance
values.

ity
●● A teach step, the shortest distance from nodes to another node is determined.

Characterizations:
●● The algorithm characterizes each node by its state. The state of a node
consists of two features
(i) Distance value
(ii) Status label
rs
ve
●● Distance value of a node is a scalar representing an estimate of its distance
from nodes.
●● Status label is an attribute specifying whether the distance value of a node is
equal to the shortest distance to node s or not.
ni

●● The status label of a node is Permanent if its distance value is equal to

the shortest distance from nodes. Otherwise, the status label of a node is
Temporary.
U

●● The algorithm maintains and step-by-step updates the states of the nodes. At
each step one node is designated as current

Algorithm Steps
ity

Step 1. Initial Stage

●● As sign the zero distance value to nodes, and label it as Permanent. [The
state of node s is (0,p).]
m

●● Assign to every node a distance value of ∞ and

Label them as Temporary. [The state of every other node is (∞,t).].
●● Designate the nodes as the current node
)A

Step 2.
●● Distance Value Update and Current Node Designation Update Let i be the
index of the current node.
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 31

●● Find the set J of nodes with temporary labels that can be reached from the
Notes

e
current node i by a link (i, j). Update the distance values of these nodes.
●● For each j∈J, the distance valuedj of node j is updated as follows new

in
dj=min{dj,di+cij} where cijis the cost of link (i,j), as given in the network
problem.
●● Determine a node j that has the smallest distance valuedj among all nodes

nl
●● j ∈ J,find j such that mindj=dj∗j∈J.
●● Change the label of node j permanent and designate this node as the current
node.

O
Step 3. Terminal Stage
●● If all nodes that can be reached from nodes have been permanently labeled,
then stop - we are done.

ity
●● If we cannot reach any temporary labeled node from the current node, then all
the temporary labels become permanent
●● Otherwise, go to Step 2.

rs
Problem: 3
To find the shortest path from node 1 to all other nodes using Dijkstra’s algorithm.
ve
ni
U

Step1: Initial Stage

●● Node 1 is designated as the current state.
ity

●● The state of node 1 is (0,p).

●● Every other node has state (∞,t)

Step: 2
m
)A
(c

Amity Directorate of Distance & Online Education

 32 Graph theory and Combinatorics

●● Nodes 2,3, and 6 can be reached from the current node 1.

Notes

e
●● Update distance values for these nodes d2= min{∞, 0 + 7} = 7
●● d3= min{∞, 0 + 9} = 9 d6=min{∞,0+14}=14. Now, among the nodes

in
2,3, and6, node 2 has the smallest distance value.
●● The status label of node 2 changes to permanent, so its state is (7,p), while
the status of 3 and 6 remains temporary.

nl
●● Node 2 becomes the current node

Step: 3

O
So graph is end at the step 2

ity
●● rs
We are not done, not all nodes have been reached from node1, so we will
ve
repeat again step: 2
●● Node 2 becomes the current node

Another Implementation of Step: 2

ni

●● Nodes 3 and 4 can be reached form the current node 2.

●● Update the distance values for theses nodes
d3=min{9,7+10}=9
U

d6= min{∞,7+15} = 22

●● Now, between the nodes 3 and 4 node 3 has the smallest distance value.
ity

●● The status label of node 3 changes to permanent, while the status of 6

remains temporary.
●● Node 3 becomes the current node.
m
)A
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 33

●● We are not done (Step 3 fails), so we perform another Step 2

Notes

e
Another Step: 2

in
●● Node 5 can be reached from the current node 6.
●● Update distance value for node 5.
d5= min{∞,11+9} = 20

nl
O
ity
●●
●●
rs
Now, node 5 is the only candidate, so its status changes to permanent.
Node 5 becomes the current node
ve
●● From node 5 we cannot reach any other node.
Hence, node 4 gets permanently labeled

Problem: 4
ni

Find the shortest distance from A to H on the network below

U
ity

Solution:
m

The fully labeled diagram below shows that the values, both temporary and
permanent at each vertex
)A
(c

Amity Directorate of Distance & Online Education

 34 Graph theory and Combinatorics

Notes

e
in
nl
O
Problem: 6
Find all pairs of shortest path using dynamic programming

ity
rs
ve
Solution:
ni

Let us start 12,13,14

21,23,24
U

31,32,34

41,22,43

Now by solving dynamic programming, it can be solved to taking sequence of

ity

decisions. In each we have to take a decision.

Now we have to preparing a matrix using the problem

0 3 ∞ 7
m

8 0 2 ∞
A0 =
5 ∞ 0 1
2 ∞ ∞ 0
)A

If there is no edge or an absence of an edge then it takes to be infinity

Step: 1
Let us consider the First row and first column it remains the same also the
(c

diagonals should be remains the same

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 35

Ak [i,j}=min{Ak-1 [i,j],Ak-1 [i,k]+Ak-1 [k,j]}

Notes

e
(i) A0[2,3]= A0[2,1]+ A0[1,3]
2 < 8 + ∞

in
(ii) A0[2,1]= A0[2,1]+ A0[1,4]
∞ > 8 + 7

nl
(iii) A0[2,1]= A0[2,1]+ A0[1,4]
∞ > 5 + 3

O
In the same we have to prepare the values and form a matrix A1

0 3 ∞ 7
8 0 2 15
A1 =

ity
5 8 0 1
2 8 ∞ 0

Step: 2

rs
Here also we consider the diagonals remains same and we prepare this matrix A2
by the above matrix A1 so that we have to kept the second row and second column must
be constant.
ve
(i) A1[1,3]= A1[1,2]+ A1[2,3]
2 > 3 + 2
(ii) A1[1,4]= A1[1,2]+ A1[2,4]
ni

7 < 3 + 15
In the same we have to prepare the values and form a matrix A1
U

0 3 5 7
8 0 2 15
A2 =
5 8 0 1
ity

2 5 7 0

Step: 3
Here also we consider the diagonals remains same and we prepare this matrix A2
m

by the above matrix A1 so that we have to kept the third row and third column must be
constant .
)A

(i) A2[1,2]= A2[1,3]+ A2[3,2]

3 > 5 + 8
In the same we have to prepare the values and form a matrix A1
(c

Amity Directorate of Distance & Online Education

 36 Graph theory and Combinatorics

0 3 5 6
Notes

e
7 0 2 3
A3 =
5 8 0 1

in
2 5 7 0

nl
Step: 4
Like wise we prepared the matrix A4

0 3 5 6

O
5 0 2 3
A4 =
3 6 0 1
2 5 7 0

ity
Ak [i,j}=min{Ak-1 [i,j],Ak-1 [i,k]+Ak-1 [k,j]}

Travelling salesman problem:

rs
●● There are number of cities a salesman must visit e.g.: A,B,C&D
●● The distance / time/cost between every pair of cities is given
●● The salesman starts from his home city, he must visit every city exactly once
ve
and return to his home city
●● The problem is to find the route shortest distance / time/ cost

Phase –I
ni

●● TSP can be first solved as Assignment problem (AP) by using Hungarian

method to find optimum solution.
●● Then check the TSP condition
U

●● If the condition is satisfied, then the AP solution will be the optimum solution
even for TSP.

If not go to Phase –II

ity

●● I - the solution can be adjusted by inspection

●● II-form a single circuit
●● III- The iterative procedure –Branch and Bound method
m

Problem: 6
A travelling salesman has planned to visit from a particular city, visit each city only
once and return to the starting city. The travelling cost in rupees is given in the table
)A

below. Find the least cost route.

(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 37

Notes

e
in
nl
O
ity
Solution:

Step: 1 (Row Reduction Method)

First assign the ∞ for the diagonal values

rs
ve
ni
U

Step: 2 (Column Reduction Method )

ity
m
)A
(c

Amity Directorate of Distance & Online Education

 38 Graph theory and Combinatorics

Step: 3
Notes

e
in
nl
O
ity
Step: 4

rs
ve
ni
U

Step: 5
ity
m
)A
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 39

Step: 6
Notes

e
in
nl
O
ity
ACDBA

75+15+75+35

rs
=Rs. 200

Bipartite Graphs
ve
Definition: 2.11
A Bipartite graph is a graph whose vertices can be divided into two disjoint
and independent setsU {\displaystyle U} U and V {\displaystyle V} V such that every
ni

edge connects in a vertex in U {\displaystyle U} U to a vertex V. It is called Bigraph.

Equivalently each edge is either (u, v) which connects edge a vertex from set U to
vertex from set V or (v, u) which connects edge a vertex from set V to vertex from set U.
U

If both sets have equal cardinality (means both sets have an equal number of
vertices), then it is called balanced bipartite graph and if each vertex in set U has an
edge to all the vertices in another set V and vice versa then that bipartite graph is called
ity

complete bipartite graph.

Some of the properties are:

●● Every tree is bipartite.
m

●● Cycle graphs with an even number of vertices are bipartite.

●● Every planar graph whose faces all have even length is bipartite.
●● The complete bipartite graph on m and n vertices, denoted by Kn,m is the
)A

bipartite graph G = ( U , V , E ) {\displaystyle G=(U,V,E)}, where U and V are

disjoint sets of size m and n, respectively, and E connects every vertex in U
with all vertices in V. It follows that Km,n has mn edges.
●● Hypercube graphs, partial cubes, and median graphs are bipartite.
(c

Amity Directorate of Distance & Online Education

 40 Graph theory and Combinatorics

Notes

e
in
nl
O
ity
rs
ve
ni

The above graphs are called Bipartite graphs

Definition: 2.12
U

A graph is said to complete bipartite graph whose vertices can be partitioned into
two subsets v1 and v2 such that no edge has both endpoints in the same subset, and
every possible edge that could relate vertices in different subsets is part f the graph.
ity

The structure of bipartite graph is (V1,V2,E) where v1 ∈ V1 and v 2 ∈ V2

m
)A
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 41

The above graphs are called complete bipartite graphs

Notes

e
in
nl
O
Examples of complete bipartite graphs:

ity
rs
ve
Questions:
1. A simple digraph in which there is exactly one edge directed from every vertex to
ni

every other vertex is a___________ digraph.

a) complete symmetric
U

b) complete asymmetric
c) symmetric
d) asymmetric.
ity

2. The vertices and edges in a _________need not be distinct.

a) walk
b) not a walk
m

c) a or b
d) a and b.
)A

3. A graph is said to__________ whose vertices can be partitioned into two subsets
a) bipartite graph
b) complete graph
c) graph
(c

d) none of the above.

Amity Directorate of Distance & Online Education

 42 Graph theory and Combinatorics

4. Cycle graphs with an ______ number of vertices are bipartite.

Notes

e
a) Odd
b) even

in
c) both a and b
d) none.

nl
5. In Dijkstra’s algorithm characterizes each node by its state. The state of a node
consists of two features are Distance value __________.
a) Nodes

O
b) vertices and edges
c) Status label

ity
d) none .
6. If both sets have equal cardinality (means both sets have an equal number of
vertices, then it is called ____________.
a) Bipartite graph
b) complete graph
c) balanced bipartite graph rs
ve
d) digraph
7. If we remove any one edge from a Hamiltonian circuit, we are left with a path.
This path is called a ___________
a) Euler walk
ni

b) Hamiltonian path
c) fundamental circuit
U

d) none of the above.

8. A simple graph in which there exist an edge between every pair of vertices is
called a ____________.
ity

a) complete graph
b) simple graph
c) bipartite graph
m

d) none of the above .

9. If G1 and G2 are said to be __________ then
G1  G2 = G2  G1
)A

G1  G2 = G2  G1
G1 ⊕ G2 = G2 ⊕ G1

a) Commutative
(c

b) Associative

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 43

c) additive
Notes

e
d) none of the above
10. If G1 and G2 are _________then G1 intersection G2 is a null graph

in
G1 ⊕ G2 = G1  G2

a) Vertex disjoint

nl
b) edge disjoint
c) both a and b

O
d) none of the above

Answer:
1. a

ity
2. a
3. a
4. b
5.
6.
c
c
rs
ve
7. b
8. a
9. a
ni

10. b

Problems:
U

1. If a set S={1,2,3...,n}, among these we have selected m elements, for which the
sum of these elements is k. What will be the best logic for this?
2. How to arrange the vertices (which have different lengths) of a graph on a
straight line such that the bandwidth of the graph is minimized?
ity

3. What are the ways to transfer a graph from one Relation space to a Euclidean
space with less time complexity?
4. Derive the relationship between vertex connectivity and edge connectivity.
m

Exercises:
1. Explain Shortest Path Problem
)A

2. Write algorithm for shortest path problem

3. Define Vertex connectivity and Edge Connectivity
4. Define Hamiltonian cycle and Hamiltonian Circuit
(c

Amity Directorate of Distance & Online Education

 44 Graph theory and Combinatorics

1. For what values of n does the graph Kn contain an Euler trail? An Euler tour?
Notes

e
A Hamilton path? A Hamilton cycle?
Solution:

in
Euler trail: K1, K2, and Kn for all odd n -3.
Euler tour: Kn for all odd n -3.

nl
Hamilton path: Kn for all n -1.
Hamilton cycle: Kn for all n- 3
2.

O
(a) For what values of m and n does the complete bipartite graph Km;n contain an
Euler tour?
(b) Determine the length of the longest path and the longest cycle in Km;n, for all

ity
m; n.
Solution:
(a) Since for connected graphs the necessary and sufficient condition is that the

rs
degree of each vertex is even, m and n must be even positive integers.
(b) The length of the longest cycle is 2. Any cycle must be even, and it must
alternate vertices from the two sides. Thus it cannot be longer than twice the
ve
smaller side (and such a cycle clearly exists). By a similar reasoning, we get
that if m = n, the longest path contains all the 2m vertices, so its length is 2m -1,
and if m 6= n, the length of the longest path is 2 and starting and ending in the
larger class.
ni

3.
(a) Find a graph such that every vertex has even degree but there is no Euler tour.
(b) Find a disconnected graph that has an Euler tour.
U

Solution:
(a) Take a graph that is the vertex-disjoint union of two cycles. It is not connected,
so there is no Euler tour.
ity

(b) The empty graph on at least 2 vertices is an example. Or one can take any
connected graph with an Euler tour and add some isolated vertices.
4. Determine the girth and circumference of the following graphs.
m

Solution:
The graph on the left has girth 4; it’s easy to _nd a 4-cycle and see that there is no
3-cycle.
)A

It has circumference 11, since below is an 11-cycle (a Hamilton cycle).

The graph on the right also has girth 4. It also has circumference 11, since below is
an 11-cycle.
Actually, one can check that the two graphs are isomorphic.
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 45

Module - III
Notes

e
Key Learning Objectives:

in
At the end of this module, you will be able to:

●● Discuss basic information on trees

nl
●● Describe rooted trees
●● Determine the path length in rooted trees and spanning trees

O
●● Perform a study of fundamental circuits
●● Interpret spanning trees of a weighted graph
●● Implement the concepts of cut sets and cut vertices

ity
●● Discuss basic concepts of minimum spanning tree

rs
ve
ni
U
ity
m
)A
(c

Amity Directorate of Distance & Online Education

 46 Graph theory and Combinatorics

Unit- III
Notes

e
Definition: 3.1

in
A Tree is a connected graph without any circuits.

nl
Theorem: 3.2
Prove that there is one and only one path between every pair of vertices in a tree T.

O
Proof:
Since T is connected, there must exist atleast one path between any pair of
vertices in T.

ity
Now suppose that between two vertices and ‘a‘ and ‘b’ of T. There are two distinct
Paths.

The union of these two paths will contain a circuit which is impossible in a tree.

Hence there is a unique path between any pair of vertices.

Theorem: 3.3
rs
ve
Prove that in a graph G there is one and only one path between every pair of
vertices G is a tree.

Proof:
ni

In a graph G, there is one and only one path between every pair of vertices.

⇒ G is connected
U

Suppose, G has a circuit.

⇒ There is atleast one pair of vertices a, b such that, there are two distinct paths
between a and b.
ity

Since G has one and only one path between every pair of vertices, G can have no
circuit.

Hence G is a tree.
m

Theorem: 3.4
Prove that a tree with ‘n‘ vertices has n-1 edges
)A

Proof:
The Theorem will be proved by induction on the number of vertices.

It is easy to see that the theorem is true for n = 1,2,3.

(c

Assume that the theorem for all trees with f ever than n vertices.

Let us now consider a tree T with n vertices

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 47

Notes

e
in
nl
O
In T let ek be an edge with end vertices Vi and Vj. According to theorem, there is no
other path between Vi and Vj except ek .

Deletion of ek from T will disconnect the graph as shown in figure.

ity
Further more T-ek consist of exactly two components, and since there were no
circuits in T to begin with, each of these component is a tree.

Let the number of vertices of tree t1 and t2 be n1 and n2 respectively such that

rs
n1+n2=n

Also n1<n and n2<n

By induction hypothesis, the number of edges in t1 and t2 are n1-1 and n2 -1

ve
respectively.

Thus T-ek consist of

n1-1+ n2-1= n1 + n2-2

ni

⇒ n-2 edges
Therefore T has exactly n-1 edges.
U

Theorem: 3.5
Prove that a connected graph with ‘n’ vertices and ‘n-1’ edges is a tree
ity

Proof:
Let G be a connected graph with ‘n’ vertices and ’n-1’ edges

We show that G contains no circuits

m

Assume to the contrary that G contain circuit. Remove an edge from a circuit so
that, the resulting graph is again connected.
)A

Continue this process of removing one edge from one circuit at a time till the
resulting graph ‘H’ is a tree

As H has ‘n’ vertices so number of edges in H is n-1. Now the number of edges in
G is greater than the number of edges in H.
(c

So, n-1 > n-1, which is not possible

Hence G has no circuits and therefore is a tree.

Amity Directorate of Distance & Online Education

 48 Graph theory and Combinatorics

Definition: 3.6
Notes

e
A graph G is said to be minimally connected if removal of anyone edge from it
disconnects the graph. Clearly a minimally connected graph has no circuits.

in
Theorem: 3.7

nl
Prove that a graph is a tree if it is minimally connected.

Proof:

O
Let the graph G be minimally connected.

Then G has no circuits and therefore is a tree.

Conversely,

ity
Let G be a tree.

Then G contains no circuits and deletion of any edge from G disconnect the graph.

Hence G is minimally connected.

Theorem: 3.8
rs
Prove that a Graph G with n vertices, n-1 edges no circuits is connected.
ve
Proof:
Suppose that there exist a circuitless graph G with `n’ vertices and (n-1) edges
which is disconnected.
ni

In that case, G will consist of two or more circuitless component. Without loss of
generality, Let G consist of two components g1 and g2.
U

Add an edge e between a vertex v1 in g1 and v2 in g2.

Since there was no path between v1 and v2 in G, adding e did not create a circuit
This G  e is a circuitless connected graph (i.e., a tree) of n vertices and n edges which
ity

is not possible. Hence G is connected.

Note:
●● G is connected and is circuitless (or)
m

●● G is connected and has (n-1) edges (or)

●● G is circuitless and has (n-1) edges (or)
●● There is exactly one path between every pair of vertices in G (or)
)A

●● G is minimally connected graph.

Definition: 3.9
In a tree, the vertex is said to be pendent if its degree is one.
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 49

Theorem: 3.10
Notes

e
Prove that in any tree (with two or more vertices) there are atleast two pendent
vertices.

in
Proof:
Suppose that the tree T has `n’ vertices.

nl
Case: (i)
n=2

O
T is of the form

ity
Hence we have two pendent vertices.

Case: (ii)
n>2

rs
Suppose T has n-1 edges.

We know that,
ve
Every edge contribute two degrees. There are 2(n-1) degrees in the graph.

There 2(n-1) degree assigned to n vertices. In a tree we cannot have any vertex of
degree zero.

Therefore atleast two vertices having degree one.

ni

Therefore atleast two pendent vertices in a tree.

U

Definition: 3.11
In a connected graph G, the distance d(vi , vj) between two of its vertices vi and vj is
the length of the shortest path between them.
ity
m
)A

Definition: 3.12
(c

Let x be a nonempty set. The function ‘t’ is said to be metric of an ‘x’. If it satisfies
the following condition.

Amity Directorate of Distance & Online Education

 50 Graph theory and Combinatorics

(i) Non-negativity f(x,y)≥0 and f(x,y)=0 iff x=y.

Notes

e
(ii) Symmetry : f(x,y) = f(y,x)
(iii) Triangle inequality: f(x,y) ≤ f(x,z) if (z,y) for some z.

in
Definition: 3.13

nl
The eccentricity E(v) of a vector V in a graph G in the distance from V to the
vertex farthest from V in G

(i.e,) E(v) = max d (v,vi )

O
Vi ∈ G

Example:

ity
rs
ve
d(v,v1) =1, d(v,v3) =1, d(v,v2) =2 , d(v,v4) =2 , d(v,v5) =3
ni

Therefore E(v) =3

Definition: 3.14
U

A vertex with minimum eccentricity in graph G is called a center of G.

ity
m

E(d) =3 E(b) =1

E(c) =2 E(a) =2
)A

Therefore ‘b’ is the center.

Definition: 3.15
(c

If a graph G has two center we refer to such centers as bicenters.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 51

Theorem: 3.16
Notes

e
Prove that every tree has either one or two centers.

in
Proof:

Case: (i)

nl
T has one vertex.

The vertex itself its center.

O
Case: (ii)
T has two vertices.

T is of the form

ity
Therefore ‘T’ has two centers.

Case: (iii)
T has more than two vertices by theorem,

There are atleast two pendent vertices. rs

ve
Let v be any vertex in T.

The maximum distance, max d(v,vi) from given vertex v to any other vertex vi
occurs only when vi is a pendent vertex.
ni

Delete all the pendent vertices from T.

The remaining graph Tꞌ is still a tree.

U

Tꞌ is also true with eccentricity of each vertex in one less than of T.

The center of Tꞌ are the center of T, only by deleting all the pendent vertices in Tꞌ.

We get another tree T” with the same center.

ity
m
)A
(c

Amity Directorate of Distance & Online Education

 52 Graph theory and Combinatorics

Notes

e
Proceeding in this way until there is left a vertex or an edge.

in
If we have one vertex then the tree has one center. If we have one edge then the
tree has two center the end vertices of the edge.

Therefore the tree has one or two centers.

nl
Remark:
If a tree T has two centers, the two centers must be adjacent.

O
Definition: 3.17
An eccentricity of a center in a tree defined as the radius of tree.

ity
Definition: 3.18
The diameter of a tree T is defined as the length of longest path in tree.

Definition: 3.19

rs
A tree in which one vertex called the root is distinguished from all the others is
called a rooted trees.
ve
Non-rooted trees are called free tree.
ni
U

Definition: 3.20
A binary tree is defined as a tree in which is exactly one vertex of degree two, and
ity

each of the remaining vertices is of tree one or three.

m
)A

Note:
The number of vertices is a binary tree is always odd.
(c

Definition: 3.21
A non-pendent vertex is a tree is called internal vertex.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 53

Definition: 3.22
Notes

e
In a binary tree a vertex vi is said to be at level li, if vi at a distance of li from the root.

in
nl
O
ity
Definition: 3.23
The maximum level lmax of any vertex in binary tree is called the height of a tree.

Note:
Minimum possible height of an n-vertex binary tree is min rs
ve
lmax =┌log2(n+1)-1┐ where ┌n┐ denotes the smallest integer ≥ n.

Max lmax =n-1 /2.

ni

Definition: 3.24
Weighted path length is defined as every pendent vertex vj of a binary tree has
associated with it a positive real number wj.
U

Given w1,w2,……,wn the problem is to construct a binary tree that minimize.

∑wjlj where lj is the level of pendent vertex vj, and the sum is taken over all pendent
vertices.
ity

Definition: 3.25
A tree t is said to be a spanning tree of a connected graph G if T is a subgraph of
m

G and T contains all vertices of G.

Spanning tree is sometimes referred to as a skeleton or scaffolding of G.

)A

Definition: 3.26
A collection of tree is called a forest.
(c

Amity Directorate of Distance & Online Education

 54 Graph theory and Combinatorics

Notes

e
in
nl
O
ity
Remark:
The complement of the complete graphKn is the empty graph with n vertices.

rs
ve
ni
U
ity

Theorem: 3.27
Every connected graph has atleast one spanning tree.
m

Proof :
Let G be a connected graph.
)A

Case: (i)
G has no circuits.

Since G is connected .G is a tree

(c

Therefore G itself is a spanning tree.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 55

Case: (ii)
Notes

e
G has circuits.

in
Delete an edge from a circuit in G, the remaining graph is also a connected graph.

If there are more circuits in the remaining graph repeat the produce till on edge
from the last circuit from graph that contain all the vertices of G.

nl
Hence it is spanning tree.

Definition: 3.28

O
An edge in a spanning tree T is called branch of T.

Definition: 3.29

ity
An edge of G that is not in a given spanning tree T is called a chord.

Chord is sometimes referred to a tie or a link.

Note:

rs
The subgraph T is a collection of chords it is referred to as the chord said the (or tie
set or cotree).
ve
Theorem: 3.30
With respect to any of its spanning trees, a connected graph of n vertices and e
edges has n-1 branches and e-n+1 chords.
ni

Proof:
Let G be a graph with `n’ vertices and e edges, suppose that T is any spanning tree
U

in G.

⇒ T is a tree with n vertices .

⇒ T has n-1 edges.
ity

The number of branches in T is n-1.

Therefore the remaining e-(n-1) are chords of T.

There are e-n+1 chords of T.

m

Definition: 3.31
Let G be a graph with ‘k’ -components ‘n’ vertices, ‘e’ edges
)A

Rank(r)=n-k

Nullity( µ )=e-n+k

Note:
(c

Rank of G = number of branches in any spanning tree of G.

Amity Directorate of Distance & Online Education

 56 Graph theory and Combinatorics

Nullity of G = number of chords in G.

Notes

e
Rank+Nullity = number of edges in G.

in
Nullity of a graph is also referred to as its cyclomatic number or first betti number.

Definition: 3.31

nl
Let T be any spanning tree in a connected graph G. Adding any one chord to T will
create exactly one circuit such a circuit, formed by adding a chord to a spanning tree is
called a fundamental circuit.

O
ity
Problem: 1
In the graph rs
ve
ni
U

{e1, e4}, {e6, e7}, {e1, e2, e3}, {e8}, {e3, e4, e5, e6}, {e2, e5, e7}, {e2, e5, e6} and {e2, e3, e4}
arecut sets. Will find that there are any other cut sets?

Solution:
ity

In a graph G= (V, E), a pair of subsets V1 and V2 of V satisfying

V =V1∪V2, V1∩V2=∅, V1=∅, V2=∅,

is called a cut (or a partition) of G, denoted V1, V2 .

m

Usually, the cuts V1, V2 and V2, V1 are considered to be the same.
)A

Definition: 3.32
The distance between two spanning tree Ti and Tj of a graph G is defined G as the
no of edges of G present in one tree but not in other.

The distance may written as d(Ti,Tj)

(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 57

Notes

e
in
nl
O
ity
Remark:
rs
ve
The distance between the spanning tree of a graph is a metric.

ie.,) i) d(Ti,Tj) ≥ 0 d(Ti,Tj) = 0 iff Ti = Tj

ii) d(Ti, Tj) = d(Ti,Tj)
ni

iii) d(Ti,Tj) ≤ d(Ti,Te) + d(Tk,Tj)

Remark:
U

Max d(Ti,Tj) =1/2 max N(Ti ⊕ Tj)

≤ r, the rank of G

Max d(Ti,Tj) ≤ μ
ity

Max d(Ti,Tj) ≤ min (μ,r)

Definition: 3.33
m

For a spanning tree T0 of a graph G, let max d(Ti,Tj)denote the maximal distance
between T0 and any other spanning tree of tree. Then T0 is called a central tree of G if
max d(To,Ti) ≤ d(T,Tj) for every tree T of G.
)A

Definiton: 3.34
The tree graph of a given graph G is defined as a graph in which each vertex
corresponds to a spanning tree of G, and each edge corresponds to a cyclic
interchange between the spanning tree is of G representing by the two and vertices of
(c

the edge.

Amity Directorate of Distance & Online Education

 58 Graph theory and Combinatorics

Definition: 3.35
Notes

e
A graph G is called a labeled graph if its edges or vertices are assigned data of
one kind or another. In particular, G is called a weighted graph if each edge e of G is

in
assigned a nonnegative number w(e) called the weight or length of v.

The below figure shows a weighted graph where the weight of each edge. The
weight (or length) of a path in such a weighted graphG is defined to be the sum of

nl
the weights of the edges in the path. In the below figure the length of a shortest path
between P and Q is 14;

O
One such path is (P,A1,A2,A5,A3,A6,Q)

Definition: 3.36
A graph G is said to be complete if every vertex in G is connected to every other

ity
vertex in G. Thus a complete graph G must be connected. The complete graph with n
vertices is denoted by Kn.

Definition: 3.37
●●

rs
A graph G is said to be regular of degree k or k-regular if every vertex has
degree k. In other words, a graph is regular if every vertex has the same
degree.
ve
●● The connected regular graphs of degrees 0, 1, or 2 are easily described.
●● The connected 0-regular graph is the trivial graph with one vertex and no
edges.
ni

●● The connected 1-regular graph is the graph with two vertices and one edge
connecting them.
●● The connected 2-regular graph with n vertices is the graph which consists of a
U

single n-cycle.
●● The 3-regular graphs must have an even number of vertices since the sum of
the degrees of the vertices is an even number.
●● In general regular graphs can be quite complicated. For example, there are
ity

nineteen 3-regular graphs with ten vertices.

●● The complete graph with n vertices Knis regular of degree n − 1.
m
)A
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 59

Notes

e
in
nl
O
ity
rs
ve
ni
U
ity
m
)A
(c

Amity Directorate of Distance & Online Education

 60 Graph theory and Combinatorics

Remark:
Notes

e
Let G be a graph with n >1 vertices. Then the following are equivalent:

in
(i) G is a tree.
(ii) G is a cycle-free and has n − 1 edges.
(iii) G is connected and has n − 1 edges.

nl
It shows that a finite tree T with n vertices must have n−1 edges.

O
ity
Definition: 3.38
If a graph G is a weighted graph (i.e.if there is a real number associated with each

rs
edge of G), then the weight of a spanning tree T of G is defined as the sum of the
weight of all the branches in T.
ve
Definition: 3.39
A spanning tree with the smallest weight in a weighted graph is called a shortest
spanning tree (or) shortest distance spanning tree (or) minimal spanning tree.
ni

Questions:
1. A tree is a _________connected graph.
a) cyclic
U

b) acyclic
c) a or b
ity

d) a and b
2. Every tree is a _________graph.
a) simple
b) cyclic
m

c) loop
d) parallel
)A

3. A tree with n vertices has__________edges.

a) n
b) n-1
(c

c) n-2
d) n+1

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 61

4. Every nontrivial tree has atleast two vertices of degree________.

Notes

e
a) zero
b) one

in
c) two
d) none

nl
5. Every connected graph contains a________tree.
a) cyclic

O
b) spanning
c) minimal
d) all

ity
6. A vertex with minimum eccenricity is the_______of tree.
a) center
b) length

rs
c) a or b
d) a and b
7. Every tree has either_________or two centers.
ve
a) zero
b) one
c) a or b
ni

d) a and b
8. A minimum-weight spanning tree is an_______tree.
U

a) optimal
b) minimum
c) maximum
ity

d) none
9. The shortest path problem is to the shortest paths from one vertex to all other
vertices in a_______ graph.
m

a) simple
b) tree
c) weighted
)A

d) none
10. The tree T is __________.
a) a binary tree
(c

b) a full binary tree

Amity Directorate of Distance & Online Education

 62 Graph theory and Combinatorics

c) a ternary tree
Notes

e
d) a full ternary tree

in
Answer:
1. b
2. a

nl
3. b
4. b

O
5. b
6. a
7. b

ity
8. a
9. c
10. b

Exercises:

rs
1. Define Decomposition graph with give an example and justify your answer.
ve
2. Prove that, a connected graph G is an Euler graph if and only if it can be
decomposed into circuits.
3. Explain complete graph.
ni

4. Show that, there is one and only one path between every pair of vertices in a
tree T.
5. In a graph G, there is one and only one path between every pair of vertices,
U

prove that G is a tree.

6. Show that, a graph G with n vertices, n-1 edges and no circuits is connected.
7. Prove that a graph is a tree if and only if it is minimally connected.
ity
m
)A
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 63

Unit- IV
Notes

e
Algorithm for shortest spanning tree:

in
Start from v1 and connect it to its nearest neighbour say vk. Now consider v1 and vk
as once a graph and connect this subgraph to its closest neighbour, let this new vertex

nl
be vi.we continue this process until all vertices have been connected by (n-1) edges. Let
us now illustrate this method of finding a shortest spanning tree.

O
ity
rs
ve
Shortest spanning tree in a weighted graph
− 10 16 11 10 17
10 − 9.5 ∞ ∞ 19.5
16 9.5 − 7 ∞ 12
ni

11 ∞ 7 − 8 7
10 ∞ ∞ 8 − 9
17 19.5 12 7 9 −
U

Theorem: 4.1
A spanning tree (of a given of a weighted connected graph G) is a shortest
ity

spanning tree of G if there exists no of the spanning tree of G at a distance of one from
T whose weight is smaller than of T.

Proof:
Suppose that T1 is shortest spanning tree.
m

There exist no spanning tree at a distance of one from T, whose weight is smaller
than that of T.
)A

Conversely,

Suppose that T1 is a spanning tree in G such that there exist no other spanning
tree in G such that there exist no other spanning tree at a distance of one from T1
whose weight is smaller than T1.
(c

To prove: T1 is a shortest distance spanning tree.

Suppose, T2 is also shortest distance spanning tree.

Amity Directorate of Distance & Online Education
 64 Graph theory and Combinatorics

Case: (i)
Notes

e
If T1 = T2

in
nl
O
Then T1 is the shortest spanning tree.

Case: (ii)

ity
If T1 ≠ T2

rs
ve
There exist atleast one edge which is in T2 but not in T1.

Let e be an edge in T2 but not in T1.construct a fundamental circuit in T1 by adding

the edge e to T1.
ni

But not all of the branch in T1 that forms the fundamental circuit which may be in T2.

Each branch in T1 that lies in the fundamental circuit has a weight smaller or equal
to that of e.
U

(i.e.) Weight of ek< weight of e 

→ (1)
For all edge ek in the fundamental circuit in T1 among all these edges in the circuit
there exist one edge which is not in T2 say bj from a fundamental circuit in T2 containing
ity

e by adding bj.

T2 is the shortest spanning tree.

ie.,)Weight of bj ≥ weight of e 
→ (2)
m

Eqn (1) is tree for all ek is in the fundamental circuit of T1.

bj is also one of the edges in the fundamental circuit.

)A

weight of bj ≤ weight of e 
→ (3)
From (2) and (3)

Weight of ej = weight of e
(c

Construct a new spanning tree T1 = T1  e-bj

Weight of T’1 = weight of T1

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 65

Repeat this producer by taking and soon and producing a series of tree of equal
Notes

e
weight T1,T1,T1,……. each a unit distance closure to T2 and we get T2 itself.

weight of T1 = weight of T2

in
Therefore T1is the shortest spanning tree.

Definition: 4.2

nl
In a connected graph G, a cutset is a set of edges whose removal from G leaves
G disconnected, provided removal of these edges disconnects G.

O
The set of edge {a,c,d,f} is a cutset. There are many other cutset such as
{a,b,g},{a,b,e,f} and {d,h,f} edge {k} alone as also a cutset .

The set of edges {a,c,h,d} is not a cutset . Because one of its proper subset {a,c,h}

ity
is a cutset.

rs
ve
Remark:
The another name of cutsets are minimal cutsets or proper cutsets or simple
ni

cutsets or cocycle.

Every edge of a tree is cutset.

U

Theorem: 4.3
Prove that the ringsum of any two cutsets in a graph is either a third cutsets or an
edge disjoint union of cutsets.
ity

Proof:
Let G be a connected graph graph and s1 and s2 be two cutset in the given
connected graph.
m

Let v1 and v2 be the unique and disjoint partitioning of the vertex set V of G
corresponding to s1.

Let v3 and v4 be the partitioning corresponding to s2, clearly,

)A

V1  V2 = V &

V1  V2 = φ
V3  V4 = V &
(c

V3  V4 = φ

Amity Directorate of Distance & Online Education

 66 Graph theory and Combinatorics

Notes

e
in
nl
O
ity
rs
ve
ni
U
ity
m
)A
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 67

Consider v5 = (v1  v4)  (v2  v3) Notes

e
= v1 ⊕ v3

in
V6 = (v1  v3)  (v2  v4)
=v2 ⊕ v4

nl
The ringsum of two cutsets s1 ⊕ s2 can be seen to consist only of edges that join
vertices in v5 to those in v6.

Also there are no edges outside s1 ⊕ s2 that joint vertices in v5 to those in v6.

O
Thus the set of edges s1 ⊕ s2 produces a partitioning of V into v5 and v6.

Such that,

ity
V5  V6 = V &
V5  V6 =φ

rs
ve
ni

Hence s1 ⊕ s2 is a cutset if the subgraphs containing v5 and v6 each remain

connected. After s1 ⊕ s2 is removed from G. otherwise s1 ⊕ s2 is an edge disjont union of
cutsets.
U

Definition:4.4
Consider a spanning tree T in a given connected graph G. Let ci be a chord with
respect to T, and let the fundamental circuit made by ci be called Γ consisting of k -
ity

branches b1,b2,b3,……..bk in addition to the chord ci.

i.e.,) Γ ={ci,,b1,b2…..bk} is a fundamental circuit with respect to T.

Theorem: 4.5
m

With respect to a given spanning tree T, a chord ci that determine a fundamental

circuit occurs in every fundamental cutset associated with the branches in Γ and in
v1 ∈ V1

no other.
)A

Proof:
Consider a spanning tree T of the given connected graph G.

Let ci be the chord with respect to T and Γ be the fundamental circuit consisting of
(c

k. branches of T together with ci.

Amity Directorate of Distance & Online Education

 68 Graph theory and Combinatorics

i.e.,) Γ =[ci,b1,b2,.….bk]
Notes

e
Where b1,b2,…..bk are branches of T. We know that every branches of a spanning
tree has a fundamental cutset associated with it.

in
Let s1 be the fundamental cutset associated with the branch b1.

Let s1={b1,c1,c2,….cq}

nl
Where c1,c2, . . . .cq are chords of T.

b1 ∈ s1  Γ

O
We know that N(s1  Γ ) is even

s1 and Γ must have exactly one more common edges and that should be ci

ity
⇒ ci =s1
By a similar argument we can prove that ci ∈ the cutsets s2,s3,. . . . sk respectively.

The chord ci is in every fundamental cutset.

rs
Now let s’ be the fundamental cutset other than s1,s2,.. . . sk.

Claim: ci ∉ s’ ⇒ ci ∈ s1

Assume the contrary

ve
i.e) ci ∈ s’

s’ = {ci,b’1,b’2,. . . b’k}

Now Γ n s’ ={ci}
ni

Which is a contrary to fact that there is an even no of edges in common to a circuit

and a cutset.
U

Ci ∉ s’

Theorem: 4.6
ity

With respect to a given spanning tree T, a branch bi that determine a fundamental

cutset S is contained in every fundamental circuit associated with the chords in S and in
no others.

Proof:
m

Let G be a given connected graph. Let S be a fundamental cutset associated with a

branch bi of T in G.
)A

Ie) S = {bi,c1,c2,. . . ck}

Let Γ 1 be a fundamental circuit associated with the chord c1.

Ie) Γ 1 = {c1,b1,b2,. . . bm}

(c

Since S is a cutset and Γ 1 is a circuit. There will be an even no of edges in

common to S and Γ 1.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 69

Obviously c1 belongs to both ε & Γ 1.

Notes

e
Therefore bi must belong to Γ 1.

in
By a similar argument we can prove that b1 belongs to every fundamental circuit
determine by the chords

C2,C3,. . . Ck.

nl
Claim: bi does not occurs in any other fundamental circuit.

Let Γ ’ be a fundamental circuit other than Γ 1, Γ 2,. . . Γ k.

O
Assume bi ∈ Γ .

Γ ’ ={c1,b1,. . . bi,bi+1,. . . bm}

ity
Γ ’  S ={bi}
Which is a contradiction to that Γ ’  S must have an even no of edges.

Then bi ∈ Γ ’.

rs
Therefore bi dose not occur in any other fundamental circuit.

Definition: 4.7
ve
The number of edges in the smallest cutset is defined as the edge connectivity of G.

Remark:
The edge connectivity of a tree is one.
ni

Definition: 4.8
The vertex connectivity of a connected graph G is defined as the minimum
U

number of vertices whose removal from G leaves the remaining graph disconnected.

Definition: 4.9
ity

A connected graph is said to be seperable. If its vertex connectivity is one. All

other connected graph are called non-seperable.

Definition: 4.10
m

In a seperable graph a vertex whose removal disconnects the graph is called a

cut-vertex a cut-node or and articulation point.
)A

Remark:
The edge connectivity and vertex connectivity of a disconnected graph is zero.

Every tree is seperable.

(c

Amity Directorate of Distance & Online Education

 70 Graph theory and Combinatorics

Theorem: 4.11
Notes

e
A vertex V in a connected graph G is cut vertex if there exist two vertices x and y in
G such that every path between x and y passes through V.

in
Proof:
Let v be cut vertex of G. Assume the contrary that is for every pare of vertices x

nl
and y there exist a path that not pass through V.

Therefore if v remove the vertex V the graph is connected.

O
⇒ we cannot a cut vertex.
Which is a contradiction.

Hence every path between x and y passes through V.

ity
Conversely,

Assume that there exist vertices x and y such that every path between x and y
passes through V.

rs
If we remove the vertex then there is no path between x and y.

G-{v} is disconnected.
ve
Therefore V is a cut vertex in G.

Applications:
Suppose we are given n stations that are to be connected by means of e lines
ni

(telephone lines, bridges, railroads, tunnels or highway), where e≥n-1. What is the
best way of connecting by best we m can that the network should be as invulnerable
to destruction are indidue station and individual line as possible. In other words,
construct a graph with ‘m’ vertices and ‘e’ edges that has the maximum possible edge
U

connectivity and vertex connectivity.

Example:
ity

The below graph has

m
)A

n=8, e=16 and has vertex connectivity of one and edge connectivity of three.
(c

Another graph the same number of vertices and edges can be drawn as shown in
the below figure.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 71

Notes

e
in
nl
O
It can be easily be seen that the edges connectivity as well as the vertex

ity
connectivity of this graph is four.

Consequently, even after any three station are bombed or any three lines
destroyed, the remaining station can still continue to communicate with each other.

Theorem: 4.12

rs
Prove that the Edge connectivity of a graph G cannot exceed the degree of the
vertex with the smallest degree in G.
ve
Proof:
Let Vi be the vertex with the smallest degree in G.
ni

Let d(Vi) = k (say)

There are k-edges incident on Vi and no other vertex will have number of edges
incident on it.
U

If we remove k- edges, the vertex v will be isolated.

That is, the graph will be disconnected.

ity

Edge connectivity ≤k
Edge connectivity ≤ d (v i )

Theorem: 4.13
m

The vertex connectivity of any graph G can never exceed the edge connectivity of G.

Proof:
)A

Let be the vertex connectivity and be the edge connectivity of G.

Since is the edge connectivity of G. There is a cutest S with edges

The removal of S from G partition S vertex set into two disjoint subsets.
(c

By removing at most vertices from v1 and v2 on which the edges in a incidents, all
the edges in S will be removed from G leaving it disconnected.

Amity Directorate of Distance & Online Education

 72 Graph theory and Combinatorics

As the vertex connectivity of a connected graph is the smallest number of vertices

Notes

e
whose removal disconnects the graph.

Therefore α≤β

in
Hence the theorem

nl
Theorem: 4.14
Every cutest in a non seperable graph with more than two vertices contains atleast
two edges

O
Proof:
Let G be a non – seperable graph with more than two vertices.

ity
Therefore Vertex connectivity ≥ 2

We know that, vertex connectivity ≤ edge connectivity

Edge connectivity ≥ vertex connectivity ≥ 2

rs
Therefore edge connectivity ≥ 2

Therefore every cutest in G must contain two or more edges.

ve
Theorem: 4.15
The maximum vertex connectivity one can achieve with a graph G of ‘n‘ vertices
and ‘e’ edges (e > n-1) is the integral part of the number 2e/n that is (2e/n)
ni

Proof:
Let G be a graph with ‘n’ vertices and ‘e’ edges
U

We know that
n

∑ d (v ) = 2e
i =1
i
ity

The total 2e degrees is divided among n vertices

Therefore there must be atleast one vertex in G whose degree is equal to or less
than the number 2e/n

The vertex connectivity of G cannot exceed this number.

m

Vertex connectivity ≤ smallest degree

2e
Vertex connectivity ≤
)A

n
Vertex connectivity ≤ edge connectivity <2e/n

To show that this value can actually be achieved, one can first construct an n
vertex regular graph of (2e/n)
(c

Definition: 4.16
A graph G is said to be k-connected if the vertex connectivity of G is k
Amity Directorate of Distance & Online Education
 Graph theory and Combinatorics 73

Therefore a one connected graph is the same as separable graph.

Notes

e
Remark:

in
●● A connected graph G is k-connected if every pair of vertices in G is joined by
k or more paths do not interested and at least one pair of vertices is joined by
exactly k-non intersecting paths.

nl
●● The edge connectivity of graph G is k if every pair of vertices in G is joined
by k or more edge disjoined paths (ie, paths that may intersect but have no
edges common) and at least one pair of vertices is joined by exactly k-edges

O
disjoined paths.

Theorem: 4.17
Prove that the ring sum of two circuits in a graph G is either a circuit or an edge

ity
disjoined union of circuits

Proof:
Let circuit 1 and circuit 2 be any two circuits in a graph G.

rs
If the two circuits have no edges or vertices common, their ring sum of circuit 1 ⊕
circuit 2 is a disconnected subgraph of G, and is obviously an edge disjoined union of
circuits.
ve
If, on the other hand circuit 1 and circuit 2 do have edges and vertices is common
we have the following possible situation.

Since the degree of every vertex in a graph (i.e.), circuit is two, every vertex ‘v’ in
ni

subgraph has degree d(v) where d(v) =2 , ‘v’ is in circuit 1 only or circuit 2 only or if
one of the edges formally incident on ‘v’ was in both and circuit 1 and circuit 2 or d(v)=
4 if circuit 1 and circuit 2 just intersect at ‘v’
U

There is no other type of vertex in circuit 1 ⊕ circuit 2.Thus circuit 1 ⊕ circuit 2

is an Euler graph ,and Therefore consist of either a circuit or an edge disjoined union of
circuits.
ity

Theorem: 4.18
If the graph Ghasnvertices and m edges, then the following statements are
equivalent:
m

(i) G is a tree.
(ii) There is exactly one path between any two vertices in G and G has no loops.
(iii) G is connected and m = n − 1.
)A

(iv) G is circuitless and m = n − 1.

(v) G is circuitless and if we add any new edge to G, then we will get one and only
one circuit.
(c

Proof:
(i)⇒(ii)

Amity Directorate of Distance & Online Education

 74 Graph theory and Combinatorics

Let G be a tree, then it is connected and circuitless.

Notes

e
Thus, there are no loops in G. There exists a path between any two vertices of G.
we know that there is only one such path.

in
(ii)⇒(iii)

G is connected. Let us use induction on m.

nl
By induction, If m= 0, G is trivial and the statement is obvious.

Induction Statement Proof:

O
Let e be an edge in G. Then G−e has ℓ edges. If G−e is connected, then there exist
two different paths between the end vertices of e

Therefore so (ii) is false. Therefore, G−e has two components G1 and G2.

ity
Let there be n1 vertices and m1 edges in G1. Similarly, let there be n2 vertices and
m2 vertices in G2. Then,

n = n1 + n2 and m = m1 + m2 + 1.

rs
The Induction Hypothesis states that

m1 = n1 − 1 and m2 = n2 − 1, so m=n1+n2−1 =n−1.

(iii)⇒(iv)
ve
Consider the hypothesis: There is a circuit in G. Let e be some edge in that circuit.
Thus, there are n vertices and n−2 edges in the connected graph G

(iv)⇒(v)
ni

If G is circuitless, then there is at most one path between any two vertices. If G has
more than one component, then we will not get a circuit when we can draw an edge
between two different components. By adding edges, we can connect components
U

without creating circuits

So G is connected.
ity

When we add an edge between vertices that are not adjacent, we get only one
circuit. Otherwise, we can remove an edge from one circuit so that other circuits will not
be affected and the graph stays connected, in contradiction to (iii)⇒(iv).

Similarly, if we add a parallel edge or a loop, we get exactly one circuit.

m

(v)⇒(i)

G is not a tree
)A

i.e. it is not connected. When we add edges as we did previously, we do not create
any circuits

Since spanning trees are trees, and it is also true for spanning trees.
(c

Theorem: 4.19
Prove that a connected graph has at least one spanning tree.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 75

Proof:
Notes

e
Consider the connected graph G with n vertices and m edges.

in
If m=n−1, then G is a tree. Since G is connected, m≥n–1.

If m≥ n, where there is a circuit in G.

We remove an edge e from that circuit. G − e is now connected.

nl
We repeat until there are n−1 edges.

Then, we are left with a spanning tree.

O
Remark.
●● We can get a spanning tree of a connected graph by starting from an arbitrary

ity
sub-forest M (as we did previously). Since there is no circuit whose edges are
all in M, we can remove those edges from the circuit which are not in M.
●● The subgraph G1 of G with n vertices is a spanning tree of G (thus G is
connected) if any three of the following four conditions hold:
(i) G1 has n vertices.
(ii) G1 is connected.
(iii) G1 has n − 1 edges. rs
ve
(iv) G1 is circuitless.

Theorem: 4.20
ni

If a tree is not trivial, then prove that there are at least two pendant vertices.

Proof:
U

If a tree has n(≥2) vertices, then the sum of the degrees is 2(n−1).

If every vertex has a degree ≥2, then the sum will be ≥2 n .

On the other hand, if all but one vertex have degree ≥2, then the sum would be ≥1
ity

+ 2(n−1) = 2n−1.

Remark:
A forest with k components is sometimes called a k-tree. (So a 1-tree is a tree.)
m

Example:
)A
(c

4-tree

Amity Directorate of Distance & Online Education

 76 Graph theory and Combinatorics

Theorem: 4.21
Notes

e
The edge set F of the connected graph G is a cut set of Gif and only if

in
(i) F includes at least one branch from every spanning tree of G, and
(ii) H⊂F, then there is a spanning tree none of whose branches is in H.

nl
Proof:
Let us first consider the case where F is a cut set.

Then, (i) is true.

O
If H⊂F then G−H is connected and has a spanning tree T. This T is also a spanning
tree of G.

Hence, (ii) is true.

ity
Let us next consider the case where both (i) and (ii) are true.

Then G−F is disconnected. If H⊂F there is a spanning tree T none of whose

branches is in H.

rs
Thus T is a subgraph of G − H and G − H is connected.

Hence, F is a cut set.

ve
Theorem: 4.22
Prove that the subgraph C of the connected graph G is a circuit if and only if

(i) C includes at least one link from every cospanning tree of G, and
ni

(ii) D is a subgraph of C and D=C, then there exists a cospanning tree none of
whose links is in D.
U

Proof:
Let us first consider the case where C is a circuit.

Then, C includes at least one link from every cospanning tree so (i) is true. If D is a
ity

proper subgraph of C, it obviously does not contain circuits, i.e. it is a forest.

We can then supplement D so that it is a spanning tree of G

i.e. some spanning tree T of G includes D and D does not include any link of T∗.
m

Thus, (ii) is true.

Now we consider the case where (i) and (ii) are both true.
)A

Then, there has to be at least one circuit in C because C is otherwise a forest and
we can supplement it so that it is a spanning tree of G.

We take a circuit C′ in C. Since (ii) is true, C′=C is not true, because C′ is a circuit
and it includes a link from every cospanning tree.
(c

Therefore, C=C′ is a circuit.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 77

Theorem: 4.23
Notes

e
Prove that a circuit and a cut set of a connected graph have an even number of
common edges.

in
Proof:
We choose a circuit C and a cut set F of the connected graph G.

nl
G−F has two components G1= (V1, E1) and G2= (V2, E2).

If C is a subgraph of G1 or G2, then the theorem is obvious because they have no

O
common edges.

Let us assume that C and F have common edges. We traverse around a circuit by
starting at some vertex v of G1.

ity
Since we come back to v, there has to be an even number of edges of the cut V1,
V2 in C.

Theorem: 4.24

rs
A fundamental circuit corresponding to links of the cospanning tree T∗ of a
connected graph is formed exactly by those branches of T whose corresponding
fundamental cut set includes c.
ve
Proof:
There exists a fundamental circuit C that corresponds to link c of T∗. The other
edges b1, . . . , bk of C are branches of T.
ni

We denote Ii as the fundamental cut set that corresponds to branch bi.. Then, bi is
the only branch of T which is in both C and Ii.
U

On the other hand, c is the only link of T∗ in C. By Theorem 2.6, we know that the
common edges of C and Ii are bi and c, equivalently c is an edge of Ii.

Then, we show that there is no c in the fundamental cut sets Ik+1, . . . , In−1 that
correspond to the branches bk+1, . . . , bn−1 of T .
ity

For instance, if c were in Ik+1, then the fundamental cut set Ik+1 and the circuit C
would have exactly one common edge.

So c is only in the fundamental cut sets I1, . . . , Ik.

m

Definition: 4.25
A spanning tree of a connected graph is a subtree that includes all the
)A

vertices of that graph. If it is a spanning tree of the graph G, then

G − T =def. T ∗is the cospanning tree.

Example:
(c

Amity Directorate of Distance & Online Education

 78 Graph theory and Combinatorics

Notes

e
in
nl
(G) Spanning tree

O
ity
Cospanning tree

rs
Theorem: 4.26
The fundamental cut set corresponding to branch b of the spanning treeT of a
connected graph consists exactly of those links of T∗ whose corresponding fundamental
ve
circuit includes b.

Proof:
Let I be a fundamental cut set that corresponds to the branch b of T. Other edges
ni

c1, . . . , ck of I are links of T ∗.

Let Ci denote the fundamental circuit that corresponds to ci. .Then, ci is the only link
of T∗ in both I and Ci.
U

On the other hand, b is the only branch of T, the common edges of I and Ci are b
and ci, in other words, b is an edge of Ci.

Then, we show that the fundamental circuits Ck+1, . . . , Cm−n+1corresponding to the

ity

links ck+1, . . . , cm−n+1do not include b.

For example, if b were in Ck+1, then the fundamental circuit

Ck+1and the cut set I would have exactly one common edge.
m

Hence, the branch b is only in fundamental circuits C1, . . . , Ck.

Minimum spanning trees of a weighted graph:

)A

The minimum spanning tree of a weighted graph constructed using

●● Prim’s algorithm (using vertex)

●● Kruskal’s algorithm (using edges)
(c

Algorithm for prim’s algorithm:

●● Step:1 Select any connected vertices with minimum weight

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 79

●● Step: 2 Select the unvisited vertex which is adjacent of the visited vertices
Notes

e
with minimum weight
●● Step: 3 Repeat step:2 until all vertices are visited

in
Problem: 1
Find the minimum spanning tree using prim’s algorithm

nl
O
ity
Solution:

Step: 1
Let us take the vertex ‘a’

rs
ve
S={a}

The remaining vertices={b,c,d,e,f,g}

Starting vertex A={ φ }

ni

The lighted edges={a,b}

U

Step: 2
Let S={a,b}
ity
m

S={a,b}

)A

The remaining vertices={c,d,e,f,g}

Starting vertex A={{a,b}}

The lighted edges are bd or ac and both have the weight 8.
(c

But we have to visit the weight from an adjacent to the unvisited vertex

So we have to select the vertex ‘b’

Amity Directorate of Distance & Online Education

 80 Graph theory and Combinatorics

Step: 3
Notes

e
in
nl
S={a,b,d}

The remaining vertices={c,e,f,g}

O
Starting vertex A={{a,b},{b,d}}

The lighted edges are dc

ity
Step: 4

rs
ve
ni

S={a,b,c,d}

The remaining vertices={e,f,g}

U

Starting vertex A={{a,b},{b,d},{d,c}}

The lighted edges are cf

ity

Step: 5
m
)A
(c

S={a,b,c,d,f}

The remaining vertices={e,g}

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 81

Starting vertex A={{a,b},{b,d},{d,c},{c,f}}

Notes

e
The lighted edges are fg

in
Step: 6

nl
O
ity
S={a,b,c,d,f,g}

rs
The remaining vertices={e}

Starting vertex A={{a,b},{b,d},{d,c},{c,f},{f,g}}

The lighted edges are ge

ve
Step: 7
ni
U
ity

S={a,b,c,d,e,f,g}

The remaining vertices={ φ }

Starting vertex A={{a,b},{b,d},{d,c},{c,f},{f,g},{g,e}}

m

The lighted edges are ge

Now the minimum spanning tree is completed

)A

Problem: 2
Find the minimum spanning tree using kruskal’s algorithm
(c

Amity Directorate of Distance & Online Education

 82 Graph theory and Combinatorics

Notes

e
in
nl
O
Step: 1

ity
Remove all loops and parallel edges

rs
ve
ni

Step: 2
Arrange all edges in their increasing order of weight

(i) BD=2
U

(ii) D T=2
(iii) AC=3
ity

(iv) CD=3
(v) CB=4
(vi) BT=5
(vii) AB=6
m

(viii) SA=7
(ix) SC=8
)A

Step: 3
Add the edge which has least weightage
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 83

Notes

e
in
nl
Next least cost is 3, so the edges are AC or CD

O
ity
Next least cost is 4, but if we take 4 it becomes closed circuit so we discard to
avoid circuit.

rs
So next least cost is 5, and it becomes also closed circuit. So discard it
ve
And next least cost is 6, and it becomes also closed circuit. So discard it

Next least cost is 7 and 8

ni
U
ity

Therefore the minimal spanning tree is completed.

Questions:
1. The number of internal vertices of T is________.
m

a) 6
b) 7
)A

c) 8
d) 9.
2. The number of vertices of T at level 1 or 2 is_________
a) 2
(c

b) 4

Amity Directorate of Distance & Online Education

 84 Graph theory and Combinatorics

c) 6
Notes

e
d) 8
3. The number of binary trees with 5 vertices is ______

in
a) 2
b) 3

nl
c) 4
d) 5

O
4. The number of rooted trees with three vertices is________
a) 2
b) 3

ity
c) 4
d) 5
5. The number of rooted binary trees with 5 vertices is_________

rs
a) C(10, 5)
b) C(10, 5)/6
c) C(10,5)/5
ve
d) C(10,5)/10
6. The number of chords of a spanning tree T of a connected graph G with n vertices
and e edges is__________.
ni

a) e-1
b) n-1
U

c) n-e+1
d) e-n+1
7. The number of spanning trees of a connected graph with n vertices and n edges
ity

is_________.
a) n
b) n-1
c) n+1
m

d) none of these
8. The number of chords of a spanning tree T of a connected graph G with n vertices
)A

and e edges is__________.

a) e-1
b) n-1
c) n-e+1
(c

d) e-n+1

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 85

9. The number of spanning trees and minimal spanning trees of the complete graph
Notes

e
with 4 vertices with distinct weight is_________.
a) 16 and 16

in
b) 16 and 1
c) C(6,3) and 1

nl
d) C(6,3) and C(6,3)
10. The chromatic number of a tree with n vertices is________.
a) n

O
b) n-1
c) 3

ity
d) 2
11. A tree with 3 or more vertices remains a tree when ________
a) any vertex is removed

rs
b) any edge is removed
c) a pendant vertex is removed
d) none of these
ve
12. A tree with 5 vertices can be ________.
a) D5
b) P5
ni

c) C5
d) W5
U

13. Every tree is__________

a) bipartite
b) not a bipartite
ity

c) a or b
d) a and b
14. A full binary tree with n vertices has________ leaves.
m

a) n
b) (n+1)/2
c) n-1
)A

d) (n-1)/2
15. The number of vertices n in a full binary tree is always_______.
a) odd
(c

b) even

Amity Directorate of Distance & Online Education

 86 Graph theory and Combinatorics

c) a or b
Notes

e
d) a and b

in
Answer:
1. b
2. c

nl
3. d
4. b

O
5. b
6. d
7. d

ity
8. d
9. b
10. d
11. c
12. b
rs
ve
13. a
14. b
15. a
ni

Problems:
1. Show that There exists a least number f(s,t) such that the vertices of any simple
digraph with minimum outdegree at least f(s,t) can be partitioned into two
U

sets inducing subdigraphs with minimum outdegree at least s and at least t,

respectively. [It is known that f(1,1)=3.
2. Prove that every tournament of order 2n-2 contains every oriented tree of order
ity

n)
3. What is the maximum cyclic edge-connectivity of a 5-connected planar graph?
4. Does every planar subgraph have an induced forest with at least half of the
vertices?
m
)A
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 87

Module IV
Notes

e
Key Learning Objectives:

in
At the end of this module, you will be able to:

●● Describe directed graphs and directed trees

nl
●● Discuss the concept of connectedness
●● Practise studying network flows

O
●● Interpret the max flow - min cut theorem
●● Prepare the matrix representation of a graph
●● Describe planar graphs

ity
●● Compare combinational and geometric duals
●● Prepare Kuratowski’s graphs
●● Detect planarity

rs
●● Study thickness and crossing
ve
ni
U
ity
m
)A
(c

Amity Directorate of Distance & Online Education

 88 Graph theory and Combinatorics

Unit- V: Directed graphs

Notes

e
Definition: 5.1

in
A graph G is said to be a directed graph or digraph it consists of vertex set and
the edge set V={v1 ,v2,…..} and E={e1 ,e2,…..} such that every pair of vertices representing

nl
any edge by an ordered or a direct pair (v1,v2) where v1 is the tail and v2 is the head of
the edge

O
Fig 3.1 b and 3.1c represents the directed graph

ity
rs
ve
ni

Properties:
U

(i) Number of edges incident out of vertex vi, d+(vi) is said to be out degree or out
valene
(ii) Number of edges incident into of vertex vi, d-(vi) is said to be in degree or in
valene
ity
m
)A

d+(v1) = 2 d-(v1) = 0
d+(v2) = 1 d-(v2) = 1
(c

d+(v3) = 0 d-(v3) = 2
d+(v4) = 1 d-(v4) = 1

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 89

(iii) For an isolated vertex in degree = out degree = 0

Notes

e
(iv) In Pendent vertex d+(vi) +d-(vj) = 1

in
nl
O
(v) In a simple digraph there are no self loops and no parallel edges

ity
(vi) In a Asymmetric digraph that have atmost one directed edge between the pair
of vertices and that are allowed to have self loops

rs
ve
ni

(vii) In a symmetric digraph if there is a vertex from ‘a’ to ‘b’ and the same for vertex
‘b’ to vertex ‘a’, it satisfies for all vertex ‘ab’
U
ity
m
)A

(viii) In complete symmetric digraph is a simple digraph in which there is exactly one
edge directed from each vertex to every other vertex
(c

Amity Directorate of Distance & Online Education

 90 Graph theory and Combinatorics

Notes

e
in
nl
(ix) In a complete asymmetric digraph is also a simple graph there is an exactly one

O
edge between a pair of vertices. It is also called as a tournament.

ity
(x) A digraph is balanced for all vi indegree = outdegree .

Definition: 5.2

rs
A directed walk from a vertex vi to an vj is an alternating sequence of vertices and
edges, beginning with vi and ending with vj such that each edge is oriented form the
ve
vertex preceding it to the vertex following it (no edge retracting)

Definition: 5.3
ni

A semi walk in a directed graph is a walk in the corresponding undirected graph

Definition: 5.4
U

A directed circuit is a closed circuit starting from vi and ending on vi such that each
edge is oriented from the vertex preceding it to the vertex following it
ity

Definition: 5.5
A semi circuit is a closed walk which the orientation is not considered
m
)A
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 91

In the above figure,

Notes

e
●● Directed walk: v1 e2 v4 e3 v1 e4 v2
●● Semi walk: v1 e2 v4 e7 v5

in
●● Directed circuit: v1 e1 v4 e6 v4 e3 v1
●● Semi circuit: v1 e2 v4 e7 v5 e9 v3 e1 v1

nl
Types of connectedness:

Strongly connected graph:

O
A graph is said to be strongly connected if there is at least one directed path from
one vertex to every other vertex

ity
Weakly connected graph:
rs
A graph is said to be weakly connected if its corresponding undirected graph is
ve
connected
ni
U

Definition: 5.6
A graph is said to be strongly connected fragment, within each component of G, the
ity

maximal strongly connected subgraph is called as fragments

m
)A
(c

Amity Directorate of Distance & Online Education

 92 Graph theory and Combinatorics

Definition: 5.7
Notes

e
The condensation G of a digraph G is a digraph in which each strongly connected
fragment is replaced by a vertex and all directed edge from one strongly connected

in
component to another are replaced by a single directed edge

Remark:

nl
●● The condensation of strongly connected graph is simply a vertex.
●● The condensation of a digraph has no directed circuit.

O
ity
rs
ve
Definition: 5.8

A diagraph D is said to be euler if it contains a closed walk which traverses every

edge of D exactly once, such a walk is called a euler walk. A digraph D is said to be
ni

unicursal if it contains an open euler walk.

U
ity

In the above figure the euler line is a b c d e f

m

Network Flows
)A

Definition: 5.9
A directed network D, with each arc (i, j) assigned a positive real number cij called
the capacity of the arc, and two distinguished vertices called a sink [t] and a source (s),
a second set of non-negative real numbers, xij, assigned to the arcs is called an (s, t)
(c

-feasible flow if 0 <= xij<= cij for all arcs, and n xij n = xji for all vertices i, except s
and t, if there is no arc (i, j) then xij = 0. ∑ ∑
j =1
j =1

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 93

Definition: 5.10
Notes

e
A partition of the vertices of the directed network D into two sets, one containing
s, called S, and the other containing t, called T, is called a cut. A cut can be determined

in
from, and is usually identified with, the set of arcs starting in S and ending in T. The
capacity of a cut is the sum of the capacities on these arcs. A minimum cut is a cut with
the minimum capacity.

nl
Remark:
●● The value of any feasible flow is less than or equal to the capacity of any cut.

O
●● If the value of some feasible flow equals the capacity of some cut, then the
flow is a maximum flow and the cut is a minimum cut.
●● For any arc (i, j) of the directed network, the slack, sij = cij - xij.

ity
Algorithm:
●● A simple algorithm for finding a larger flow is to find a path from s to t such
that all the arcs on the path have positive slack.
●● We can add the smallest slack on these arcs to each xij of the path to obtain a

●●
rs
larger flow. Unfortunately, this procedure does not always lead to a maximum flow.
Now consider the underlying graph of the directed network. Given a chain
from s to t in this graph, orient the all the edges of the chain from s to t.
ve
●● Then compare this orientation with the orientation that these arcs have in the
original network. If the orientation is the same for an arc, it will be called a
forward arc, if the orientation is opposite, the arc is called a backward arc.
●● A flow augmenting chain in the directed network, is a chain from s to t in
ni

the underlying graph, so that each forward arc has positive slack and each
backward arc has a positive flow.
●● Given a flow augmenting chain, a larger flow is obtained by taking the
U

smallest value of the positve slacks on forward arcs and the positve flows
on backward arcs and adding this value to the flow on all forward arcs and
subtracting it from the flow on all backward arcs of the chain.
ity

Note:
 A feasible flow is maximum if and only if there are no flow augmenting
chains in the network.
 The value of the maximum flow of a directed network is the capacity of a
m

minimum cut.

Labeling Algorithm:
)A

●● Some of the arcs of the directed network will be considered as members of

two sets, F and B.
●● F consists of all arcs that have positive slack and B consists of all arcs that
have a positive flow. Notice that these sets are not disjoint.
(c

●● The algorithm proceeds by scanning vertices and labeling some of them in the
scanning process.

Amity Directorate of Distance & Online Education

 94 Graph theory and Combinatorics

●● When all the labeled vertices have been scanned, the algorithm stops. If t has
Notes

e
been labeled, then an augmenting chain can be found, otherwise there is no
augmenting chain.

in
●● Initially, the vertex s is labeled with the label
●● A labeled, but unscanned vertex, call it i.
●● Scan i, meaning

nl
1. for all arcs (i,j), if (i,j) is in F and j is unlabeled, then label j with (i+).
2. for all arcs (j,i), if (j,i) is in B and j is unlabeled, then label j with (i-).

O
●● Repeat until all labeled vertices have been scanned, or t has been labeled.
●● If t is labeled, then the augmenting chain is determined by starting with t and
following the labels back until you get to s.

ity
●● The +’s and -’s show whether the arc is a forward or backward arc.

Flows and Matchings in Bipartite Graphs

In a bipartite graph (X, Y) we can form a directed network as follows:

rs
●● Add a vertex s, and join it with arcs going from s to each vertex in X. Give all
of these arcs a capacity of 1.
●● Connect all the edges in the bipartite graph from X to Y.
ve
●● Give all of these arcs an infinite capacity.
●● Finally, add a vertex t, and join it with arcs from each vertex in Y to t. Note all
of these arcs a capacity of 1and say this directed network N.
●● Notice that any feasible flow of N where all the flow values are integers
ni

(necessarily either 0 or 1), corresponds to a matching of the original graph

(the matching consists of the edges which correspond to arcs between X and
Y which have positive flow), and vice versa, any matching of the graph gives
U

rise to a feasible flow.

●● The value of the flow is the number of edges in the matching.

Direct relation between flows and matchings:

ity

A covering of the graph corresponds to in the directed network.

Let A be a subset of X and B a subset of Y. We will consider the set K = A U B. Let

S = {s} U (X-A) U B and T = {t} U (Y-B) U A, which is clearly a cut in the network.
m

Theorem: 5.11
‘K’ is a covering of the graph if and only if (S,T) is a cut of the network of finite capacity.
)A

Proof:
Let ‘K’ be a covering of the graph.

Suppose (S,T) is a cut of finite capacity.

(c

Then no arcs from X to Y are in this cut, in particular there are none from X-A to
Y-B in the cut.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 95

So, all arcs from A go to Y-B, and all arcs from X-A go to B.
Notes

e
Therefore, every edge of the graph has one vertex in A or one vertex in B, and so,
K is a covering.

in
The arcs in the cut that start at s all go to A, and those that end at t all start in B,
each having capacity 1, so the capacity of the cut is |A U B|.

nl
Now, suppose that K is a covering of the graph. Let A be the intersection of K with
X and B the intersection of K with Y.

Then Form S and T as before.

O
Consider an arc from a vertex in S to one in T. If this arc starts at s, it must go to A.
If it starts in X-A, it must go to B since K is a covering, so such an arc does not go to T.
If it starts in B, then it must go to t, and so, has capacity 1.

ity
Thus, all arcs from S to T have finite capacity

So the cut (S, T) has finite capacity.

Finally, for a bipartite graph, the size of a maximum matching equals the size of a

rs
minimum cover.

Flows and Cuts in Networks

ve
A capacitated single source-single sink network is a directed graph D, with
each arc (i, j) assigned a positive real number cij called the capacity of the arc, and two
distinguished vertices called a sink (t) and a source (s). The source has out-degree of
at least one and the sink has in-degree of at least one. In almost all of the examples we
ni

consider the source will have in-degree of 0 and the sink will have out-degree of 0, but
this is not required in the definition. We shall refer to these as (s,t)-networks.

Given an (s,t)-network, a second set of non-negative real numbers, xij, assigned to

U

the arcs is called an (s,t) - feasible flow if

1. 0 xij cij for all arcs, and

2. xij = xji for all vertices i, except s and t
ity

j j

(Where it is understood that if there is no arc (i, j) then xij = 0.).

m
)A
(c

At s and t, the conservation law (the second condition above) does not hold, and
we define

Amity Directorate of Distance & Online Education

 96 Graph theory and Combinatorics

vt = xjt - xtj and vs = xsj - xjs.

Notes

e
j j j j

Now, if we sum over all the vertices i, we get:

in
i
( j
xij - j
xji ) = vs - vt.

But the left hand side of this equation is 0, so vs = vt and we call this common value
v, and define it to be the value of the flow.

nl
For the above figure, the value of that flow is 3.

O
Definition: 5.12
A partition of the vertices of the (s,t)-network D into two sets, one containing
s, called S, and the other containing t, called T, is called a cut. A cut can also be
determined from, and is usually identified with, the set of arcs starting in S and ending

ity
in T. The capacity of a cut is the sum of the capacities on these arcs. A minimum cut
is a cut with the minimum capacity.

Remark:
●●
●●
rs
The value of any feasible flow is less than or equal to the capacity of any cut.
If the value of some feasible flow equals the capacity of some cut, then the
flow is a maximum flow and the cut is a minimum cut.
ve
●● For any arc (i,j) of the (s,t)-network, its slack is defined as, sij = cij - xij.
●● A simple algorithm for finding a larger flow is to find a path from s to t such
that all the arcs on the path have positive slack.
ni

●● We can add the smallest slack on these arcs to each xij of the path to obtain a
larger flow. Unfortunately, this procedure does not always lead to a maximum flow.
●● Consider the underlying graph of the directed network. Given a path from s to
U

t in this graph, orient the all the edges of the path from s to t.
●● Now, compare this orientation with the orientation that these arcs have in the
original network. If the orientation is the same for an arc, it will be called a
forward arc, if the orientation is opposite, the arc is called a backward arc.
ity

●● A flow augmenting path in the (s,t)-network, is a path from s to t in the

underlying graph, so that each forward arc has positive slack and each
backward arc has a positive flow.
●● Given a flow augmenting path, a larger flow is obtained by taking the
m

smallest value of the positive slacks on forward arcs and the positive flows
on backward arcs and adding this value to the flow on all forward arcs and
subtracting it from the flow on all backward arcs of the path.
)A

Labeling Algorithm for Finding Flow Augmenting Paths:

●● Some of the arcs of the (s,t)-network will be considered as members of two
sets, F and B.
●● F consists of all arcs that have positive slack and B consists of all arcs that
(c

have a positive flow. Notice that these sets are not disjoint. A forward arc in F
or a backward arc in B is called a usable arc.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 97

Matchings, Transversals and Vertex Covers:

Notes

e
A set M of edges in a graph G is called a matching of G if no two edges in set
M have an endpoint in common. A maximum matching of graph G is a matching of G

in
with the greatest number of edges while a maximal matching is a matching which is not
contained in any larger matching. While any maximum matching is certainly maximal,
the reverse is not generally true.

nl
O
In the above graph, {a}, {b,d}, and {c,e} are all maximal matchings, but only the last
two are maximum matchings.

ity
Job Assignment Problem revisited:
Given a bipartite graph (X, Y) we can form an (s,t)-network as follows:

●● Add a vertex s, and join it with arcs going from s to each vertex in X. Orient all

●●
the edges in the bipartite graph from X to Y.

rs
Finally, add a vertex t, and join it with arcs from each vertex in Y to t. Give all
of the arcs leaving s or entering t a capacity of 1, but the arcs between X and
ve
Y will get infinite capacity. Call this network N.
●● Note that any feasible flow of N where all the flow values are integers
(necessarily either 0 or 1), corresponds to a matching of the original graph
(the matching consists of the edges which correspond to arcs between X and
ni

Y which have positive flow), and vice versa, any matching of the graph gives
rise to a feasible flow.
●● The value of the flow is the number of edges in the matching.
U

Definition: 5.13
Let G be a graph and C a subset of the vertices of G. The set C is a vertex cover
ity

of graph G if every edge of G is incident with at least one vertex in C. A minimum vertex
cover is a vertex cover with the least number of vertices.
m
)A

Theorem: 5.14
K is a vertex covering of the graph if and only if (S,T) is a cut of the network of
(c

finite capacity.

Amity Directorate of Distance & Online Education

 98 Graph theory and Combinatorics

Proof:
Notes

e
Suppose (S,T) is a cut of finite capacity.

in
Then no arcs from X to Y are in this cut, in particular there are none from X-A to
Y-B in the cut.

So, all arcs from A go to Y-B, and all arcs from X-A go to B. Therefore, every edge

nl
of the graph has one vertex in A or one vertex in B, and so, K is a covering.

The arcs in the cut that start at s all go to A, and those that end at t, all start in B,
each having capacity 1, so the capacity of the cut is |A B|.

O
Suppose that K is a vertex covering of the graph. Let A be the intersection of K with
X and B the intersection of K with Y.

Form S and T as before. Consider an arc from a vertex in S to one in T. If this arc

ity
starts at s, it must go to A (and so, has capacity 1). If it starts in X-A, it must go to B
since K is a vertex covering, so such an arc does not go to T.

If it starts in B, then it must go to t, and so, has capacity 1. Thus, all arcs from S to
T have finite capacity, so the cut (S, T) has finite capacity.

Theorem: 5.15
rs
Let G be a bipartite graph. Then the size of a maximum matching equals the size
ve
of a minimum vertex cover.

Proof:
Let G be a bipartite graph.
ni

Consider the associated network of the bipartite graph.

Change the infinite capacities to 2’s, this will not change any flows.
U

All capacities are integers, so the maximum flow exists and will have integer
values.

The value of this flow is the number of edges in a maximum matching, call it n.
ity

Let ‘n’ is the minimum capacity of any (S, T) cut.

The trivial cut, S = {s}, has finite capacity, so the minimum will have to be finite. Any
cut with the minimum capacity corresponds to a covering with |A B| = n. But now,
m

there exists a cover with n vertices and a matching with n edges

So the cover must be a minimum cover.

)A

Matrix Representation of graphs

Introduction:
A matrix is a convenient and useful way of representing a graph to a computer.
(c

Matrices lend themselves easily to mechanical manipulations. Many known results of

matrix algebra can be applied to the study of structural properties of graph.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 99

Definition:5.16
Notes

e
Let G be a graph with ‘n’ vertices, ‘e’ edges and no self loops. Define an n by
‘e’ matrix A=[aij] whose ‘n’ rows correspond to the ‘n’ vertices and the ‘e’ columns

in
correspond to the ‘e’ edges .

The matrix element

nl
aij = 1, if the ‘j’th edge ej is incident on the ‘i’th vertex vi and

aij = 0, otherwise

O
ity
rs
ve
ni

Edges

0 0 0 1 0 1 0 0
U

0
 0 0 0 1 1 1 1
Vertices 0 0 0 0 0 0 0 1
 
1 1 1 0 1 0 0 0
ity

0 0 1 1 0 0 1 0
 
1 1 0 0 0 0 0 0
m

Graph and incidence matrix

●● In the above graph, such a matrix is called the vertex –edge incidence matrix
or simply matrix.
)A

●● The incidence matrix contains only two elements 0 and 1.

●● Such a matrix is called binary matrix or a (0,1) matrix.

Observations:
(c

●● Each edge is incident on exactly two vertices, each column of ‘A’ has exactly
two 1’s

Amity Directorate of Distance & Online Education

 100 Graph theory and Combinatorics

●● The number of 1’s in each row is equals the degree of the corresponding
Notes

e
vertex
●● A row with all 0’s therefore represents an isolated vertex.

in
●● Parallel edges in a graphg produce identical columns in its incidence matrix

Definition: 5.17

nl
Each row in an incidence matrix A(G) may be regarded as a vector over in the
vector space of a graph G

O
 A1 
A 
 2
 A3 
A(G ) =  . 
 

ity
 . 
 
 An 
Definition: 5.18
Let ‘g’ be a subgraph of a graph ‘G’ and let A(g) and A(G) be the incidence

rs
matrices of ‘g’ and ‘G’ respectively. A(g) is a submatrix of A(G). There is a non-to-one
correspondence between each ‘n’ by ‘k’ submatrix of A(G) and a subgraph of G with ‘k’
edges of vertices in G.
ve
Theorem: 5.19 (Matrix Tree Theorem)
Given a loopless graph G with vertex set v1 , v 2 ,.......v n , let− aϋij be the number of
*
edges having vi and v j as their end vertices. Let Q be the matrix with entry (i, j)
ni

*
being − a ij where i ≠ j and d (vi ) when i = j . If Q is the matrix obtained by
th
deleting any one row, say sth row and any one column, say t column of Q, then
τ (G ) = (−1) det Q . [i, e., τ (G ) is a confactor Ast of G ] .
s +t *
U

Example:

Now we use the Matrix Tree Theorem to fine τ (G ) of the graph G given in Figure
ity
m
)A

 1 −1 0 0 0 0 
 
−1 3 −1 −1 0 0 
 0 −1 3 −1 −1 0 
 
 0 −1 −1 3 −1 0 
 0 0 − 1 − 1 3 − 1
(c

 
 0 0 0 0 − 1 − 1
 
matrix Q
Amity Directorate of Distance & Online Education
 Graph theory and Combinatorics 101

We omit the second row and the second column to find the confactor A22
Notes

e
1 0 0 0 0
0 3 -1 -1 0

in
A22 = (-1) 2+2 0 -1 3 -1 0
0 -1 -1 3 -1
0 0 0 -1 1

nl
3 -1 -1 0
1 -1 3 -1 0 + 0 + 0 + 0 + 0 = 8

O
-1 -1 3 -1
Problem: 1

Find τ (G ) for the graph G given to figure

ity
Solution:
rs
ve
The matrix Q is
4 −2 −2 0 
 
 2 5 −1 − 2
−
 − 2 −1 5 − 2
 
ni

 0 −2 −2 4 
 

5 −1 − 2
U

τ (G ) = the cofactor A11 2

= (−1) − 1 5 − 2
−2 −2 4

= 5(20-4)+1(-4-4)-2(2+10)
ity

= 80-8-24

= 48
m

Theorem: 5.20 (Cayley’s Formula)

Prove that if n ≥ 2,τ (k ) = n n − 2
)A

Proof:
Let the vertices of the complete graph k n be labeled with integers 1,2,......n.
n−2
We note that there are n sequences of length n-2 with entries from ( 1,2,......n. ).
Denote the set of all such sequence by S.
(c

We establish a bijection between the set of all spanning trees of k n and the set S.
Using the labels of the vertices, we consider the vertex set v( k n ) as an ordered set.

Amity Directorate of Distance & Online Education

 102 Graph theory and Combinatorics

Notes To each spanning tree T, we associate a sequence f (T ) = (t1 , t 2 ,........t n − 2 ) as

e
follows. Let s1 be the first vertex of degree one in T and t1 be the unique vertex which
is adjacent to s1 in T. Now T − s1 is a tree. Let s 2 be the first vertex of degree one in

in
T − s1 and t 2 be the unique vertex which is adjacent to s 2 in T − s1 . Let s3 be the
first vertex of degree one in T − {s1 , s 2 } . After n − 2 iterations a single edge remains.
Thus we have produced a sequence f (T ) = (t1 , t 2 ,........t n − 2 ) of length n − 2 . For

nl
example for the spanning tree T given in the figure 40, the associated sequence f(T) is
(3,3,1,3,3,5,1,1).

O
ity
Let rs
Next we define a function of the produces a tree from each sequence x.

x = (t1 , t 2 ,....., t n − 2 ) be the given sequence.

ve
First we begin with n isolated vertices (and edge set E = Φ ).

Let s1 be the least number that does not appear in (t1 , t 2 ,....., t n ) .
Join the vertices s1 and t1 by an edge and set E = {s1 , t1 }.
ni

Let s2 ¹ s1 be the least number that is not in (t 2 , t 3 ,....., t n ) .

Join the vertices s 2 and t 2 by an edge and let E = {s1t1 , s 2 t 2 }.
U

Let s3 be the least number such that s3 Ï {s1 , s2 } and s3 is not in .

Join s3 and t 3 by an edge and set E = {s1t1 , s 2 t 2 , s3 t 3 }.
Continue in this way until we get n − 2 edges s1t1 , s 2 t 2 ,......., s n − 2 t n − 2 .
ity

T is now obtained by adding the edge joining the remaining two vertices of
{1,2,..., n} − {s1 , s 2 ,..., s n − 2 }.
The resulting graph is denoted by g (t1 , t 2 ,....t n − 2 ).
m

At each step we add one edge, there are n-1 edges including one that is added at
the last step.

After the i th step in construction of g(x)

)A

we have n-i components.

So after adding the edge at the last step.

We get a connected graph so g(x) is a connected graph with n vertices and n-1
(c

edges and hence it is a spanning tree.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 103

Questions:
Notes

e
1. The Konigsberg bride problem was a long-standing problem until solved by euler
in_________by means of a graph.

in
a) 1930
b) 1936

nl
c) 1940
d) 1946
2. A graph G is said to be eulerian if it contains an ___________.

O
a) euler tour
b) euler circuit

ity
c) a or b
d) a and b
3. If degree of every vertex is even in graph G, then 2(G) and hence G contains
atleast___________cycle.
a) zero
b) one rs
ve
c) two
d) three
4. If a closed trail C contains an edge e, then it contains a__________containing e.
ni

a) cycle
b) not a cycle
c) a or b
U

d) a and b
5. An open trial that contains every edge of G exactly__________is an open euler
trial.
ity

a) once
b) two
c) three
m

d) none
6. A cycle that contains every vertex of G is a Hamilton__________of G.
)A

a) path
b) walk
c) cycle
d) none
(c

Amity Directorate of Distance & Online Education

 104 Graph theory and Combinatorics

7. The complete graph Kn on n vertices is hamiltonian.

Notes

e
a) 3n
b) 2 < n

in
c) 2n
d) all

nl
8. The complete bipartie graphs Km;n (2 m; n) are hamiltonian if and only if m is equal
to__________.
a) n-1

O
b) n
c) n+1

ity
d) none
9. The graph K3;3 is __________.
a) planar

rs
b) non-planar
c) a or b
d) a and b
ve
10. The Euler’s formula for any connected plane graph is_________.
a) n+e+f=2
b) n-e-f=2
ni

c) n-e+f=2
d) all
U

Answer:
1. b
2. a
ity

3. b
4. a
5. a
m

6. c
7. a
)A

8. b
9. b
10. c
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 105

Problems:
Notes

e
1. Write down the properties of Incidence Matrix.
2. Prove that, the reduced incidence matrix of a tree is non-singular.

in
3. If B is a circuit matrix of a connected graph G with e edges and n vertices, prove
that rank of B = e-n + 1.

nl
4. Define Cut-set Matrix with give an example.
5. Explain path matrix.

O
ity
rs
ve
ni
U
ity
m
)A
(c

Amity Directorate of Distance & Online Education

 106 Graph theory and Combinatorics

Unit-VI: Planar and Dual Graphs

Notes

e
We studied properties of subgraphs, such as paths, circuits, spanning trees, and

in
cut-sets, in a given connected graph G.

In this chapter we shall subject the entire graph G to the following important
question: It is possible to draw G in a plane without its edges crossing over

nl
The question of planarity is of great significance from a theoretical point of view. In
addition, planarity and other related concepts are useful in any practical situations. For
instance, in the design of a printed-circuit board, the electrical engineer must know if he

O
can make the required connections without an extra layer of insulation. The solution to
the puzzle if three utilities, posed in Chapter 1, requires the knowledge of whether or
not the corresponding graph can be drawn in a plane.

ity
Combinatorial Versus Geometric Graphs
As mentioned in Chapter 1, a graph exists as an abstract object, devoid of any
geometric connotation of its ability of being drawn in three-dimensional Euclidean
space. For example, an abstract graph G1 can be defined as

G1 = ( V, E, ψ ),

rs
Where the set V consists of the five objects named a, b, c, d, and e, that is,
ve
V={a,b,c,d,e}

And the set E consists of seven objects (none of which is in set) named 1,2,3,4,5,6,
and 7, that is,
ni

E={1,2,3,4,5,6,7},

And the relationship between the two sets is defined by the mapping ψ , which
consists of
U

é
ê1 ¾¾®(a, c)
ê 2 ¾¾
®( c , d )
ê
ê
ity

ê3 ¾¾®(a, d )
ê
ψ = êê 4 ¾¾
®( a , b )
ê5 ¾¾®(b, d )
ê
ê
ê 6 ¾¾
®( d , e )
m

ê
ê 7 ¾¾
®(b, e)
ë
)A

Here, he symbol 1 → (a,c) says that object 1 from set E is mapped onto the
(unordered) pair (a,c) of objects from set V.

Now it so happens that this combinatorial abstract object G1 can also be

represented by means of a geometric figure. In fact, the sketch in Fig 2-13 is one such
geometric representation of this graph. Moreover, it is also true that any graph can be
(c

represented by means of such a configuration in three-dimensional Euclidean space.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 107

It is important to realize that what is sketched in Fig 2-13 is merely one (out of
Notes

e
infinitely many) representation of the graph G1 and not the graph G1 itself. We could
have, for instance, twisted some of the edges or could have placed e within the triangle

in
a,d,b and thereby obtained a different figure representing G1. However, when there is
no chance of confusion, a pictorial representation of the graph has been and will be
regarded as the graph itself.

nl
This convenient slurring over is done deliberately for the sake of simplicity and
clarity. Learning graph theory for the first time without any diagrams would be extremely
difficult and little fun. Graph and a geometric representation of a graph.

O
Definition: 6.1
A graph G is said to be planar if there exists some geometric representation of
G which can be drawn on a plane such that no two of its edges intersect. A graph that

ity
cannot be drawn on a plane without a crossover between its edges is called nonplanar.

A drawing of a geometric representation of a graph on any surface such that no

edges intersect is called embedding. Thus, to declare that a graph G is nonplanar.
We have to show that of all possible geometric representation of G none can be

rs
embedded in a plane. Equivalently, a geometric graph G is planar if there exists a graph
isomorphic to G that is embedded in a plane. Otherwise, G is nonplanar. An embedding
of a planar graph G on a plane is called a plane representation of G.
ve
For instance, consider the graph represented by. The geometric representation
clearly is not embedded in a plane, because the edges e and f are interesting. But if we
redraw edge foutside the quadrilateral, leaving the other edges unchanged, we have
embedded the new geometric graph in the plane, thus showing that the graph which is
ni

being represented is planar. As another example, the two isomorphic diagrams in are
different geometric representations of one and the same graph. One of the diagram is a
plane representation; the other one is not. The graph, of course, is planar. On the other
U

hand, you will not be able to draw any of the three configurations in on a plane without
edges intersecting. The reason is that the graph which these three different diagrams in
represent is nonplanar.

Γ ) or by one of
ity

A graph G [which may be given by an abstract notation G = ( V, E,

its geometric representation] is planar or nonplanar

Kuratowski’s Two Graphs

m

Theorem: 6.2
Prove that the complete graph of five vertices is nonplanar.
)A

Proof
Let the five vertices in the complete graph be named v1,v2,v3,v4 and v5.

A complete graph, as you may recall, is a simple graph in which every vertex is
(c

joined to every other vertex by means of an edge.

This being the case, the “meeting” of edges at a vertex is not considered an intersection.

Amity Directorate of Distance & Online Education

 108 Graph theory and Combinatorics

Notes

e
in
nl
O
ity
rs
ve
ni
U

The five vertices of complete graph

A circuit going from v1 to v2 to v3 to v4 to v5 to v1—that is, a pentagon. See Fig

ity

5-1(a). This pentagon must divide the plane of the paper into two regions, one inside
and the other outside (Jordan Curve Theorem).

Since vertex v1 is to be connected to v3 by means of an edge, this edge may be

drawn inside or outside the pentagon (without intersecting the five edges drawn
m

previously).

Suppose that we choose to draw a line from v1 to v3 inside the pentagon. (If we
choose outside, we end up with the same argument.)
)A

Now we have to draw and edge from v2 to v4 and another one from v2 to v5. Since
neither of these edges can be drawn inside the pentagon without crossing over the
edge already drawn, we draw both these edges outside the pentagon. The edge
connecting v3and v5 cannot be drawn outside the pentagon without crossing the edge
(c

between v2 and v4.

Therefore, v3 and v5 have to be connected with an edge inside the pentagon.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 109

Now we have yet to draw and edge between v1 and v4.

Notes

e
This edge cannot be placed inside or outside the pentagon without a crossover.
Thus the graph cannot be embedded in a plane.

in
A complete graph with five vertices is the first of the two graphs of Kuratowski.

The second graph of Kuratowski is a regular connected graph with six vertices and

nl
nine edges, shown in its two common geometric representation in where it is fairly easy
to see that the graphs are isomorphic.

Employing visual geometric arguments similar to those used in proving, it can be

O
shown that the second graph of Kuratowskiis also nonplanar.

ity
rs
ve
Remark:
ni

●● Kuratowski’s second graph is also nonplanar.

●● Both are regular graphs.
U

●● Both are nonplanar.

●● Removal of one edge or a vertex makes each a planar graph.
●● Kuratowski’s first graph is the nonplanar graph with the smallest number
of vertices, and Kuratowski’s second graph is the nonplanar graph with the
ity

smallest number of edges.

●● Thus the both are the simplest nonplanar graphs.

Different Representations of a Planar Graph

m

In following the proof of the Theorem 5-1, it may have appeared that one’s ability
to draw a planar graph in a plane depended on his ability to draw many crooked lines
through devious routes. This is not the case. The following important and somewhat
)A

surprising result, due to Fary, tells us that there is no need to bend edges in drawing a
planar graph to avoid edge intersections. Any simple planar graph can be embedded in
plane such that every edge is drawn as a straight line segment.

Definition: 6.3
(c

A plane representation of a graph divides the plane into regions also called
windows, faces, or meshes) as shown in Figure A region is characterised by set of

Amity Directorate of Distance & Online Education

 110 Graph theory and Combinatorics

Notes

e
in
nl
Straight line representation of the graph

O
Theorem: 6.4
A planar graph may be embedded in a plane such that any specified region (i.e.,

ity
specified by the edges forming it) can be made the infinite region.

Proof:
In terms of the regions on the sphere, we see that there is no real difference

rs
between the infinite region and the finite regions on the plane.

Therefore, when we talk of the regions in the plane representation of a graph, we

include the infinite region.
ve
Also, since there is no essential difference between an embedding of a planar
graph on a plane or on a sphere (a plane may be regarded as the surface of a sphere
or infinitely large radius), the term “plane representation” of a graph is often used to
include spherical as well as planar embedding.
ni

Theorem: 6.5
Prove that a connected planar graph with n vertices and e edges has e-n+2
U

regions.

Proof
ity

It will suffice to prove the theorem for a simple graph, because adding a self-loop or
a parallel edge simply adds one region to the graph and simultaneously increases the
value of e by one.

We can also remove all edges that do nor form boundaries of any region. Three
m

such edges are shown in Fig. 5-4. Addition (or removal) of any such edge increases
(or decreases) e by one and increases (or decreases) n by one, keeping the quality e-n
unaltered.
)A

Since any simple planar graph can have a plane representation such that each
edge is a straight line, any planar graph can be drawn such that each region is a
polygon (a polygon net). Let the polygonal net representing the given graph consist of f
regions or faces, and let kpbe the number if p-sided regions. Since each edge is on the
boundary of exactly two regions,
(c

3.k3 + 4.k4 + 5.k5 +…+r.kr =2.e.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 111

Where k is the number of polygons, with maximum edges.

Notes

e
Also,

in
k3 + k4 + k5 + …+ kr=f.

The sum of all angles subtended at each vertex in the polygonal net is

2 π n.

nl
π
Recalling that the sum of all interior angles of a p-sided polygon is (p-2), and the
π
sum of the exterior angles is (p+2), let us compute the expression as the grand sum

O
of all interior andlesf-1 finite regions plus the sum of the exterior angles of the polygon
defining the infinite region. This sum is

π (3-2). k + π (4-2).k + …+ π (r-2).k + 4 π

3 4 r

= π (2e-2f) + 4 π .

ity

Equating we get

2 π (e-f) + 4 π = 2 π n.

rs
or e-f +2 = n.

Therefore, the number of regions is

f =e-n+2.
ve
Theorem: 6.6
In any simple, connected planar graph with f regions, n vertices, and e edges
ni

(e>2), the following inequalities must hold:

3
1. e≥ f
2
U

2. e ≤ 3n − 6

Proof:
ity

Since each region is bounded by at least three edges and each edge belongs to
exactly two regions,

2e ≥ 3f,

or e ≥ 3∕2f.
m

Substituting for f from Euler’s formula is inequality

e ≥ 3∕2 (e-n+2)
)A

or e ≤ 3n-6.

It is often useful in finding out if graph is nonplanar. For, example, in the case of K3,
the complete graph of five vertices.
(c

Amity Directorate of Distance & Online Education

 112 Graph theory and Combinatorics

Notes

e
in
nl
O
n = 5, e = 10, 3n - 6 = 9 <e.

Thus the graph violates and hence it is not planar.

ity
Incidentally, this is an alternative and independent proof of the non-planarity of
Kuratowski’s first graph.

Every simple planar graph must satisfy the above equation the mere satisfaction of
this inequality does not guarantee the planarity of a graph.

rs
For example, Kuratowski’s second graph, K3.3, satisfies because

e = 9,
ve
3n- 6 = 3.6-6 = 12.

Yet the graph is non-planar.

To prove the non-planarity of Kuratowski’s second graph, we make use of the

ni

additional fact that no region in the graph can be bounded with fewer than four edges.

Hence, if this graph were planar, we would have

2e ≥ 4f,
U

And substituting for f from Euler’s formula,

2e ≥ 4(e-n+2),
ity

or 2.9 ≥ 4(9-6+2),

or 18 ≥ 20, a contradiction.

Hence the graph cannot be planar.

m

Plane Representation and Connectivity:

In a disconnected graph the embedding of each component can be considered
)A

independently. Therefore, it is clear that a disconnected graph is planar if and only if its
components is planar. Similarly, in a separable (or 1-connected) graph the embedding
of each block (i.e., maximal non-separable subgraph) can be considered independently.
Hence a separable graph is planar if and only if each of its blocks is planar.

Therefore, in questions of embedding or planarity, one need consider only

(c

nonseparable graphs.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 113

We must define the meaning of unique embedding can be made to coincide by

Notes

e
suitably rotating one sphere with respect to the other and possibly distorting regions
(without letting a vertex cross an edge). If of all possible embeddings on a sphere no

in
two are distinct, the graph is said to have a unique embedding on a sphere (or a unique
plane representations).

For example, consider two embeddings of the same graph.

nl
The embedding (b) has a region bounded with five edges, but embedding (a)
has no region with five edges. Thus, rotating the two spheres on which (a) and (b) are
embedded will not make them coincide. Hence the two embeddings are distinct, and

O
the graph has no unique plane representation.

On the other hand, the embeddings when considered on a sphere, can be made to
coincide. (Remember that edges can be bent, and in a spherical embedding there is no

ity
infinite region), when a graph is uniquely embeddable in a sphere.

Remark:
The spherical embedding of every planar 3-connected graph is unique.

rs
ve
ni
U

Geometric Dual
Consider the plane representation of a graph in figure with six regions or faces F1,
F2, F3, F4, F5, and F6.
ity

Let us place six points p1,p2, . . . ,p6. One in each of the regions, as shown in figure.
Next let us join these six points according to the following procedure:
m
)A
(c

Amity Directorate of Distance & Online Education

 114 Graph theory and Combinatorics

Notes

e
in
nl
Construction of a dual graph

O
If two regions Fi and Fj are adjacent (i.e., have a common edge), draw a line joining
points pi and pj that intersects the common edge between Fi and Fj exactly once.

If there is more than one edge common between Fi and Fj, draw one line between
points pi and pj for each of the common edges.

ity
For an edge e lying entirely in one region, say Fk, draw a self-loop at point pk
intersecting e exactly once.

By this procedure we obtain a new graph G* consisting of six vertices, p1, p2, . .

rs
, p6 and of edges joining these vertices. Such a graph G* is called a dual (or strictly
speaking, a geometric dual) of G.

Clearly, there is a one -to-one correspondence between the edges of graph G and
ve
its dual G* - one edge of G* intersecting one edge of G.

Observations that can be made about the relationship between a planar

graph G and its dual G* are
ni

1. An edge forming a self-loop in G yields a pendant edge in G*.

2. A pendant edge in G yields a self-loop in G*.
3. Edges that are in series in G produce parallel edges in G*.
U

4. Parallel edges in G produce edges in series in G*.

5. The result of the general observation that the number of edges constituting the
boundary of a region Fi in G is equal to the degree of the corresponding vertex
ity

pi in G*, and vice versa.

6. Graph G* is also embedded in the plane and is therefore planar.
7. Considering the process of drawing a dual G* from G, it is evident that G is a
m

dual of G* [see figure]. Therefore, instead of calling G* a dual of G, we usually

say that G and G* are dual graphs.
8. If n, e, f, r, and µ denote as usual the numbers of vertices, edges, regions,
)A

rank, and nullity of a connected planar graph G, and if n*, e*, f*, r*, and µ * are
the corresponding numbers in dual graph G*, then
n* = f,

e* = e,
(c

f* = n.

Using the above relationship, one can immediately get

Amity Directorate of Distance & Online Education
 Graph theory and Combinatorics 115

r* = µ,
Notes

e
µ * = r.

in
nl
O
ity
Unique the above Dual Graphs:
rs
In an incident on a pendant vertex is called a pendant edge.
ve
Other words, are all duals of a given graph isomorphic. From the methods of
constructing a dual, it is reasonable to except that a planar graph G will have a unique
dual if and only of it has a unique plane representation or unique embedding on a
ni

sphere.

For instance, in figure the same graph (isomorphic) had two distinct embeddings,
(a) and (b). Consequently, the duals of these isomorphic graphs are non-isomorphic,
U

as shown in figure

The graphs in figure, however, are 2-isomorphic. This stated without proof, is a
generalization of this example.
ity

Theorem: 6.7
All duals of a planar graph G are 2-isomorphic; and every graph 2-isomorphic; to a
dual of G is also a dual of G.
m

Proof:
It is quite appropriate to refer to a dual as the dual of a planar graph.
)A

Since a 3-connected planar graph has a unique embedding on a sphere, its dual
must also be unique. In other words, all duals of a 3-connected graph are isomorphic.

Combinatorial Dual:
(c

We have defined and discussed duality of planar graphs in a purely geometric

sense. The following provides us with an equivalent definition of duality independent of
geometric notions.
Amity Directorate of Distance & Online Education
 116 Graph theory and Combinatorics

Theorem: 6.8:
Notes

e
A necessary and sufficient condition for two planar graphs G1 and G2 to be duals
of each other is as follows: There is a one-to-one correspondence between the edges

in
in G1 and the edges in G2 such that a set of edges in G1 forms a circuit if and only if the
corresponding set in G2 forms a cut-set.

nl
Proof:
Let us consider a plane representation of a planar graph G. Let us also draw
(geometrically) a dual G* of G. Then consider an arbitrary circuit Γ in G. Clearly, Γ

O
will form some closed simple curve in the plane representation of G dividing the plane
into two areas. (Jordan Curve Theorem).

Thus the vertices of G* are partitioned into two nonempty, mutually exclusive
subsets one inside Γ and the other outside.

ity
In other words, the set f edges Γ * in G* corresponding to the set Γ in G is a cut-
set in G*. (No proper subset of Γ * will be a cut-set in G*).

Likewise it is apparent that corresponding to a cut-set S in G such that S is a

rs
circuit. This proves the necessity portion

To prove the sufficiency, let G be a planar graph and let G’ be a graph for which
there is a one-to-one correspondence between the cut-sets of G and circuits of G’, and
ve
vice versa.

Let G* be a dual graph of G. There is a one-to-one correspondence between the

circuits of G’ and cut-sets of G, and also between the cut-sets of G and circuits of G*.
ni

Therefore there is a one-to-one correspondence between the circuits of G’ and G*,

implying that G’ and G* are 2-isomorphic G’ must be a dual of G.
U

Dual of a Subgraph:
Let G be a planar graph and G* be its dual. Let a be an edge in G, and the
corresponding edge in G* be a*. Suppose that we delete edge a from G and then try to
find the dual of G-a. If edge a was on the boundary of two regions, removal of a would
ity

merge these two regions into one. Thus the dual (G - a)* can be obtained from G* by
deleting the corresponding edge a* and them fusing the two end vertices of a* in G*-a*.
On the other hand, if edge a is not on the boundary, a* forms a self-loop. In that case
G*-a* is the same as (G-a)*. Thus if a graph G has a dual G*, the dual of any subgraph
m

of G can be obtained by successive application of this procedure.

Dual of a Homeomorphic Graph:

)A

Let G be a planar graph and G* be its dual. Let a be an edge in G, and the
corresponding edge in G* be a*. Suppose that we create an additional vertex in G by
introducing a vertex of degree two in edge a (i.e., a now becomes two edges in series).
It will simply add an edge parallel to a* in G*. Likewise, the reverse process of merging
two edges i series will simply eliminate one of the corresponding parallel edges in G*.
(c

Thus if a graph G has a dual of any graph homeomorphic to G can be obtained from G*
by the above procedure.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 117

So far we have been studying duality for planar graphs only. This was forced upon
Notes

e
us because the very definition of duality depended on the graph being embedded in a
plane. However, now that provides us with an equivalent abstract definition of duality

in
(namely, the correspondence between circuits and cut-sets), which does not depend on
a plane representation of a graph, we will see if the concept of duality can extended to
nonplanar graphs also. In other words, given a non-planar graph G, can we find another
graph G’ with one-to-one correspondence between their edges such that every circuit in

nl
G corresponds to a unique cut-set in G’, and vice versa. The answer to this question is
no, as shown in the following important theorem, due to Whitney.

O
Theorem: 6.9
Prove that a graph has a dual if and only is it is planar.

ity
Proof:
We need prove just the “only if” part. Modu That is, we have only to prove that a
nonplanar graph does not have a dual. Let G be a nonplanar graph. Then according to
Kuratowski’s theorem, G contains K5 or K3,3 or a graph homeomorphic to either of these.

rs
We have already seen that a graph G can have a dual only if every subgraph g of G
and every graph homeomorphic to g has a dual. This if we can shoe that neither K5 nor
K3,3 has a dual, we have proved the theorem.
ve
This we shall prove by contradiction as follows:

(a) Suppose that K3,3 has a dual D. Observe that the cut-sets in K3,3, correspond
to circuits in D and vice versa. Since K3,3 has no cut-set consisting of two edges, D has
no circuit consisting of two edges. That is, D contains no part of parallel edges. Since
ni

every circuit in K3,3 is of length four or six, D has no cut-set with less than four edges.
Therefore, the degree of every vertex in D is at least four. As D has no parallel edges
and the degree of every vertex is at least four, D must have at least five vertices each
U

of degree four or more. That is D must have at least (5 x 4)/2 =10 edges. This is a
contradiction, because K3,3 has nine edges and so must its dual. Thus K3,3 cannot have
a dual. Likewise,

(b) Suppose that the graph K5 has a dual H. Note that K5 has (1) 10 edges, (2)
ity

no pair of parallel edges, (3) no cut-set with two edges, and (4) cut-sets with only
four or six edges. Consequently, graph H must have (1) 10 edges, (2) no vertex with
degree less than three, (3) no pair of parallel edges, and (4) circuits of length four and
six only. Now graph H contains a hexagon ( a circuit of length six), and length three
m

or a pair of parallel edges [see figure ]. Since both of these are forbidden in H and
H has 10 edges, there must be at least seven vertices in H. The degree of each of
these vertices is at least three. This leads to H having at least 11 edges. Which is a
)A

contradiction.

There is yet another equivalent combinatorial definition of duality, and also given

Two planar graphs G and G* are said to e duals (or combinatorial duals) of each
other if there is a one-to-one correspondence between the edges of G and G* such that
(c

if g is any subgraph of G and h is the corresponding subgraph of G*, then

Rank of (G*-h) = rank of G* - nullity of g.

Amity Directorate of Distance & Online Education

 118 Graph theory and Combinatorics

This relationship is shown diagrammatically in figure

Notes

e
in
nl
O
Rank of (G*-h)=Rank of G*-Nullity of g

Combinatorial duals

ity
As an example, consider the graph in figure and its dual in figure 5-11. Take the
subgraph {e4, e5, e6, e7} in figure and the corresponding subgraph {e*4, e*5, e*6, e*7} in
figure

Rank of (G* - {e*4, e*5, e*6, e*7}) = rank of {e*1, e*2, e*3} = 2,

Rank of G* = 3,

nullity of {e4, e5, e6, e7} = 1, rs

ve
and

2 = 3-1.

Clearly, this definition is also independent of the geometric connotation. It is

ni

therefore often preferred for proving results in purely algebraic fashion. However, in
deciding whether or not two given graphs are dual the combinatorial definitions are
difficult to use.
U

The proof of equivalence of combinatorial and geometric duals is quite involved.

Since the geometric and combinatorial duals are one and the same, we simply refer to
them as the dual, rather than the geometric or combinatorial dual.
ity

Self- Dual Graphs:

If a planar graph G is isomorphic to its own dual, it is called a self-dual graph. It
can be easily shown that the four-vertex complete graph is a self-dual graph. Self-dual
graphs have interesting properties and pose some unsolved problems.
m

More on Criteria of Planarity

)A

Set of Basic Circuits:

A set C of circuits in a graph is said to be a complete set of basic circuits if (i) every
circuit in the graph can be expressed as a ring sum of some or all circuits in C, and (ii)
no circuit in C can be expressed as a ring sum of others in C.
(c

The significance of complete sets of basic circuits will be clearer in Chapter

6, in relation to the vector space of a graph. It may, however, be mentioned here

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 119

that whereas a set of fundamental circuits (as defined in Chapter 3 with respect to a
Notes

e
spanning tree always constitutes a complete set of basic circuits, the converse does not
hold for all graphs

in
In a planar graph a complete set of basic circuits has an additional property

In a plane representation of a planar, connected graph G the set of circuits forming

the interior regions constitutes a complete set of basic circuits. For any circuit Γ in G

nl
can be expressed as the ring sum of the circuits defining the regions contained in Γ .
Observe that every edge appears in at most two of these basic circuits. Thus for every
planar graph G we can find a complete set of basic circuits such that no edge appears

O
in more than two of these basic circuits. This result and its converse (Proof of which can
be found in [5-6]) lead to another well-known characterization of planar graphs.

Theorem: 6.10:

ity
A graph G is planar if and only if there exists a complete set of basic circuits (i.e.,
all µ of them, µ being the nullity of G) such that no edge appears in more than two of
these circuits.

Proof:

rs
A three of these classic characterizations suffer from two shortcomings. First,
they are extremely difficult to implement for a large graph. Second in case the graph is
ve
planar they do not give a plane representation of the graph.

These drawbacks have prompted recent discoveries of several map-construction

methods, where the testing of planarity itself is based on several other construction
ni

methods, some of them quite similar, have been implemented on digital computers. In
most of these methods, the given graph is first reduced to one or more simple, non-
separable graphs with every vertex of degree three or more and with e ≤ 3n – 6. Then
the construction algorithm is applied such that either one succeeds in will be said on
U

such algorithms.

Some algorithms are better than others, but all are laborious and time- consuming.
The search for a simple, elegant, and practical characterization of a planar graph is far
ity

from over.

Thickness and Crossing:

When a graph has a pair of edges that cross, it’s known as a crossing on the
m

graph. Counting up all such crossings gives you the total number for that drawing of the
graph. Therefore, one of the main problems is to minimize the number of crossings by
adjusting the positions of the vertices.
)A

This is usually done to improve the readability of graphs or to solve some real-
world application such circuit board design or road or railroad crossings. The minimum
number of crossings in any drawing of a graph is known as the crossing number.

Computing the crossing number is an NP-hard problem, which means there is no

(c

algorithm that can find the minimum number of crossings of any graph in deterministic
polynomial time. For rectilinear complete graphs, we know the crossing number for
graphs up to 27 vertices, the rectilinear crossing number.

Amity Directorate of Distance & Online Education

 120 Graph theory and Combinatorics

Since this problem is NP-hard, it would be at least as hard to have software

Notes

e
minimize or draw the graph with the minimum crossing, except for graphs where we
already know the crossing number. In all other cases, it is best to have a triangular

in
convex hull. You can discover why by manually drawing complete graphs with a low
number of vertices.

Local Crossing Number:

nl
Though there is a lot of study around the crossing number, one area that has
received less attention is the local crossing number. The local crossing number of a
drawing of a graph is the largest number of crossings on a single edge. The minimum

O
local crossing in any drawing of a graph is the local crossing number for that graph.

Questions:

ity
1. If a plane graph has k components, than n-e+f is equal to_________
a) k-1
b) k

rs
c) k+1
d) none
2. If K5 were planar graph, then e is less then or equal to_________
ve
a) n-6
b) 2n-6
c) 3n-6
ni

d) all.
3. The graph kn is planar if and only if________
U

a) n>5
b) n = 3
c) n<4
ity

d) n = 4
4. A 3-regular graph with four or more vertices is__________
a) an euler graph
m

b) not an euler graph

c) a wheel
d) a cycle
)A

5. Km;n is planar if and only if

a) m 2 or n 2
b) m 2 and n 2
(c

c) m = 2 or n = 2
d) m = 2 and n = 2

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 121

6. A tree with 3 or more vertices __________

Notes

e
a) has an euler line and a hamiltonian circuit
b) has an euler line but not a hamiltonian circuit

in
c) has neither an euler nor a hamiltonian circuit
d) has a hamiltonian circuit but not an euler line.

nl
7. The graph Q3 has_____________
a) has an euler line and a hamiltonian circuit

O
b) has an euler line but not a hamiltonian circuit
c) has neither an euler nor a hamiltonian circuit
d) has a hamiltonian circuit but not an euler line

ity
8. The graph Q4 has_____________
a) has an euler line and a hamiltonian circuit
b) has an euler line but not a hamiltonian circuit

rs
c) has neither an euler nor a hamiltonian circuit
d) has a hamiltonian circuit but not an euler line
9. The graph K5 has __________
ve
a) has an euler line and a hamiltonian circuit
b) has an euler line but not a hamiltonian circuit
c) has neither an euler nor a hamiltonian circuit
ni

d) has a hamiltonian circuit but not an euler line

10. The graph K5 has____________
U

a) is planner
b) is nonplanar after the removal of any edge
c) is planar after the removal of any edge
ity

d) none of these

Answer:
1. c
m

2. c
3. c
)A

4. b
5. a
6. c
7. d
(c

8. a

Amity Directorate of Distance & Online Education

 122 Graph theory and Combinatorics

9. a
Notes

e
10. c
11. d

in
12. c
13. b

nl
Questions:
1. Define planar graphs with an example

O
2. Explain Hamiltonian cycle and hamiltonain circuit
3. Relationship between Hamiltonian cycle and Hamiltonian circuit

ity
rs
ve
ni
U
ity
m
)A
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 123

Module V
Notes

e
Key Learning Objectives:

in
At the end of this module, you will be able to:

●● Implement the concept of partitions and counting functions

nl
●● Divide the number of partitions into odd or unequal parts
●● Discuss the concepts of necklaces, Euler’s function, set of symmetries, and
enumeration in the odd and even cases

O
ity
rs
ve
ni
U
ity
m
)A
(c

Amity Directorate of Distance & Online Education

 124 Graph theory and Combinatorics

Unit- VII
Notes

e
Partitions:

in
In number theory and combinatorics, a partition of a positive integer, also called
an integer partition, is a way of writing n as a sum of positive integers. Two sums that

nl
differ only in the order of their summands are considered the same partition. (If order
matters, the sum becomes a composition.) For example, 4 can be partitioned in five
distinct ways:

O
4
3+1
2+2
2+1+1

ity
1+1+1+1

The order-dependent composition 1 + 3 is the same partition as 3 + 1, while the

two distinct compositions 1 + 2 + 1 and 1 + 1 + 2 represent the same partition 2 + 1 + 1.

A summand in a partition is also called a part. The number of partitions of n is

partition of n.
rs
given by the partition function p(n). So p(4) = 5. The notation λ⊢n means that λ is a

Partitions can be graphically visualized with Young diagrams or Ferrers diagrams.

ve
They occur in a number of branches of mathematics and physics, including the study of
symmetric polynomials and of the symmetric group and in group representation theory
in general
ni

Example:

The seven partitions of 5 are:

U

●● 5
●● 4+1
●● 3+2
●● 3+1+1
ity

●● 2+2+1
●● 2+1+1+1
●● 1+1+1+1+1
m

In some sources partitions are treated as the sequence of summands, rather than
as an expression with plus signs. For example, the partition 2 + 2 + 1 might instead
be written as the tuple (2, 2, 1) or in the even more compact form (22, 1) where the
)A

superscript indicates the number of repetitions of a term.

Odd parts and distinct parts:

Among the 22 partitions of the number 8, there are 6 that contain only odd parts:
(c

●● 7+1
●● 5+3

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 125

●● 5+1+1+1
Notes

e
●● 3+3+1+1
●● 3+1+1+1+1+1

in
●● 1+1+1+1+1+1+1+1
Alternatively, we could count partitions in which no number occurs more than once.
Such a partition is called a partition with distinct parts. If we count the partitions of 8

nl
with distinct parts, we also obtain 6:

●● 8

O
●● 7+1
●● 6+2
●● 5+3

ity
●● 5+2+1
●● 4+3+1
This is a general property. For each positive number, the number of partitions with
odd parts equals the number of partitions with distinct parts, denoted by q(n).

rs
For every type of restricted partition there is a corresponding function for the
number of partitions satisfying the given restriction. An important example is q(n).
ve
Partition of a set:
A partition of a set X is a set of non-empty subsets of X such that every element x
in X is in exactly one of these subsets (i.e., X is a disjoint union of the subsets).
ni

Equivalently, a family of setsP is a partition of X if and only if all of the following

conditions hold

●● The family P does not contain the empty set (that is ∅ ∉ P).
U

●● The union of the sets in P is equal to X. The sets in P are said to coverX.
●● The intersection of any two distinct sets in P is empty. The elements of P are
said to be pairwise disjoint.
ity

The sets in P are called the blocks, parts or cells of the partition

The rank of P is |X| − |P|, if X is finite.

Examples:
m

●● The empty set has exactly one partition, namely for any non-empty set X, P =
{X} is a partition of X, called the trivial partition.
●● Particularly, every singleton set {x} has exactly one partition, namely { {x} }.
)A

●● For any non-empty proper subsetA of a set U, the set A together with its
complement form a partition of U.
●● The set { 1, 2, 3 } has these five partitions (one partition per item):
o { {1}, {2}, {3} }, sometimes written 1|2|3.
(c

o { {1, 2}, {3} }, or 12|3.

o { {1, 3}, {2} }, or 13|2.

Amity Directorate of Distance & Online Education

 126 Graph theory and Combinatorics

o { {1}, {2, 3} }, or 1|23.

Notes

e
o { {1, 2, 3} }, or 123 (in contexts where there will be no confusion with the
number).

in
●● The following are not partitions of { 1, 2, 3 }:
o { {}, {1, 3}, {2} } is not a partition (of any set) because one of its elements
is the empty set.

nl
o { {1, 2}, {2, 3} } is not a partition (of any set) because the element 2 is
contained in more than one block.

O
o { {1}, {2} } is not a partition of {1, 2, 3} because none of its blocks contains
3; however, it is a partition of {1, 2}.

Partitions of n:

ity
A partition of n is a combination (unordered, with repetitions allowed) of positive
integers, called the parts, that add up to n

In other words, a partition is a multiset of positive integers, and it is a partition of

rs
n if the sum of the integers in the multiset is n. It is conventional to write the parts of a
partition in descending order,

Ferrers diagram and conjugate partition:

ve
The Ferrers diagram, also called Young diagram, of a partition λ +n is a
rectangular array of n boxes, or cells, with one row of length j for each part j of λ . For
example, the diagram of (7,5,2,2) is
ni
U
ity

The conjugate of a partition λ +n is the partition of n whose diagram you get by

reacting the diagram of λ about the diagonal so that rows become columns and
columns become rows. We use the notation for the conjugate of λ .

In our example above, with λ =(7,5,2,2), the diagram of

m
)A
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 127

Partition function and Recurrence relation:

Notes

e
P(n) sometimes also denoted gives the number of ways of writing the integer
as a sum of positive integers, where the order of addends is not considered significant.

in
By convention, partitions are usually ordered from largest to smallest For example,
since 4 can be written

nl
4
= 4 (1)

= 3 + 1 (2)

O

= 2 + 2 (3)

= 2 + 1 + 1 (4)

= 1 + 1 + 1 + 1, (5)

ity
it follows that P (4) = 5. P (n) is sometimes called the number of unrestricted
partitions, and is implemented Partitions P[n].

The values of P (n) for n = 1, 2, ..., are 1, 2, 3, 5, 7, 11, 15, 22, 30,

rs
The first few prime values of P (n) are 2, 3, 5, 7, 11, 101, 17977, 10619863, …..,
corresponding to indices 2, 3, 4, 5, 6, 13, 36, 77, 132, ...The largest known giving a
probable prime is 1000007396 with 35219 decimal digits, while the largest known
giving a proven prime is 221444161 with 16569 decimal digits
ve
ni
U

When explicitly listing the partitions of a number n, the simplest form is the so-
ity

called natural representation which simply gives the sequence of numbers in the
representation (e.g., (2, 1, 1) for the number 4 = 2+1+1). The multiplicity representation
instead gives the number of times each number occurs together with that number
(e.g., (2, 1), (1, 2) for 4=2.1+1.2). The Ferrers diagram is a pictorial representation of a
m

partition. For example, the diagram above illustrates the Ferrers diagram of the partition
6+3+3+2+1=15 .

Euler gave a generating function for P (n) using the q-series

)A

(q)∞ ≡ Õ (1-q m ) (6)

m =1

¥
n (3 n +1)/2
= S (-1) q
n
(7)
-¥
(c

2 5 7 12 15 22 26
= 1 − q − q + q + q − q − q + q + q + .... (8)

Amity Directorate of Distance & Online Education

 128 Graph theory and Combinatorics

Here, the exponents are generalized pentagonal numbers 0, 1, 2, 5, 7, 12, 15, 22,
Notes

e
26, 35, ... and the sign of the th term (counting 0 as the 0th term) is(-1)[(k+1)/2] (with
the floor function). Then the partition numbers p(n) are given by the generating function

in
¥
1
(q)∞
= å P ( n) q
n=0
n
(9)

= 1 + q + 2q 2 + 3q 3 + 5q 4 + .... (10)

nl
The number of partitions of a number n into m parts is equal to the number of
partitions into parts of which the largest is m, and the number of partitions into at most

O
m parts is equal to the number of partitions into parts which do not exceed m. Both
these results follow immediately from noting that a Ferrers diagram can be read either
row-wise or column-wise

For example, if for all , then the Euler transform is the number of

ity
partitions of into integer parts.

Euler invented a generating function which gives rise to a recurrence equation in

,

rs
n
1 1
P(n)=å (-1) k +1[ P(n- k (3k -1))+ P(n- k (3+1))] (11)
k =1 2 2

Other recurrence equations include

ve
¥
P(2n+1)= P(n)+å [ P(n-4k 2 -3k )+ P(n-4k 2 +3k )]-
k =1 (12)
¥

å (-1) [ P(2n+1-3k
k 2
+k )+ P(2n+1-3k 2 -K )]
ni

k =1

and
1 n-1
P(n)= å σ1 (n-k ) P(k )
U

(13)
n k =0

where σ1 (n) is the divisor function as well as the identity

[( 24 n +1-1)/6]
1 (14)
å (-1) k P(n- k (3k +1))=0
ity

k =[-( 24 n +1 +1)/6] 2

where is the floor function and is the ceiling function.

A recurrence relation involving the partition function Q is given by

m

[ n 2]
P(n)=å Q(n-2k ) P(k ) (15)
k =0
¥ ¥ (1-q 5 n )
å P(5n+1)q =Õ (1-q n
(mod 5) (16)
)A

5 n-4
n=0
n =1 )(1-q 5 n-1 )

¥ ¥ (1- q 5 n )
å P(5n +1)q n
=2Õ 5 n-3
(mod 5) (17)
n=0
n=1 (1- q )(1- q 5 n-2 )

¥ ¥ (1-q 5 n-4 )(1-q 5 n-1 )(1-q 5 n )

å P(5n+3)q n =3 Õ (mod 5) (18)
(c

n=0
n =1 (1-q 5 n-3 ) 2 (1-q 5 n-2 ) 2

P (n)-P (n-1)-P (n-2)+ P (n-5)+ P (n-7)-P (n-12)-P (n-15)+...=0 (20)

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 129
k
where the sum is over generalized pentagonal numbers £n and the sign of the th
Notes

e
term is (-1)[( k +1) 2] , as above. Ramanujan stated without proof the remarkable identities
¥
(q 5 )5¥

in
å P(5k +4)q =5 (q)
k =0
k
6
(21)
¥

and

nl
¥
(q 7 )3¥ ( q 7 )¥
7
å P(7k +5)q k =7
k =0
4
( q )¥
+ 49 q
(q )8¥
(22)

To obtain the asymptotic solution

O
1
P ( n)  eπ 2n 3
(23)
4n 3

é æ öù
ê sinh çç π n- 1 ÷÷ ú

ity
ê ÷
1 ¥ d ê ç
çè k 24 ÷ø úú
P (n)= å k
π 2 k =1
A ( n ) k ê
dn ê 1
ú (24)
n - ú
ê ú
êë 24 úû
where

h =1
é k -1 j æ hj é hj ù 1 ö 2π ihn ù
Ak (n)=å δ GCD(h, k ),1exp êêπ i å çç -ê ú- ÷÷-
ç ÷
ë j =1 k è k êë k úû 2 ø k úû
ú

rs (25)
ve
δ mn is the Kronecker delta, and is the floor function. The remainder after terms
is

N æ K n ö÷
R( N )<CN -1 2 + D sinh ççç ÷÷ (26)
n èç N ø÷
ni

where C and D are fixed constants

To found an algebraic formula for the partition function P(n) as a finite sum of
U

algebraic numbers as follows. Define the weight-2 meromorphic modular form F(z) by

E2 ( z )-2 E2 (2 z )-3E2 (3 z )+6 E2 (6 z ) (27)

F ( z)
2 η2 ( z )η2 (2 z )η2 (3z )η3 (6 z )
ity

2π iz
were q=e , E2 (q ) is an Eisenstein series, and η(q) is a Dedekind eta function.
Now define
æ 1 d 1 ÷ö
R( z )=-çç + ÷ F ( z ) (28)
çè 2π i dz 2π y ÷÷ø
m

where z= x+iy . Additionally let Qn be any set of representatives of the

equivalence classes of the integral binary quadratic form Q( x, y )=ax 2 +bxy+cy 2 such
that 6|a with a>0 and bº1(mod12) , and for each Q( x, y ) , let α Q be the so-called CM point
)A

in the upper half-plane, for which Q(α Q ,1)=0 . Then

Tr (n)
P(n)= (29)
24n-1
where the trace is defined as
(c

Tr (n)= å R (α Q ) (30)
Q ÎQn

Amity Directorate of Distance & Online Education

 130 Graph theory and Combinatorics

Let fO ( x) be the generating function for the number of partitions PO (n) of n

Notes

e
containing odd numbers only and f D ( x) be the generating function for the number of
partitions PD (n) of nwithout duplication, then

in
fO ( x) = f D ( x) (31)
¥ ¥

= Õ åx ik
(32)

nl
k =1,3,.. i = 0

1
= ¥ (33)
Õ k =1,3,..
1-x k

O
¥
= Õ(1+ x )
k =1
(34)
k

1
= (-q; x)¥ (35)
2

ity
2 3 4 5
= 1+ x+ x +2 x +2 x +3 x +..... (36)

The identity
¥ ¥
Õ (1+ z k
)=Õ (1-z 2 k -1 )-1 (37)

rs
k =1 k =1

is known as the Euler identity

The generating function for the difference between the number of partitions into
ve
an even number of unequal parts and the number of partitions in an odd number of
unequal parts is given by
¥

Õ (1-z k
)=1-z-z 2 + z 5 + z 7 -z12 -z15 +... (38)
ni

k =1

where

 ( −1)n 1
for k of the form n (3 n ±1)
(40)
ck =  0 2
U

otherwise

Let PE (n) be the number of partitions of even numbers only, and let PEO (n)( PDO (n)) be
the number of partitions in which the parts are all even (odd) and all different. Then the
ity

generating function of PDO (n) is given by

∞

f DO ( x) = ∏ 1 + x k (41) k =1,3,..
2
(− x; x
= )∞ (42)
m

The number of partitions with no even part repeated is the same as the number in
which no part occurs more than three times and the same as the number in which no
part is divisible by 4, all of which share the generating functions
)A

∞
1 + x2k
P3 (n) = ∏
k =1 1 − x
2 k −1 (43)

= (1 + x ∏ k
+ x 2 k + x 3k ) (44)
(c

k =1

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 131

∞
1 − x4k
∏ Notes

e
= k (45)
k =1 1 − x

in
( x 4 )∞
= (46)
( x)∞

nl
In general, the generating function for the number of partitions in which no part
occurs more than times is
∞ d
Pd (n) = ∏ ∑ xik

O
(47)
k =1 i = 0

1 − x ( d +1) k
∞

= ∏ 1 − xk
(48)

ity
k =1

The generating function for the number of partitions in which every part occurs 2, 3,
or 5 times is
∞
P2,3,5 (n) ∏ (1 + x 2k
+ x3k + x5 k ) (49)
=

rs
k =1

= (1 + x∏
k =1
2k
)(1 + x3k ) (50)
ve
∞
1 − x4k 1 − x6k
= ∏1− x
k =1
2k
1 − x3k
(51)

( x 4 )∞ ( x 6 )∞
ni

= (52)
2
( x )∞ ( x3 )∞

The first few values are 0, 1, 1, 1, 1, 3, 1, 3, 4, 4, 4, 8, 5, 9, 11, 11, 12, 20, 15, 23, ....
U

The number of partitions in which no part occurs exactly once is

∞
P1 (n) = (1 + x 2 k + x3k + ...) (53)
∏ k =1
ity

∞
1 − xk + x2k
= ∏
k =1 1 − xk
(54)

∞
1 + x3k
= ∏ 2 k (55)
m

k =1 1 − x

∞
1 − x6k
= ∏ 2k 3 k
k =1 (1 − x )(1 − x )
(56)
)A

Remark:
1. The number of partitions of n in which no even part is repeated is the same as
the number of partitions of n in which no part occurs more than three times and
also the same as the number of partitions in which no part is divisible by four.
(c

2. The number of partitions of n in which no part occurs more often than d times is
the same as the number of partitions in which no term is a multiple of d+1.
Amity Directorate of Distance & Online Education
 132 Graph theory and Combinatorics

3. The number of partitions of in which each part appears either 2, 3, or 5 times

Notes

e
is the same as the number of partitions in which each part is congruent mod 12
to either 2, 3, 6, 9, or 10.

in
4. The number of partitions of in which no part appears exactly once is the same
as the number of partitions of in which no part is congruent to 1 or 5 mod 6.
5. The number of partitions in which the parts are all even and different is equal to

nl
the absolute difference of the number of partitions with odd and even parts. P(n)
satisfies the inequality
1

O
P (n) ≤ [ P (n + 1) + P (n − 1)] (58)
2
P(n,k) denotes the number of ways of writing n as a sum of exactly k terms or,
equivalently, the number of partitions into parts of which the largest is exactly k. (Note

ity
that if “exactly k” is changed to “k or fewer” and “largest is exactly k,” is changed to
“no element greater than k,” then the partition function q is obtained.) For example,
P(5,3)=2, since the partitions of 5 of length 3 are {3,1,1}and {2,2,1}, and the partitions of
5 with maximum element 3 are {3,2}and {3,1,1}.

rs
The P(n,k) such partitions can be enumerated in the Integer Partitions

[n, k ].

P(n,k) can be computed from the recurrence relation

ve
P(n, k )= P(n − 1, k − 1) + P(n − k , k ) (59)

with P (n, k ) = 0 for k > n, P(n, n) =

1 and P(n, 0) = 0 . The triangle of P(k , n) is
given by
ni

1
1 1
U

1 1 1
1 2 1 1 (60)
1 2 2 1 1
ity

1 3 3 2 1 1

The number of partitions of with largest part is the same as P(n,k).

The recurrence relation can be solved exactly to give

m

P(n,1)
= 1 (61)

1
P(n, 2) =  2n − 1 + (−1) n  (62)
4
)A

1  2  2 
P(n,3) =  6n − 7 − 9(−1) n + 16 cos  π n   (63)
72   3 
where P(n,k)=0for . The functions P(n,k) can also be given explicitly for the
(c

first few values of in the simple forms

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 133

1 
Notes

e
P(n, 2) =  2 n  (65)

in
 1 2
P(n,3) = 12 n  (66)

where is the floor function and is the nearest integer function defines

nl
1
tk ( n) =
n + k (k − 3) (67)
4
and then yields

O
1 
P(n, 2) =  2 t2 (n)  (68)

ity
P(n,3) = 1 2 
12 t3 (n)  (69)

 144
1 3 1 
t4 ( n ) − t4 ( n ) 
48
 
P(n,4) =   1 t 3 ( n ) − 1 t ( n ) (70)

rs
 144 4
12
4 


Then obtained the exact asymptotic formula

P ( n)
= ∑ Pk (n) + O(n −1 4 )
ve
(71)
k <α n

where is a constant. However, the sum diverges

∞
∑ P (n) (72)
ni

k
k =1

Counting Functions:
U

Combinatorics can be traced back more than 3000 years to India and China.
For many centuries, it primarily comprised the solving of problems relating to the
permutations and combinations of objects. The use of the word combinatorial” can be
traced back to Leibniz in his dissertation on the art of combinatorial in 1666. Over the
ity

centuries, combinatorics evolved in recreational pastimes.

These include the Konigsberg bridges problem, the four-colour map problem,
the Tower of Hanoi, the birthday paradox and Fibonacci’s ‘rabbits’ problem. In the
modern era, the subject has developed both in depth and variety and has cemented
m

its position as an integral part of modern mathematics. Undoubtedly part of the reason
for this importance has arisen from the growth of computer science and the increasing
use of algorithmic methods for solving real-world practical problems. These have led
)A

to combinatorial applications in a wide range of subject areas, both within and outside
mathematics, including network analysis, coding theory, and probability.

Addition and multiplication rules:

1) How many possible crossword puzzles are there?
(c

Suppose we have to select 4 balls from a bag of 20 balls numbered 1 to 20.

Amity Directorate of Distance & Online Education

 134 Graph theory and Combinatorics

2) How often do two of the selected balls have consecutive numbers?

Notes

e
How many ways are there of rearranging the letters in the word ALPHABET?
We observe various things about the above problems. A priori, unlike many

in
problems in mathematics, there is hardly any abstract or technical language. Despite
the initial simplicity, some of these problems will be frustratingly di cult to solve. Further,
we notice that despite these problems appearing to being diverse and unrelated, they

nl
principally involve selecting, arranging, and counting objects of various types. We will
address the problem of counting. Clearly, we would like to be able to count without
actually counting.

O
In other words, can we find out how many things there are with a given property
without actually enumerating each of them. Quite often this entails deep mathematical
insight. We now introduce two standard techniques which are very useful for counting

ity
without actually counting. These techniques can easily be motivated through the
following examples.

Example:

Let the cars in New Delhi have license plates containing 2 alphabets followed by

rs
two numbers. What is the total number of license plates possible?

Solution:
ve
Here, we observe that there are 26 choices for the first alphabet and another 26
choices for the second alphabet.

After this, there are two choices for each of the two numbers in the license plate.
Hence, we have a maximum of 26 × 26 × 10 × 10 = 67600 license plates.
ni

Example:

Let the cars in New Delhi have license plates containing 2 alphabets followed by
U

two numbers with the added condition that \in the license plates that start with a vowel
the sum of numbers should always be even”. What is the total number of license plates
possible?

Solution:
ity

Here, we need to consider two cases.

Case 1:

The license plate doesn’t start with a vowel. Then using the previous example, the
m

number of license plates equals

21 × 26 × 10 × 10 = 54600.
)A

Case 2:
The license plate starts with a vowel. Then the number of license plates equals

5 × 26 × (5 × 5+5 × 5)=6500.
(c

Hence, we have a maximum of 54600 + 6500 = 61100 license plates.

Generalization of the first example leads to what is referred to as the rule of

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 135

product and that of the second leads to the rule of addition. To understand these rules,
Notes

e
we explain the involved ideas.

Suppose we have a task to complete and that the task has some parts. Assume

in
that each of the parts can be completed on their own and completion of one part does
not result in the completion of any other part. We say the parts are compulsory to mean
that each of the parts must be completed to complete the task. We say the parts are

nl
alternative to mean that exactly one of the parts must be completed to complete the
task. With this setting we state the two basic rules of combinatorics.

Basic counting rule:

O
Let n, m1,…mn ∈ N.

Multiplication/Product rule:

ity
If a task consists of n compulsory parts and the i-th part can be completed in mi
ways, i = 1, 2,…..n, then the task can be completed in m1,m2,…….,mn ways.

Addition rule:

rs
If a task consists of n alternative parts, and the i-th part can be completed in mi
ways, i = 1; : : : ; n, then the task can be completed in
ve
m1+ m2+….. + mnways.

Let us consider the following examples.

1. How many three digit natural numbers can be formed using digits
ni

0,1,….., 9?
Identify the number of parts in the task and the type of the parts (compulsory or
alternative).
U

Solution:

The task of forming a three digit number can be viewed as lining three boxes kept
in a horizontal row. It has three compulsory parts.
ity

Part 1:
Choose a digit for the left most place.
m

Part 2:
Choose a digit for the middle place.
)A

Part 3:
Choose a digit for the rightmost place.

Apply multiplication rule.

(c

9 × 10 × 10=900

Amity Directorate of Distance & Online Education

 136 Graph theory and Combinatorics

Prime Counting Function:

Notes

e
The prime counting function is the function giving the number of primes less
than or equal to a given number ‘x’. For example, there are no primes , so

in
. There is a single prime (2) , so . There are two primes (2 and 3) , so
. And so on.

The notation for the prime counting function is slightly unfortunate because

nl
it has nothing whatsoever to do with the constant . This notation was
introduced by number theorist Edmund Landau in 1909 and has now become standard.
In the words of Derbyshire (2004, p. 38), “I am sorry about this; it is not my fault. You’ll

O
just have to put up with it.”

Letting denote the ‘n’th prime, is a right inverse of since

π(Pn) = n (1)

ity
for all positive integers. Also,

pπ ( n ) = n (2)

iff ‘n’is a prime number.

The first few values of for

rs
, 2, ... are 0, 1, 2, 2, 3, 3, 4, 4, 4, 4, 5, 5, 6, 6, 6,
.... The Wolfram Language command giving the prime counting function for a number
is PrimePi[x], which works up to a maximum value of .
ve
The notation is used to denote the modular prime counting function, i.e., the
number of primesof the form less than or equal to .

One of the most fundamental and important results in number theory is the
ni

asymptotic form of as becomes large. This is given by the prime number theorem,
which states that
π(n) ~ Li(n), (3)
U

where is the logarithmic integral and is asymptotic notation. This relation

was first postulated by Gauss in 1792.

Prime Counting Function:

ity

The prime counting function can be expressed by Legendre’s formula, Lehmer’s

formula, Mapes’ method, or Meissel’s formula. A brief history of attempts to calculate
is given by Berndt (1994). The following table is taken from Riesel (1994), where
is asymptotic notation.
m

An approximate formula due to illustrated above, is given by

n
π ( n) ≈ (4)
)A

hn
where is related to the harmonic number by . This formula is
within of the actual value for . The values for which
are 1, 109, 113, 114, 199, 200, 201, ... This quantity is positive for all .
(c

An upper bound for is given by

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 137

1.25506 n
π ( n) < Notes

e
(5)
1nn
for , and a lower bound by

in
n
< π (n) (6)
1nn

nl
for .

It gives the formula

O
n   ( j − 2 ) ! 
π (n) =−1 + ∑ ( j − 2 ) !− j    (7)
j =3   j  

for , where is the floor function.

ity
A modified version of the prime counting function is given by

 π ( p) for p composite
π 0 ( p) ≡  1
(8)
π ( p ) − 2 for p rime

π 0 ( p) ≡ ∑
∞

n =1
µ ( x) f ( x1 n )
n
(9)
rs
ve
where is the Möbius function and is the Riemann prime counting function.

It also shows that

d π ( x ) 1 ∞ µ ( n)
ni

 ∑ x1 n (10)
dx x1nx n =1 n

where is the Möbius function.

U

The smallest such that for , 3, ... are 2, 27, 96, 330, 1008, ...,
and the corresponding are 1, 9, 24, 66, 168, 437, .... The number of solutions of
for , 3, ... are 4, 3, 3, 6, 7, 6, ... .
ity

It shows that for sufficiently large ,

ex  x 
π 2 ( x)< π   (11)
1nx  e 
m

This holds for , 9, 10, 12, 14, 15, 16, 18, ... (OEIS A091886). The largest
known prime for which the inequality fails is 38358837677. The related inequality

ex  x 
[1i( x)]2 < 1i  
)A

(12)
1nx  e 
where
dt
li ( x) ∫0x (13)
ln t
(c

is true for and no smaller number.

Amity Directorate of Distance & Online Education

 138 Graph theory and Combinatorics

Riemann Counting function:

Notes

e
Riemann defined the function by

in
f ( x) ≡ ∑ 1
(1)
v
p v sx

p prime

nl
[1gx ] π ( x )
1n

=
∑n =1 n
(2)

O
1 1
=
π ( x) + π ( x1 2 ) + π ( x1 3 ) + ... (3)
2 3
Amazingly, the prime counting function is related to by the Möbius
transform

ity
∞
µ ( n)
π ( x) = ∑ f ( x1 n ) (4)
n =1 n

where is the Möbius function. More amazingly still, is connected with the

rs
Riemann zeta function by

1n [ζ ( s ) ] ∞
= ∫ f ( x) x − s −1dx (5)
s 0
ve
is also given by

1 2+iT x s
t →∞ 2π i ∫2 −iT x
f ( x) = lim 1nζ ( s )ds (6)
ni

where is the Riemann zeta function, and (5) and (6) form a Mellin transform pair.
U
ity
m

Riemann (1859) proposed that

∞ dt
f ( x) = 1i ( x) − ∑ 1i ( x p ) − 1n2 + ∫ (7)
ρ
x t (t − 1)1nt 5
2
)A

where is the logarithmic integral and the sum is over all nontrivial zeros of
the Riemann zeta function .

Actually, since the sum of roots is only conditionally convergent, it must be summed
in order of increasing even when pairing terms with their “twins” , so
(c

∑1i( x p )
=
ρ
∑
ρ
Ι ( )>0
 Li ( x p ) + Li ( x1− ρ )  (8)

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 139

This formula was subsequently proved. The integral on the right-hand side
Notes

e
converges only for , but since there are no primes less than 2, the only values of
interest are for . Since it is monotonic decreasing, the maximum therefore occurs

in
at , which has value
∞ dt
∫2 t1nt (t 2 − 1)
= 0.14001010114328692668... (9)

nl
O
ity
rs
ve
Riemann also considered the function
ni

∞
µ ( n)
R( x) = ∑ 1i ( x1 n ) (10)
n =1 n
U

1n
Sometimes also denoted obtained by replacing f ( x ) in the Riemann
function with the logarithmic integral 1i ( x1 n ) , where ζ (z) is the Riemann zeta function
and μ(n) is the Möbius function which illustrate the fact that R(x) has a series of zeros
near the origin. These occur at 10-x for x=14827.7, 15300.7, 21381.5, 25461.7, 32711.9,
ity

40219.6, 50689.8, 62979.8, 78890.2, 98357.8, ..., corresponding to x = 1.829 X 10–

14828
, 2.040 X 10–15301, 3.289 X 10–21382, 2.001 X 10–25462, 1.374 X 10–32712, 2.378 X 10–40220,
1.420 X 10–50690, 1.619 X 10–62980, 6.835 X 10–78891, 1.588 X 10–98358, ...
m
)A
(c

The quantity is plotted above.

Amity Directorate of Distance & Online Education

 140 Graph theory and Combinatorics

This function is implemented in the Wolfram Language as Riemann R[x].

Notes

e
Ramanujan independently derived the formula for , but non rigorously. The
following table compares and for small . Riemann conjectured that

in
, but this was disproved by Littlewood in 1914.

nl
O
ity
The Riemann prime counting function is identical to the Gram series
∞
(1nx) k

rs
G ( x) = 1 + ∑ (11)
k =1 kk !ζ ( k + 1)

where is the Riemann zeta function, but the Gram series is much more
ve
tractable for numeric computations. For example, the plots above show the difference
where is computed.

Questions:
ni

1. A partition of a positive integer, also called an __________ partition.

a) Rational
b) integer
U

c) both a and b
d) none
ity

2. Two sums that differ only in the order of their summands are considered the
_______ partition.
a) Same
b) different
m

c) both a and b
d) none
)A

3. A summand in a partition is also called a ________.

a) Odd part
b) even part
c) part
(c

d) both a and b

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 141

4. Partitions can be graphically visualized with _____________.

Notes

e
a) Young diagrams or Ferrers diagrams.
b) Counting functions

in
c) Necklaces
d) Euler’s function

nl
5. If we count partitions in which no number occurs more than once. Such a partition
is called a ___________.
a) Partition with distinct parts.

O
b) Counting functions
c) Necklaces

ity
d) Euler’s function

Answer:
1. b

rs
2. a
3. c
4. a
ve
5. a

Exercises:
ni

1. Define Partition
2. Explain Counting Functions
3. Explain generating functions with an example
U

4. Explain Euler’s Function

ity
m
)A
(c

Amity Directorate of Distance & Online Education

 142 Graph theory and Combinatorics

Unit- VIII
Notes

e
Necklaces:

in
nl
O
ity
rs
ve
In the technical combinatorial sense, an -ary necklace of length is a string of ‘n’
characters, each of possible types. Rotation is ignored, in the sense that is
equivalent to for any .

In fixed necklaces, reversal of strings is respected, so they represent circular

ni

collections of beads in which the necklace may not be picked up out of the plane (i.e.,
opposite orientations are not considered equivalent). The number of fixed necklaces of
length composed of types of beads is given by
U

1 v(n)
N (n, a ) = ∑ φ (di )a n di (1)
n i =1
where are the divisors of with , , ..., , is the number of
ity

divisors of , and is the totient function.

For free necklaces, opposite orientations (mirror images) are regarded as

equivalent, so the necklace can be picked up out of the plane and flipped over. The
number of such necklaces composed of beads, each of possible colors, is
m

given by

 φ ( di ) an di + na( n+1) 2
v(n)

1 ∑
for n odd
)A

'
N (n, a ) =  vi(=n1) (2)
2n  ∑ φ ( di ) an di + 12 n (1+ a ) an 2 for n even
 i=1
For and an odd prime, this simplifies to
(c

2 ρ −1 − 1
N ' ( ρ , 2) = + 2( ρ −1) 2 + 1 (3)
ρ
Amity Directorate of Distance & Online Education
 Graph theory and Combinatorics 143

Notes

e
in
nl
O
A table of the first few numbers of necklaces for and follows. Note that

ity
is larger than for . For , the necklace 110100 is inequivalent
to its mirror image 001011, accounting for the difference of 1 between and
.

Similarly, the two necklaces 0010110 and 0101110 are inequivalent to their
reversals, accounting for the difference of 2 between

n N (n, 2)
and

N' (n, 2) rs .

N' (n, 3)
ve
Sloane A000031 A000029 A027671
1 2 2 3
2 3 3 6
ni

3 4 4 10
4 6 6 21
5 8 8 39
U

6 14 13 92
7 60 46 1219
8 108 78 3210
ity

9 188 126 8418

10 352 224 22913
11 632 380 62415
12 1182 687 173088
m

13 2192 1224 481598

Let us consider the problem of finding the number of distinct arrangements of

people in a ring such that no person has the same two neighbors two or more times.
)A

For 8 people, there are 21 such arrangements.

Example:

A jeweler sells six-beaded necklaces in his shop. Given that there are two different
colors of beads, how many varieties of necklaces does he need to create in order to
(c

have every possible permutation of colors?

Amity Directorate of Distance & Online Education

 144 Graph theory and Combinatorics

Notes

e
in
Given 6 beads we have 2 choices of color per bead, so there are 64 ways to color
an unmovable beaded necklace.

nl
However, if we are given the freedom to rotate and reflect it, there are only 13 distinct
varieties. Furthermore, if the colors are interchangeable (i.e. a completely light necklace
is equivalent to a completely dark necklace), then there are only 8 distinct arrangements.

O
ity
rs
The set of necklaces with 0, 1, or 2 beads of one color.

We might guess that it has something to do with symmetry. The hexagon has 12
ve
symmetries: rotation by 0, 60, 120, 180, 240, or 300 degrees, three reactions through
opposite vertices, and three reactions through opposite sides. There is no obvious
relationship between the number of possibilities for a xed necklace, the number of
symmetries, and the number of distinct varieties, but surely there is one. In this paper,
ni

we use combinatorics and group theory to work through the problem of the six-beaded
necklace and others like it.

One of the most important and beautiful themes unifying many areas of modern
U

mathematics is the study of symmetry. Many of us have an intuitive idea of symmetry,

and we often think about certain shapes or patterns as being more or less symmetric
than others. A square is in some sense “more symmetric” than a rectangle, which in
ity

turn is “more symmetric” than an arbitrary four-sided shape. Can we make these ideas
precise? Group theory is the mathematical study of symmetry, and explores general
ways of studying it in many distinct settings. Group theory ties together many of the
diverse topics we have already explored – including sets, cardinality, number theory,
isomorphism, and modular arithmetic.
m

Shapes and Symmetries:

Many people have an intuitive idea of symmetry. The shapes in Figure 38 appear
)A

symmetric, some perhaps more so than others. However, despite our general intuitions
about symmetry, it may not be clear how to make this statement precise. Can it make
sense to discuss “how much” symmetry a shape has? Is
(c

Figure 38: Some symmetric polygons.

Amity Directorate of Distance & Online Education
 Graph theory and Combinatorics 145

there some way to make precise the idea that the regular pentagon is “more
Notes

e
symmetric” than the equilateral triangle, or that the circle is “more symmetric” than any
regular polygon? In this section we will explore symmetry and the way in which it arises

in
in various contexts with which we are familiar, especially in the geometry of regular
polygons (2D) and regular polyhedra (3D), such as the Platonic solids. The study of
symmetry is a recurring theme in many disparate areas of modern mathematics, as well
as chemistry, physics, and even economics.

nl
To help us explore the idea of symmetry, we begin by considering a single concrete
example, the equilateral triangle below.

O
ity
Rotation symmetries
An equilateral triangle can be rotated by 120, 240, or 360 angles without really
changing it. If you were to close your eyes, and a friend rotated the triangle by one
of those angles, then after opening your eyes you would not notice that anything had

rs
changed. In contrast, if that friend rotated the triangle by 31 or 87, you would notice that
the bottom edge of the triangle is no longer perfectly horizontal.

Many other shapes that are not regular polygons also have rotational symmetries.
ve
The shapes illustrated in Figure 39, for example, each have rotational symmetries. The
first example can be rotated only 180, or else 360 or 0.
ni

Figure 39: Several shapes with rotational symmetries.

U

The third shape can be rotated any integer multiple of 90. The fourth shape can be
rotated any integer multiple of 72. The fifth shape can be rotated any integer multiple of 60.

More generally, we say that a shape has rotational symmetry of order n if it can
ity

be rotated by any multiple of 360 /n without changing its appearance. We can imagine
constructing other shapes with rotational symmetries of arbitrary order. If the only
rotations that leaves a shape unchanged are multiples of 360, then we say that the
shape has only the trivial (order n = 1) symmetry.
m

Mirror reflection symmetries:

Another type of symmetry that we can find in two-dimensional geometric shapes is
)A

mirror reflection symmetry. More specifically, we can draw a line through some shapes
and reflect the shape through this line without changing its appearance. This is called a
mirror reflection symmetry.

Further consideration of the equilateral triangle (cf. Figure 40) shows that there
(c

are actually three distinct mirror lines through which we can reflect the shape without
changing its appearance.

Amity Directorate of Distance & Online Education

 146 Graph theory and Combinatorics

If we were to reflect the triangle through any other line, the shape as a whole
Notes

e
would look different.

in
Counting symmetries:
One way in which we can quantify the “amount” of symmetry of an object is by
counting its number of symmetries. For example, we might count the number of

nl
rotational symmetries of an object, along with its mirror reflection symmetries.
However, counting the symmetries of a shape can be challenging. It is not immediately
clear which symmetries we should count and which, if any, we should not count.

O
To understand why we might not count certain symmetries, consider rotating the
equilateral triangle by 120, 240 and 360. Of course the numbers by which we are
rotating the triangle are different, and so we might be inclined to count each of them
separately. But notice that we can also rotate the triangle by 480, 600 and 720. Should

ity
we count those as different symmetries? If we do count them, then what would stop
us from counting an infinite number of rotational symmetries for a triangle, or for that
matter, any shape?

One way to limit the number of symmetries we count involves coloring, or

rs
otherwise labeling, the shape. For example, we can color each edge of the equilateral
triangle, as illustrated in Figure 42. Symmetries can then be captured as changes of
colors that leave the uncolored shape fixed. Any triangle in either row can be obtained
from any other triangle in that row through a rotation; triangles can be obtained from
ve
triangles in the other row through reflections.

Using this coloring allows us to count symmetries carefully. If changing the

shape in two different ways results in the same coloring, then we should count those
ni

two symmetries as the same. For example, rotating the equilateral triangle by 120 or
480 results in the same coloring, so we count those as the same symmetry. Likewise,
rotating the triangle by 0 and 360 also result in Figure 42: Equilateral triangle with
edges colored.
U
ity
m

Any triangle in either row can be obtained from another triangle in the row through
a rotation; triangles can be obtained from triangles in the other row through reflections.
The same coloring, so we count those the same as well. To reduce confusion, we use a
)A

number between 0 and 360 (not including 360 itself) to describe the angle of a rotation;
thus, we prefer 120 to 480, despite their equivalence rotations, or “doing nothing” to
360 rotations, despite their equivalence. We are therefore left with six symmetries
of the triangle – the rotations (0, 120, and 240), and three reflections, one for each
of the mirror planes passing through a corner and the center of the triangle. These
(c

symmetries can be pictured by how they transform the colored triangle.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 147

Symmetries of the square:

Notes

e
A square is in some sense “more symmetric” than a triangle because it has more
symmetries. Figure 43 below shows a square with colored edges arranged in different

in
ways. Again you might notice that any two squares in the same row can be obtained
from one another through rotations, whereas those in distinct rows can only be obtained
from one another through a reflection. Some thought will show that there are no other

nl
rotations or mirror reflection symmetries, and so these figures represent all eight
symmetries of the square.

O
ity
Squares

Although this section is concerned primarily with rotational and mirror re-flection

rs
symmetries of single objects in two dimensions, other types of symmetries arise in
infinite systems and in higher dimensions.

Rotational Symmetry:
ve
Abstractly, a spatial configuration is said to possess rotational symmetry if
remains invariant under the group . Here, denotes the group of
rotations of and is viewed as a subgroup of the automorphism group of all
ni

automorphisms which leave unchanged. A more intuitive definition of rotational

symmetry comes from the case of planar figures in Cartesian space.

For ‘d’ arbitrary, a geometric object in is said to possess rotational symmetry if

U

there exists a point so that the object, when rotated a certain number of degrees (or
radians) about said point, looks precisely the same as it did originally. This notion
can be made more precise by counting the number of distinct ways the object can
be rotated to look like itself; this number is called the degree or the order of the
ity

symmetry.

Rotational symmetry of degree corresponds to a plane figure being the same

when rotated by degrees, or by radians.
m
)A
(c

Amity Directorate of Distance & Online Education

 148 Graph theory and Combinatorics

The regular pentagon in the figure above has a rotational symmetry of order 5 due
Notes

e
to the fact that rotating it about the center point by radians, n = 0, 1, 2, 3,
4, yields the exact same figure. This is a particular example of a more general fact,

in
namely that any regular planar n-gon has rotational symmetry of order . In the case
of the regular planar -gon, the collection of all such symmetries is a group denoted by
, is isomorphic to the cyclic group of integersmodulo , and is a proper
subgroup of the dihedral group of all symmetries--rotational and otherwise--of the

nl
figure.

By the definition given above, rotational symmetry of degree 1 corresponds to

O
an object having symmetry about a point only when rotated by degrees;
clearly, this condition is satisfied only by objects which have no symmetry, i.e., those
objects whose rotational symmetry group is trivial. Therefore, the simplest possible
rotational symmetry is of order 2 and is possessed, e.g., by planar parallelograms.

ity
In some literature, rotational symmetry of order is defined by classifying the
results of rotating a figure about a line rather than about a point (Weyl 1982). In
particular, such sources define a figure to have rotational symmetry of order if the
figure which remains identical after a -radian rotation about (which is called the

rs
axis of rotation). These two perspectives yield the same result, however; for example,
in the figure above, the -radian clockwise rotation of the pentagon about its
center point can equivalently be viewed as a -radian clockwise rotation about the
segment/line determined by the center point and the top right vertex.
ve
In the -dimensional Cartesian space , the -sphere has complete
rotational symmetry in that its shape remains identical after any -radian rotation about
any line . Historically, this fact led some ancient civilizations to consider the circle and/
ni

or the sphere to be divine (Weyl 1982).

In addition to being a well-studied concept mathematically, rotational symmetry

is also a far-reaching notion due to the prevalence of such symmetry among many
U

naturally-occurring objects including snowflakes, crystals, and flowers.

Knot Symmetry:
ity

A symmetry of a knot is a homeomorphism of which maps onto itself. More

succinctly, a knot symmetry is a homeomorphism of the pair of spaces . Hoste et
al. (1998) consider four types of symmetry based on whether the symmetry preserves
or reverses orienting of and ,
m

1. preserves , preserves (identity operation),

2. preserves , reverses ,
3. reverses , preserves ,
)A

4. reverses , reverses .
This then gives the five possible classes of symmetry summarized in the table
below.
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 149

class symmetries knot symmetries

Notes

e
c 1 chiral, noninvertible
+ 1, 3 + amphichiral, noninvertible

in
– 1, 4 – amphichiral, noninvertible
i 1, 2 chiral, invertible
a 1, 2, 3, 4 + and – amphichiral, invertible

nl
In the case of hyperbolic knots, the symmetry group must be finite and either
cyclic or dihedral (Riley 1979, Kodama and Sakuma 1992, Hoste et al. 1998). The
classification is slightly more complicated for nonhyperbolic knots. Furthermore, all

O
knots with crossings are either amphichiral or invertible (Hoste et al. 1998). Any
symmetry of a prime alternating link must be visible up to flypes in any alternating
diagram of the link.

ity
rs
ve
The following tables (Hoste et al. 1998) give the numbers of -crossing knots
belonging to cyclic symmetry groups (Sloane’s A052411 for and A052412 for ) and
ni

dihedral symmetry groups (Sloane’s A052415 through A052422). Of knots with 16

or fewer crossings, there are only one each having symmetry groups , , and
(above left). There are only two knots with symmetry group , one hyperbolic (above
U

right), and one a satellite knot. In addition, there are 2, 4, and 10 satellite knots having
14-, 15-, and 16-crossings, respectively, which belong to the dihedral group .
ity

1 0 0 0 0
2 0 0 0 0
3 0 0 0 0
4 0 0 0 0
5 0 0 0 0
m

6 0 0 0 0
7 2 0 0 0
8 24 3 0 0
)A

9 173 14 0 0
10 1047 57 0 0
11 6709 210 0 0
12 37177 712 0 2
(c

13 224311 2268 1 0
14 1301492 7011 0 11

Amity Directorate of Distance & Online Education

 150 Graph theory and Combinatorics

Notes

e
1 0 0 0 0 0 0 0 0 0 0 0 0

in
2 0 0 0 0 0 0 0 0 0 0 0 0
3 0 0 0 0 0 0 0 0 0 0 0 0
4 0 0 1 0 0 0 0 0 0 0 0 0

nl
5 0 4 0 2 0 0 0 0 0 0 0 0
6 4 12 0 3 0 0 0 1 0 0 0 0

O
7 13 23 3 4 0 3 0 0 0 0 0 0
8 66 62 1 5 0 1 0 0 0 1 0 0
9 217 134 2 11 0 0 0 0 0 0 0 0

ity
10 728 309 6 18 0 8 1 2 0 0 0 0
11 2391 647 1 21 2 3 1 2 0 0 0 0
12 7575 1463 4 31 2 2 0 0 0 0 1 0

rs
15 23517 3065 50 53 3 12 0 2 1 4 0 0
16 73263 6791 15 89 0 10 1 8 1 1 0 1
ve
Euler’s Function:
Zn, it is frequently fruitful to ask whether something comparable applies to Un.
Here we look at Un in the context of the previous section. To aid the investigation, we
introduce a new quantity, the Euler phi function, written ϕ(n), for positive integers n
ni

Definition:
U

ϕ(n) is the number of non-negative integers less than n that are relatively prime to
n. In other words, if n>1 then ϕ(n) is the number of elements in Un, and ϕ(1)=1

Example:
ity

ϕ(2)=1, ϕ(4)=2, ϕ(12)=4 and ϕ(15)=8

Example:

If p is a prime, then ϕ(p)=p−1, because 1, 2, …, p−1 are all relatively prime to p,

and 0 is not.
m

For any number n, ϕ(n) turns out to have a remarkably simple form; that is, there
is a simple formula that gives the value of ϕ(n). We’ve already seen how simple it is for
)A

primes. As is typical of many results in number theory, we will work our way gradually to
any n,looking next at powers of a single prime.

Euler’s ϕ function:
ϕ (n) is the number of integers m ∈ [1,n] with m coprime to n.
(c

Or, it is the order of the unit group of the ring Z/nZ.

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 151

Euler:
Notes

e
If a is coprime to n, then

in
aϕ(n) ≡ 1(modn)

Euler’s theorem is the basis of the RSA Cryptosystem:

If integers E,D satisfy ED ≡ 1 (mod ϕ (n)), then

nl
a ED ≡ a (mod n),
For every integer a coprime to n, (In fact, this holds for all integers a if n is square

O
free, such as the product of two different large primes.)

Encrypt message “a” : b ≡aE

(mod n),

Decrypt: a = bD(mod n).

ity
To encrypt, one should know E ,n. To decrypt, D as well.

Note that it is easy to compute aE (mod n) given a, E, n and similarly it is easy to

compute bD(mod n).

Further, given ϕ (n), it is easy to come up with pairs E,D with

rs
ED ≡ 1 (mod ϕ (n)). Indeed, keep choosing random numbers D until one is found
that is coprime to ϕ (n), and then use Euclid’s algorithm to find E.
ve
As a public-key system, E, n are released to the public, but D is kept secret. Then
anyone can send you encrypted messages that only you can read.

The security of the RSA Cryptosystem is connected to our ability to compute ϕ (n):
ni

For p prime, ϕ (p)= p-1; more generally,

ϕ (p)= p k
– pk-1= pk(1-1/p).
U

By the Chinese Remainder Theorem,

ϕ (n)= n ∏ (1 − 1 / p) .
pprime
ity

p/n

So, knowing the prime factorization of n, one can compute ϕ (n) rapidly—in
deterministic polynomial time.

What about the converse? Given n and ϕ (n) can one compute the prime
m

factorization of n in deterministic polynomial time?

Yes, if n=pq , where p ,q are different primes, since

ϕ (n) = (p-1) (q-1), so that n-1- ϕ (n) = p + q.

)A

In general, the Extended Riemann Hypothesis implies that there is a deterministic,

polynomial time algorithm to compute the prime factorization of n, given n and ϕ (n).

A random polynomial time algorithm to get a nontrivial factorization:

(c

We may assume ϕ (n) < n-1 and that n and ϕ (n) are coprime.
Write ϕ (n) = 2 s m, where m is odd.
Amity Directorate of Distance & Online Education
 152 Graph theory and Combinatorics

Choose a at random from [1,n-1]. We may assume a and n are coprime.

Notes

e
Note that,

in
aϕ(n) – 1 ≡ (am – 1)(am + 1)(a2m + 1)...(a2s –1 + 1)

Theorem:

nl
If pis a prime and a is a positive integer, then

ϕ(pa)=pa−pa−1

O
Proof:
The number of non-negative integers less than n=pa that are relatively prime to n.
As in many cases, it turns out to be easier to calculate the number that are not relatively

ity
prime to n, and subtract from the total. List the non-negative integers less than pa: 0, 1,
2, …, pa−1; there are pa of them.

The numbers that have a common factor with pa (namely, the ones that are not
relatively prime to n) are the multiples of p: 0, p, 2p, …, that is, every p th number.

rs
There are thus pa/p=pa−1 numbers in this list,

so ϕ(pa)=pa−pa−1

Example:
ve
ϕ(32)=32−16=16,

ϕ(125)=125−25=100
ni

Example:

Since
U20={[1],[3],[7],[9],[11],[13],[17],[19]},
U

U4={[1],[3]},
U5={[1],[2],[3],[4]},
ity

both U20 and U4×U5 have 8 elements.

In fact, the correspondence discussed in the Chinese Remainder Theorem

between Z20 and Z4×Z5 is also a 1-1 correspondence between U20 and U4×U5:
m

[1] ↔ ([1],[1])
[3] ↔ ([3],[3])
[7] ↔ ([3],[2])
)A

[9] ↔ ([1],[4])
[11] ↔ ([3],[1])
[13] ↔([1],[3])
[17] ↔([1],[2])
(c

[19]↔ ([3],[4])

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 153

Theorem:
Notes

e
If a and b are relatively prime and n=ab, then ϕ(n)=ϕ(a)ϕ(b)

in
Proof.
We want to prove that |Un|=|Ua|⋅|Ub|

nl
As indicated in the example, we will actually prove more, by exhibiting a one to one
correspondence between the elements of Un and Ua×Ub.

We already have a one to one correspondence between the elements of Zn

O
and Za×Zb. Again as indicated by the example, we just have to prove that this same
correspondence works for Un and Ua×Ub.

That is, we already know how to associate any [x] with a pair ([x],[x]);

ity
we just need to know that [x]∈Un if and only if ([x],[x])∈Ua×Ub.

⇒ [x] is in Un if and only if (x,n)=1 if and only if (x,a)=1 and (x,b)=1 if and only if
([x],[x])∈Ua×Ub

Euler’s function on average

rs
The chance that two random integers are both divide by the prime p is 1/p2.
ve
Should be

α := ∏ (1 − 1 / p
pprime
2
).

We have
ni

-1-1
æ p 2 ÷ö æ 1 1 ÷ö
ç 2 ç
1-1/ p = ç 2 ÷÷ = ç1 + 2 + 2 + ...÷÷
çè p -1÷ø çè p p ÷ø
U

So that
−1
 ∝ 1  1 6
α = ∑ 2  = = 2.
ity

 n =1 n  ζ ( 2) π

6
Thus the probability that two random integers are coprime is .
π2
Another interpretation: choose a lattice point (m, n) at random, the probability that it
m

6
is “visible” is .
π2
Using these thoughts it is easy to see that
)A

3
∑ ϕ ( n) ≈ π
n≤ x
2
x2

As x 
→ ∝ .
Extreme values of ϕ (n) are also fairly easy:
(c

ϕ (n)
lim sup = 1,
n
Amity Directorate of Distance & Online Education
 154 Graph theory and Combinatorics

Notes

e
ϕ (n)
lim inf = 0,
n

in
in fact,
ϕ ( n)
lim inf = e −γ
n
log log n

nl
ϕ (n)
Clearly jumps around a bit: if n has only large prime factors, the ratio is
n

O
close to 1, but if n has many small prime factors, it is close to 0.

One can ask for a “distribution function”. That is, let u be a real variable in [0,1] and
consider.

ity
ϕ (n)
{n: /n ≤ u}.
n
Euler’s Totient Function:
The totient function , also called Euler’s totient function, is defined as the
number of positive integers

rs
that are relatively prime to (i.e., do not contain any factor
in common with) , where 1 is counted as being relatively prime to all numbers.
ve
Since a number less than or equal to and relatively prime to a given number is
called a totative, the totient function can be simply defined as the number of
totatives of . For example, there are eight totatives of 24 (1, 5, 7, 11, 13, 17, 19, and
23), so .
ni

The totient function is implemented in the Euler Phi[n].

The number is called the cototient of and gives the number of positive
integers that have at least one prime factor in common with .
U

is always even for . By convention, , although the Euler Phi[0]

equal to 0 for consistency with its Factor Integer[0] command. The first few values of
for , 2, ... are 1, 1, 2, 2, 4, 2, 6, 4, 6, 4, 10, ... The totient function is given by
ity

the Möbius transform of 1, 2, 3, 4, .... is plotted above for small .

For a prime ,

φ (p) = p – 1,
m

Since all numbers less than are relatively prime to . If is a power of a

prime, then the numbers that have a common factor with are the multiples of : ,
, ..., . There are of these multiples, so the number of factors relatively prime
to is
)A

α α −1
φ ( pα ) = p − p (2)
α −1
= p ( p − 1) (3)
 1  (4)
(c

= pα 1 − 
 p

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 155

Now take a general divisible by . Let be the number of positive integers

Notes

e
not divisible by . As before, , , ..., have common factors, so

in
φ p (m) = m − (5)
p
 1

nl
= m 1 −  (6)
 p

Now let be some other prime dividing . The integers divisible by are , , ...,

O
. But these duplicate , , ..., .

So the number of terms that must be subtracted from to obtain is

m m

ity
∆φq (m) = − (7)
q pq

m çæ 1 ÷ö
= ç1- ÷ (8)
p ÷ø÷

q çè

and

φ pq (m) = φ p (m) − ∆φq (m) (9) rs

ve
 1  m 1 
= m 1 −  − 1 −  (10)
 p q  p
ni

 1  1 
= m 1 −  1 −  (11)
 p  q 
U

By induction, the general case is then

æ 1ö
φ (n) = n Õççèç1- p ÷÷ø÷÷ (12)
ity

p n

 1  1   1 
= n 1 −  1 −  .... 1 −  (13)
 p1   p2   pr 
where the product runs over all primes dividing . An interesting identity relating
m

to is given by

φ (n k ) = n k −1φ (n) (14)
)A

Another identity relates the divisors of to via

∑ φ (d ) = n
d n
(15)

The totient function is connected to the Möbius function through the sum
(c

n
∑ d µ  d  = φ (n) (16)
d

Amity Directorate of Distance & Online Education

 156 Graph theory and Combinatorics

where the sum is over the divisors of , which can be proven by induction on and
Notes

e
the fact that and are multiplicative.

The totient function has the Dirichlet generating function

in
∞
φ ( n) ζ ( s − 1)
∑
n =1 n s
=
ζ (s)
(17)

nl
The totient function satisfies the inequality

φ ( n) ≥ n (18)

O
for all except and . Therefore, the only values of for which are
, 4, and 6. In addition, for composite ,

φ (n) ≤ n − n (19)

ity
also satisfies

ln ln n
lim inf φ (n) = e − y (20)
n →∞ n

rs
where is the Euler-Mascheroni constant. The values of for which
are given by 3, 4, 5, 6, 8, 9, 10, 12, 14, 15, 16, 18, 20, 22, ..

The divisor function satisfies the congruence

ve
nσ (n) = 2(mod φ (n)) (21)

{ 0(mod φ ( n ))if φ ( n ) = 2
= 2(mod φ ( n )) otherwise (22)
ni

for all primes and no composite with the exception of 4, 6, and 22,
where is the divisor function. No composite solution is currently known to

n − 1 ≡ 0(mod φ ( n )) (23)
U

A corollary of the Zsigmondy theorem leads to the following congruence,

φ (a n + b n ) ≡ 0(mod n) (24)
ity

The first few ‘n’ for which

φ (=
n) φ (n + 1) (25)

are given by 1, 3, 15, 104, 164, 194, 255, 495, 584, 975, ... which have
common values , 2, 8, 48, 80, 96, 128, 240, 288, 480, ...
m

The only for which

φ (n) = φ (n + 1) = φ (n + 2) (26)
)A

is n=5186, giving

φ=
(5186) φ=
(5187) φ (5188) − 2534 (27)

Values of shared among that are close together include

(c

φ (25930) = φ (25935) = φ (25940) = φ (25942) (28)

= 2734 (29)

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 157

φ (404471) = φ (404473) = φ (404477) (30)

Notes

e
= 28 32 52 7 (31)

in
then for every positive integer , there are primes and such that

φ ( p) + φ (q) =
2m = (32)

φ (σ (n)) = n

nl
(33)

where is the divisor function.

O
Generating Functions:
A generating function is a formal power series
∞
f ( x) = ∑ an x n

ity
(1)
n =0

Whose coefficients give the sequence .

Given a sequence of terms, Find Generating Function [ a1, a2, ... , x] attempts to

rs
find a simple generating function in whose th coefficient is .

Given a generating function, the analytic expression for the th term in the
corresponding series can be computing using Series Coefficient [expr, x, x0, n ]. The
ve
generating function is sometimes said to “enumerate”.

Generating functions giving the first few powers of the nonnegative integers are
given in the following table.
ni

np f(x) series
1 x
x + x 2 + x 3 + ....
1− x
U

x
n 2 x + 2 x 2 + 3 x3 + 4 x 4 + ....
(1-x)
ity

x( x + 1) x + 4 x 2 + 9 x3 + 16 x 4 + ....
n2 n2

(1 − x )
3

3 x( x 2 + 4 x + 1) x + 8 x 2 + 27 x3 + ....

n 3 n
m

(1 − x )
4

x( x + 1)( x 2 + 10 x + 1) x + 16 x 2 + 81x 3 + ....

n4 n4
(1 − x )
5
)A

There are many beautiful generating functions for special functions in number
theory. A few particularly nice examples are

1
(c

f(x) = (2)
( x)∞

Amity Directorate of Distance & Online Education

 158 Graph theory and Combinatorics

∞
Notes ∑ P ( n) x

e
n
= (3)
n =0

in
= 1 + x + 2 x 2 + 3 x 3 + .... (4)

for the partition function P, where is a q-Pochhammer symbol, and

nl
x
f(x) = (5)
1 − x − x2

O
∞

= ∑F x
n =0
n
n
(6)

= x+ x 2 + 2 x 3 + 3 x 4 + ..... (7)

ity
for the Fibonacci numbers . Generating functions are very useful in combinatorial
enumeration problems.

For example, the subset sum problem, which asks the number of ways to select

The generating function of

rs
out of given integers such that their sum equals , can be solved using generating
functions.

of a sequence of numbers is given by the

ve
Z-transform of in the variable

Enumeration in the odd and even codes:

ni

To enumerate a set of objects satisfying some set of properties means to explicitly

produce a listing of all such objects. The problem of determining or counting all such
solutions is known as the enumeration problem.
U

A generating function

F ( x) = ∑ an x n
n
ity

is said to enumerate

Even Codes:
An even number is an integer of the form , where is an integer. The
m

even numbers are therefore ..., , , 0, 2, 4, 6, 8, 10, ....Since the even numbers
are integrally divisible by two, the congruence holds for even . An even
number for which also holds is called a singly even number, while an
)A

even number for which is called a doubly even number. An integer which
is not even is called an odd number.

The oddness of a number is called its parity, so an odd number has parity 1, while
an even number has parity 0.
(c

The generating function of the even numbers is

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 159

2x
=2 x + 4 x 2 + 6 x3 + 8 x 4 + .... Notes

e

( x − 1) 2
The product of an even number and an odd number is always even, as can be

in
seen by writing
(2k)(2/+1) = 2 [k(2/+1)],
which is divisible by 2 and hence is even.

nl
Parity:

O
The parity of an integer is its attribute of being even or odd. Thus, it can be said
that 6 and 14 have the same parity (since both are even), whereas 7 and 12 have
opposite parity (since 7 is odd and 12 is even).

ity
A different type of parity of an integer is defined as the sum of the bits in
binary representation, i.e., the digit count , computed modulo 2. So, for example,
the number has two 1s in its binary representation and hence has parity 2
(mod 2), or 0. The parities of the first few integers (starting with 0) are therefore 0, 1, 1,
0, 1, 0, 0, 1, 1, 0, 0, ... as summarized in the following table.

N
1
2
binary
1
10
parity
1
1
N
11
12
binary
1011
1100
rs
parity
1
0
ve
3 11 0 13 1101 1
4 100 1 14 1110 1
5 101 0 15 1111 0
ni

6 110 0 16 10000 1

7 111 1 17 10001 0
U

8 1000 1 18 10010 0
9 1001 0 19 10011 1
10 1010 0 20 10100 0
ity

A generating function for parity is given by

1 1 ∞

∏
k

 − 1 − x2 2 4 7
 =x + x + x + x + .... (1)
2  1 − x k =0 
m

The constant generated by interpreting the sequence of parity digits as a binary

fraction is called the Thue-Morse constant.

The parity function obeys the sum identity

)A

2n+1

∑ (−1)
k =0
P(k )
(k + r ) n =
0 (2)

for any . For example, for and ,

(c

1 – 4 – 9 + 16 – 25 + 36 + 49 – 64 = 0 (3)

Amity Directorate of Distance & Online Education

 160 Graph theory and Combinatorics

Odd Codes:
Notes

e
An odd number is an integerof the form , where is an integer. The odd
numbers are therefore ..., , , 1, 3, 5, 7, ... which are also the gnomonic numbers.

in
Integers which are not odd are called even.

Odd numbers leave a remainder of 1 when divided by two, i.e., the congruence
holds for odd . The oddness of a number is called its parity, so an odd

nl
number has parity 1, while an even number has parity 0.

The generating function for the odd numbers is

O
x(1 + x)
x 3 x 2 + 5 x3 + 7 x 4 + ....
=+

( x − 1) 2
The product of an even number and an odd number is always even, as can be

ity
seen by writing
(2 k) (2 / + 1) = 2 [k (2 / + 1)],
which is divisible by 2 and hence is even.

rs
Parity check matrix:
Given a linear code C of length and dimension over the field , a parity check
matrix H of C is a matrix whose rows generate the orthogonal complement of
ve
C, i.e., an element w of is a code word of C if wH=0. The rows of H generate the null
space of the generator matrix G.

Generator Matrix:
ni

Given a linear code C, a generator matrix of is a matrix whose rows generate

all the elements of C, i.e., if , then every code word w of C can be
represented as
U

w = c1 g1 + c2 g2 + ... + ck gk = cG
in a unique way, where .

An example of a generator matrix is the Golay code, which consists of all

ity

possible binary sums of the 11 rows.

Golay Code:
The Golay code is a perfect linear error-correcting code. There are two essentially
m

distinct versions of the Golay code: a binary version and a ternary version.

The binary version is a binary linear code consisting of

code words of length 23 and minimum distance 7. The ternary version is a
)A

ternary linear code, consisting of code words of length 11 with minimum

distance 5.

A parity check matrix for the binary Golay code is given by the matrix ,
where is the identity matrix and M is the matrix.
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 161

Notes

e
in
nl
By adding a parity check bit to each code word in , the extended Golay

O
code , which is a nearly perfect binary linear code, is obtained. The
automorphism group of is the Mathieu group .

A second generator is the adjacency matrix for the icosahedron, with

ity
appended, where is a unit matrix and is an identity matrix.

A third generator begins a list with the 24-bit 0 word (000...000) and repeatedly
appends first 24-bit word that has eight or more differences from all words in the list.
Conway and Sloane list many further methods.

rs
Amazingly, Golay’s original paper was barely a half-page long but has proven to
have deep connections to group theory, graph theory, number theory, combinatorics,
game theory, multidimensional geometry, and even particle physics.
ve
Hamming Code:
A binary Hamming code of length (with ) is a linear code with
ni

parity-check matrix H whose columns consist of all nonzero binary vectors of length r,
each used once. Hr is an code. Hamming codes are
perfect single error-correcting codes.
U

Triangular number :
The triangular number is a figurate number that can be represented in the form
of a triangular grid of points where the first row contains a single element and each
ity

subsequent row contains one more element than the previous one. This is illustrated
above for , , .... The triangular numbers are therefore 1, , ,
, ..., so for , 2, ..., the first few are 1, 3, 6, 10, 15, 21, ….

More formally, a triangular number is a number obtained by adding all positive

m

integers less than or equal to a given positive integer ‘n’, i.e.,

Tn
≡ ∑ (1)
)A

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

n+1
= 2 ( ) (3)
(c

Amity Directorate of Distance & Online Education

 162 Graph theory and Combinatorics

Notes ( )
n
where k is a binomial coefficient. As a result, the number of distinct wine glass

e
clinks that can be made among a group of people (which is simply ) is given by the
triangular number .

in
The triangular number is therefore the additive analog
of the factorial .

nl
A plot of the first few triangular numbers represented as a sequence of binary bits
is shown above. The top portion shows to , and the bottom shows the next 510
values.

O
The odd triangular numbers are given by 1, 3, 15, 21, 45, 55, ..., while the even
triangular numbers are 6, 10, 28, 36, 66, 78, ....

gives the number and arrangement of the tetractys (which is also the

ity
arrangement of bowling pins), while gives the number and arrangement of balls
in billiards. Triangular numbers satisfy the recurrence relation

Tn2+1 − Tn2 =(n + 1)3 (4)

as well as

Tn2 (5)
Tn2 + Tn2−1 =
3Tn + Tn −1 =
T2n (6) rs
ve
3Tn + Tn +1 =
T2 n +1
(7)

1 + 3 + 5 + ..... + (2n − 1) = Tn + Tn −1 (8)

ni
U

In addition, the triangle numbers can be related to the square numbers by

(2n + 1) 2 8Tn + 1 (9)

ity

= Tn −1 + 6Tn + Tn +1 (10)

The triangular numbers have the ordinary generating function

m

x
f(x) = (11)
(1 − x)3
= x + 3 x 2 + 6 x3 + 10 x 4 + 15 x5 + .... (12)
)A

and exponential generating function

1 2 x
g(x) = (1 + 2 x + x )e (13)
2
5 5
(c

= 1 + 3 x + 3 x 2 + x3 + x 4 + ... (14)
3 8

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 163

x x2 x3 x4
Notes

e
= 1 + 3 + 6 + 10 + 15 + ..... (15)
1! 2! 3! 4!
Every other triangular number is a hexagonal number, with

in
H n = T2 n −1 (16)
In addition, every pentagonal number is 1/3 of a triangular number, with

nl
1
Pn = T3n −1 (17)
3
The sum of consecutive triangular numbers is a square number, since

O
1 1
Tr Tr = r (r + 1) + (r − 1)r (18)
2 2
1

ity
= r[(r + 1) + (r − 1)] (19)
2
= r 2 (20)

Interesting identities involving triangular, square, and cubic numbers are

2 n −1

∑ (−1)
k =1
k +1
Tk = n 2 (21)
rs
ve
n

∑k
k =1
3
= Tn2 (22)

1 2
n (n + 1) 2 (23)
ni

=
4
n

∑ (2k − 1)
k =1
3
= T2 n2 −1 (24)
U

2
= n (2n 2 − 1) (25)
Triangular numbers also unexpectedly appear in integrals involving the absolute
value of the form
ity

1 1 n 2
∫∫ 0 0
x − y dxdy =
( n + 1)( n + 2 ) (26)

All evenperfect numbers are triangular with prime . Furthermore, ever

m

evenperfect number is of the form

1 + 9Tn =
P= T3n +1
(27)
where is a triangular number with (Eaton 1995, 1996). Therefore, the
)A

nested expression
9 (0 ... (9(9(9(9 Tn + 1) +1) + 1) + 1) ... + 1) + 1 (28)

generates triangular numbers for any . An integer is a triangular number if

is a square number .
(c

The numbers 1, 36, 1225, 41616, 1413721, 48024900, ... are square triangular
numbers, i.e., numbers which are simultaneously triangular and square (Pietenpol

Amity Directorate of Distance & Online Education

 164 Graph theory and Combinatorics

1962). The corresponding square roots are 1, 6, 35, 204, 1189, 6930, ..., and the
Notes

e
indices of the corresponding triangular numbers are , 8, 49, 288, 1681, ....

Numbers which are simultaneously triangular and tetrahedral satisfy the binomial

in
coefficient equation

Tn
= (=) ( =)
n +1
2
m+2
3 Tem (29)

nl
the only solutions of which are

Te3 = T4 = 10 (30)

O
Te8 = T15 = 120 (31)

Te20 = T55 = 1540 (32)

ity
Te34 = T119
= 7140 (33)

The following table gives triangular numbers having prime indices .

rs
class Sequence
with prime indices 3, 6, 15, 28, 66, 91, 153, 190, 276, 435, 496, ...

odd with prime indices 3, 15, 91, 153, 435, 703, 861, 1431, 1891, 2701, ...
ve
ven with prime indices 6, 28, 66, 190, 276, 496, 946, 1128, 1770, 2278, ...

The smallest of two integers for which is four times a triangular number
ni

is 5, as determine by the only Fibonacci numbers which are triangular are 1, 3, 21,
and 55, and the only Pell number which is triangular is 1 (McDaniel 1996). The beast
number 666 is triangular, since
U

T66 = T36 + 666 (34)

The positive divisors of are all of the form , those of are

all of the form , and those of are all of the form ; that is, they
ity

end in the decimal digit 1 or 9.

Fermat’s polygonal number theorem states that every positive integer is a sum of
at most three triangular numbers, four square numbers, five pentagonal numbers, and
nn-polygonal numbers.
m

This case is equivalent to the statement that every number of the form is
a sum of three oddsquares . Dirichlet derived the number of ways in which an integer
can be expressed as the sum of three triangular numbers (Duke 1997). The result
)A

is particularly simple for a primeof the form , in which case it is the number of
squares mod minus the number of nonsquares mod in the interval from 1
to

The only triangular numbers which are the product of three consecutive integers
are 6, 120, 210, 990, 185136, 258474216
(c

Amity Directorate of Distance & Online Education

 Graph theory and Combinatorics 165

Questions:
Notes

e
1. The rank of P is |X| − |P|, if X is __________.
a) finite.

in
b) infinite
c) negative

nl
d) positive
2. Any non-empty set X, P = {X} is a partition of X, called the _________trivial partition.

O
a) Non trivial
b) trivial
c) both a and b

ity
d) none
3. A partition is a __________of positive integers, and it is a partition of n of the sum of
the integer.

rs
a) even partition
b) odd partition
c) both a and b
ve
d) multiset
4. The number of partitions of n in which no part occurs more often than d times is the
same as the number of partitions in which no term is a multiple of ____________.
ni

a) d-3
b) d-2
c) d-4
U

d) d-1
5. The number of partitions in which the parts are all even and different is equal to the
_____________ difference of the number of partitions with odd and even parts.
ity

a) Symmetric
b) absolute
c) both a and b
m

d) none

Answers:
)A

6. a
7. b
8. d
(c

9. d
10. b

Amity Directorate of Distance & Online Education

 166 Graph theory and Combinatorics

Questions:
Notes

e
1. Explain even codes
2. Explain odd codes

in
3. Relation between even and odd codes
4. Explain Enumeration

nl
Sample Questions:
1. A coin is flipped 10 times. How many outcomes are possible?

O
Solution:
We have 2 multiplied by itself 10 times, or
210 =1,024 possible outcomes. The set itself is obviously too large to list.

ity
2. An automobile’s license plate consists of three letters (A-Z) followed by three
numerical digits (0-9). How many plates are possible if
a) there are no restrictions (repetition is allowed)?

rs
b) all letters and numbers must be different (repetition is not allowed)?
Solution:
In the first case (a) where there are no restrictions and repetition of the same
ve
letter or numeral is allowed, there are 263 =103 =17,576,000 possible license
plates. We used exponential shorthand for the calculation. In the second case
(b) where repetition is not allowed, we see that there are 26 x 25 x 24 x 10 x 9
x 8 = 11,232,000 possible plates.
ni

3. A true-false exam has 8 questions, followed by 10 multiple-choice questions,

each with four choices. How many ways can this test be filled?
U

Solution:
For the true-false portion, there are 28 = 256 ways to randomly answer the
questions.
ity

For the multiple choice portion, there are 410 = 1,048,576 ways to randomly
answer the questions. Taken as a whole, we multiply the two results together:
28. 410 = 268,435,456 possible ways to fill in the exam.
4. Radio and television station call letters are a sequence of three or four letters,
with the first letter being a W (for stations east of the Mississippi River) and K
m

(for stations west of the Mississippi River). How many such call letter sequences
are possible?
)A

Solution:
For three letters sequences, there are two choices for the first position (W or
K), and 26 choices each for the remaining two positions, for a total of 2 x 26 x
26 = 1,352 three letter sequences. For four letter sequences, the total number
possible is 2 x 26 x 26 x 26 = 35,152. Because these form two distinct classes
(c

of possible call-letter sequences, the total number of 3-letter or 4-letter call

sequences is the sum: We sum the two to obtain 36,504 possible three-letter or
four-letter call sequences.
Amity Directorate of Distance & Online Education
 Graph theory and Combinatorics 167

Text Books:
Notes

e
1) C.L. Liu, Elements of Discrete Mathematics, Tata McGraw Hill, 2nd Edition,
2000.

in
2) N. Deo, Graph Theory with Applications to Engineering and Computer Science,
PHI publication, 3rd edition, 2009

nl
3) Harikishan, ShivrajPundir and Sandeep Kumar, Discrete Mathematics, Pragati
Publication, 7th Edition, 2010.
4) Colmun, Busby and Ross, Discrete Mathematical Structure, PHI Publication,

O
6th Edition, 2009

ity
rs
ve
ni
U
ity
m
)A
(c

Amity Directorate of Distance & Online Education