Chapter 11

Graph Theory

“The origins of graph theory are humble, even frivolous.” (N.

Biggs, E. K. Lloyd, and R. J. Wilson)

Let us start with a formal definition of what is a graph.

Definition 72. A graph G = (V, E) is a structure consisting of a set V of
vertices (also called nodes), and a set E of edges, which are lines joining
vertices.
One way to denote an edge e is to explicit the 2 vertices that are connected
by this edge, say if the edge e links the vertex u to the vertex v, we write
e = {u, v}.
Definition 73. Two vertices u, v in a graph G are adjacent in G if {u, v} is
an edge of G. If e = {u, v} is an edge of G, then e is called incident with the
vertices u and v.
Graphs can be of several types.
Definition 74. A simple graph G is a graph that has no loop, that is no
edge {u, v} with u = v and no parallel edges between any pair of vertices.
Example 106. Consider the table of fictitious flights among different cities.
Suppose all you want to know is whether there is a direct flight between any
two cities, and you are not interested in the direction of the flight. Then you
can draw a graph whose vertices are the cities, and there is an edge between
city A and city B exactly when there is at least one flight going either from
city A to the city B, or from city B to city A.

249
250 CHAPTER 11. GRAPH THEORY

Definitions
A graph G = (V,E) is a structure consisting of a set V of
vertices (nodes) and a set E of edges (lines joining vertices).

e={u,v}
u v

•Two vertices u and v are adjacent in G if {u,v} is an edge of G.

• If e = {u,v}, the edge e is called incident with the vertices u and v.

Graphs are useful to represent data.

Simple Graphs
A simple graph is a graph that has no loop (=edge {u,v} with
u=v) and no parallel edges between any pair of vertices.
London
Beijing
From\To Hong Kong Singapore Beijing Taipei London Hong Kong
Hong Kong 4 Flights
Taipei
Singapore 2 Flights 3 Flights 1 Flight 1 Flight

Beijing 1 Flight 2 Flights

Taipei

London 1 Flight 1 Flight 1 Flight Singapore

Draw a graph to see whether there are direct

flights between any two cities (in either direction)
 251

Take for example Hong Kong: there will be one edge between Hong Kong
and Singapore (this is read in the first row), and and one edge from Hong
Kong to Beijing (this is read in the first column). This graph is simple, there
is no loop, and no parallel edge.
Definition 75. A multigraph G is a graph that has no loop and at least two
parallel edges between some pair of vertices.
Example 107. We continue with our fictitious example of flights. Suppose
now we want to know how many flights are there that operate between two
cities (we are still not interested in the direction). In this case, we will need
parallel edges to represent multiple flights. For example, let us consider Hong
Kong and Singapore: there are 4 flights from Hong Kong to Singapore, and
2 flights from Singapore to Hong Kong, thus a total of 6 edges between the
two vertices representing these cities.
Very often, a relation from a vertex A to B does not yield one from B to
A, in this case, edges become arrows.
Definition 76. A directed graph G, also called digraph for short, is a graph
where edges {u, v} are ordered ({u, v} and {v, u} are not the same), that is
edges have a direction. The graph is called undirected otherwise. Parallel
edges are allowed in directed multigraph. Loops are allowed for both directed
and directed multigraphs.
Example 108. On our example of ficticious flights, a directed graph corre-
sponds to put arrows for flights going in a given direction, for example, there
is one arrow from London to Beijing, but none from Beijing to London.
People attribute the origin of graph theory to the Königsberg bridge ques-
tion, solved by the mathematician Euler in 1736. The question was, given
the map of Königsberg, which contains 7 bridges, is it possible to find a walk
that goes through the 7 bridges without crossing a bridge twice? You may
look at the map and give it a try yourself, to convince yourself of the answer.
The answer turns out to be no, it is not possible. We will see why next. But
first, we will give a name to such walks, in honour of Euler:
Definition 77. A Euler path/trail is a walk on the edges of a graph which
uses each edge in the graph exactly once. A Euler circuit/cycle is a walk on
the edges of a graph which starts and ends at the same vertex, and uses each
edge in the graph exactly once.
252 CHAPTER 11. GRAPH THEORY

Multigraphs
A multigraph is a graph that has no loop and at least 2
parallel edges between some pair of vertices.
London
Beijing
From\To Hong Kong Singapore Beijing Taipei London Hong Kong
Hong Kong 4 Flights
Taipei
Singapore 2 Flights 3 Flights 1 Flight 1 Flight

Beijing 1 Flight 2 Flights

Taipei

London 1 Flight 1 Flight 1 Flight Singapore

Draw a graph with an edge for each flight that
operates between two cities (in either direction).
4/15

Directed (Multi)graphs
A directed graph is a graph where edges {u,v} are ordered,
that is, edges have a direction. Parallel edges are allowed
in directed multigraphs. Loops are allowed for both.
From\To Hong Kong Singapore Beijing Taipei London
London
Hong Kong 4 Flights
Beijing
Hong Kong
Singapore 2 Flights 3 Flights 1 Flight 1 Flight

Beijing 1 Flight 2 Flights Taip

ei
Taipei

London 1 Flight 1 Flight 1 Flight

Draw a graph to see whether there are direct Singapore

flights between any two cities (direction matters)
 253

Origin: The Bridges of Konigsberg

a
• Königsberg (now Kaliningrad, 1 2 3
Russia) has 7 bridges. b 7 d
6 4 5
c

a Leonhard Euler
introduced
Graphs in 1736
to solve the
b d Königsberg
• People tried (without success) Bridge problem
(no solution and
to find a way to walk all 7 bridges why).
without crossing a bridge twice. c

Euler Path and Circuit

A Euler path (Eulerian trail) is a walk on the edges of a graph
which uses each edge in the original graph exactly once.
The beginning and end of the walk may or not be the same vertex.

A Euler circuit (Eulerian cycle) is a walk on the edges of a

graph which starts and ends at the same vertex, and uses
each edge in the original graph exactly once.
254 CHAPTER 11. GRAPH THEORY

Euler Circuit
• Suppose the beginning and end are the same
a
node u.
• The graph must be connected.
• At every vertex v≠ u, we reach v along one
edge and go out along another, thus the
b d number of edges incident at v (called the
degree of v) is even.
• The node u is visited once the first time we
leave, and once the last time we arrive, and
c
possibly in between (back and forth), thus the
degree of u is even.
• Since the Königsberg Bridges graph has odd
degrees, no solution!

8/15

Euler Theorem
The degree of a vertex is the number of edges
incident with it.

Theorem. Consider a connected graph G.

1. If G contains an Euler path that starts and ends at the
same node, then all nodes of G have an even degree.
2. If G contains an Euler path, then exactly two nodes of G
have an odd degree.
• Suppose G as an Euler path, which starts at v and finishes at w.
• Add the edge {v,w}.
• Then by the first part of the theorem, all nodes have even
degree, but for v and w which have odd degrees.
 255

The Euler circuit/cycle is simply an Euler path/trail whose start and end
are the same vertex.

Definition 78. The degree of a node is the number of edges incident with
it.

With this definition of degree, we next answer the question of the bridges
of Kn̈igsberg. We start with the more constrained case of starting and finish-
ing at the same vertex. We assume the graph G is connected, which means
that there is always a way to walk from any vertex to any other (possibly
using several times the same vertex or the same edge).

Theorem 4. Consider a connected graph G. Then G contains an Euler

circuit/cycle, if and only if all nodes of G have an even degree.

Proof. We prove only that if G contains an Euler circuit, then all nodes have
an even degree. We start the walk at vertex u. Now for any vertex v which is
not u, we need to walk in v using some edge, and walk out of u using another
edge. We may come back to v, but for every come back, we still need one
edge to come in, and one to walk out. Therefore the degree of v must be
even! As for the starting point u, it is visited once the first time we leave,
and the last time we arrive (2 edges), and any possible back and forth counts
for 2 edges as well, which shows that indeed, it must be that all nodes have
an even degree!
Since the Königsberg bridge graph has odd degrees, it has no Euler cycle.
We next extend the argument to an Euler path.

Theorem 5. Consider a connected graph G. Then G contains an Euler path

if and only if exactly two nodes of G have an odd degree.

Proof. We only prove that if G has an Euler path, then exactly two nodes of
G have an odd degree. Suppose thus that G has an Euler path, which starts
at v and finishes at w. Create a new graph G0 , which is formed from G by
adding one edge between v and w. Now G0 has an Euler cycle, and so we
know by the previoust theorem that G0 has the property that all its vertices
have an even degree. Therefore the degrees of v and w in G are odd, while
all the others are even and we are done. The other direction is left as an
exercise (see Exercise 96).
256 CHAPTER 11. GRAPH THEORY

Examples
Note: Euler Theorem actually states an if and only if.

Example: Wolf, Goat, Cabbage

A classical puzzle that involves graphs:

From the left bank of the river, the ferryman is

to transport the wolf, the goat and the cabbage
to the right bank.
The boat is only big enough to transport one
Boat
River

object/animal at a time, other than himself.

The wolf cannot be left alone with the goat, and
the goat cannot be left alone with the cabbage.
Ferryman
How should the ferryman proceed?

Copyright belongs to the artists 11/15

 257

Here is a classical puzzle.

Example 109. A ferryman needs to transport a wolf, a goat and a cabbage

from one side of a river to the other, the boat is big enough for himself and
one object/animal at a time. How should the ferryman proceed, knowing
that the wolf cannot be left alone with the goat, and the goat cannot be
left alone with the cabbage? One way to solve this is to use a graph that
represents the different possible states of this sytem. The initial start up
point is a state where wgcf (w=wolf, g=goat, c=cabbage and f =ferryman)
are on the left side of the river, with nothing on the right side. The first
step, the ferryman has no choice, he takes the goat on the other side, which
leads to a second step, with wc on the leftt bank, and gf on the right bank.
The ferryman returns, he then can choose: he either takes the cabbage of the
wolf. This creates two branches in the graph. Each branch leads to a couple
of states, after which (see the graph itself) we reach a state where g is on the
left, and wf c is on the right, leading to the end of the puzzle. A solution is a
path in this graph that represents the different states of the system. In this
example, it is a fairly easy graph, and there are 2 paths, each of the same
length!

Here are two more types of graphs which are important, in that you are
very likely to encounter them.

Definition 79. A complete graph with n vertices is a simple graph that has
every vertex connected to every other distinct vertex.

Definition 80. A bipartite graph is a graph whose vertices can be parti-

tioned into 2 (disjoint) subset V and W such that each edge only connects a
v ∈ V and a w ∈ W .

We have seen the notion of degree of a vertex v in an undirected graph, it

is the number of edges incident with. We note that in this case, a loop at a
vertex contributes twice. For directed graphs, you may like to know that the
notion of degree is more precise, one distinguishes in-degree and out-degree,
which we will not discussed here.

Definition 81. The total degree of an undirected

P graph G = (V, E) is the
sum of the degrees of all the vertices of G: v∈V deg(v).
258 CHAPTER 11. GRAPH THEORY

Example: Wolf, Goat, Cabbage (IV)

Either he takes
1. The cabbage, bring back the goat, leave
• f = ferryman the goat and take the wolf across, return,
• g = goat and take the goat across.
• w = wolf 2. Or the wolf, bring back the goat, leave the
• c = cabbage goat and the cabbage across, return, and
take the goat across.

w/cfg wgf/c

wgcf/ wc/gf wcf/g g/wfc gf/wc /wcfg

c/wfg cfg/w

Complete & Bipartite

A complete graph with n vertices is a simple graph that
has every vertex connected to every other distinct vertex.

A bipartite graph is a graph whose

vertices can be partitioned into 2
(disjoint) subsets V and W s.t. each edge
only connects a v ∈ V and a w ∈ W.
 259

More on Node Degree

The degree deg(v) of a vertex v in an undirected graph is the number
of edges incident with it ( a loop at a vertex contributes twice) .
In-degree and Out-degree are distinguished for directed graphs.
v2

total degree = deg(v1) + deg(v2) + deg(v3) = 0 + 2 + 4 = 6.

v1
v3

The total degree deg(G) of an undirected graph G is the sum of the

degrees of all the vertices of G: deg(v)

v V

The Handshaking Theorem

Let G = (V,E) be an undirected graph with e
edges. Then
2e =  deg(v)
vV
(Note that this even applies if multiple edges and v1
loops are present.) v2

Proof.
Choose an e ∈ E(G) with endpoints v, w ∈ V.
e contributes 1 to deg(v) and 1 to deg(w). True
v3 v4
even when v = w . Thus each edge contributes 2
to the total degree. deg(v1) = deg(v2) = deg(v3) = deg(v4) = 4
2e =  deg(v) = 4  4 = 16 and e = 8
260 CHAPTER 11. GRAPH THEORY

Adjacency Matrix
v1 v2 v3
v1 1 0 0
v1 v2
A= v2 1 1 2
v3 1 0 0

v3

A graph can be represented by a What is the

matrix A = (aij) called adjacency adjacency matrix
matrix, with of a complete
aij = the number of arrows from graph?
vi to vj

Hamiltonian Circuit
• The Icosian game (1857):
Along the edges of a dodecahedron, find a
path such that every vertex is visited a single
time, and the ending point is the same as the
starting point.
• Hamilton sold it to a London game dealer in
1859 for 25 pounds.

A Hamiltonian path of a graph G is a walk such that every vertex is

visited exactly once.
A Hamiltonian circuit of a graph G is a closed walk such that every
vertex is visited exactly once (except the same start/end vertex).

Copyright: http://mathworld.wolfram.com/IcosianGame.html
 261

The Handshaking Theorem links the number of egdes in a graph to the

numbers of vertices.
Theorem 6. Let G = (V, E) be an undirected graph with |E| edges. Then
X
2|E| = deg(v).
v∈V

Proof. The proof follows the name of the theorem. The idea of handshaking
is that if two people shake hands, there must be...well...two persons involved.
In a graph G, this becomes, if there is an edge e between v and w, then e
contributes to 1 to the degree of v and to 1 to the degree of s. This is also
true when v = w. Therefore each edge contributes 2 to the total degree.
To represent a graph, a useful way to do so is to use a matrix.
Definition 82. The adjacency matrix of a graph G is a matrix A whose
coefficients are denoted aij , where aij , the coefficient in the ith row and jth
column, counts the number of arrows from vi to vj .
You may replace the term arrow by edge in this definition if you graph is
undirected.
Example 110. The adjacency of a complete graph with 4 vertices is
 
0 1 1 1
1 0 1 1
 .
1 1 0 1
1 1 1 0
The first row reads that v1 is connected to v2 , v3 , v4 , but not to itself (since
it is a simple graph by definition).
We saw earlier Euler paths as walks going through exactly every edge of
a graph. If you replace ”exactly every edge” by ”exactly every vertex”, this
becomes a Hamiltonian path!
Definition 83. A Hamiltonian path of a graph G is a walk such that every
vertex is visited exactly one. A Hamiltonian circuit of a graph G is a closed
walk such that every vertex is visited exactly one, except the same start/end
vertex.
Hamiltonian paths are harder to characterize than Euler paths. In par-
ticular, finding an algorithm that will identify Hamiltonian paths in a graph
is hard!
262 CHAPTER 11. GRAPH THEORY

Hamiltonian vs Eulerian

• Path (or trail) vs Circuit (or cycle): for circuits, the

walk starts and finishes at the same vertex, not
needed for path.
• Eulerian: walk through every edge exactly once.
• Hamiltonian: walk through every vertex exactly.

Examples
Euler circuit
Hamiltonian path
No Hamiltonian circuit

No Euler circuit, but Euler path

Hamiltonian path
No Hamiltonian circuit
 263

Graph Isomorphism

v1 v1
0 1 0 0 1
1 0 1 0 0
v3 v4 v5 v2
0 1 0 1 0
0 0 1 0 1
1 0 0 1 0

v5 v2 v4 v3

Finally, it is useful to pay attention that one draws a graph, it is just a

visualization...and several visualizations may be different, giving the impres-
sion that the graphs are different, while they are actually the same. If two
graphs differ by their labeling, but their adjacency structure is the same, we
say that these graphs are isomorphic. More formally:

Definition 84. A graph isomorphism between two graphs G and H is a

bijection f between the set of vertices of G and the set of vertices of H
such that any two vertices u, v in G are adjacent if and only if f (u), f (v) are
adjacent in H.
264 CHAPTER 11. GRAPH THEORY

Exercises for Chapter 11

Exercise 96. Prove that if a connected graph G has exactly two vertices
which have odd degree, then it contains an Euler path.

Exercise 97. Draw a complete graph with 5 vertices.

Exercise 98. Show that in every graph G, the number of vertices of odd
degree is even.

Exercise 99. Show that in very simple graph (with at least two vertices),
there must be two vertices that have the same degree.

Exercise 100. Decide whether the following graphs contain a Euler path/cycle.