Chapter

Ver.

1
Chapter Summary
 Graphs and Graph Models
 Graph Terminology and Special Types of Graphs
 Representing Graphs and Graph Isomorphism
 Connectivity
 Euler and Hamiltonian Graphs
 Shortest-Path Problems
 Planar Graphs
 Graph Coloring

2
Section

3
§10.1: Graphs and Graph Models
 Introduction to Graphs
 Graph Taxonomy
 Graph Models

4
What are Graphs?
 Discrete structure consisting of vertices and edges

vertex

edge

5
Simple Graphs
 each edge connects two different vertices
 No two edges connect the same pair of vertices
 No loop (edge that connects a vertex to itself)

If this edge is added,

the graph is no
longer simple

6
Example of a Simple Graph
 Let V be the set of states in the far-southeastern U.S.:
V={FL, GA, AL, MS, LA, SC, TN, NC}
 Let E={{u,v}|u adjoins v}
TN NC

each edge is a set MS SC

AL

GA
LA FL

7
Multigraphs
 There may be multiple edges connecting the same two
vertices.
 When m different edges connect the vertices u and v, we
say that {u, v} is an edge of
an edge of
multiplicity m. multiplicity 2

8
Pseudographs
 Like a multigraph, but may include loops.
 Loop: an edge that connects a vertex to itself
 e.g., nodes are campsites in a loop
national park, edges are hiking
trails through the woods.

9
Directed Graphs
 A directed graph (or digraph) G=(V,E) consists of a
nonempty set V of vertices (or nodes) and a set E of edges
(or arcs).
 Each edge is associated with an ordered pair of vertices.
 The edge associated with (u, v) is v
u
said to start at u and end at v.
an ordered pair

10
Simple Directed Graph
 A directed graph that has no loops and no multiple edges.

v
u
a b

c
not a simple
directed graph

11
Directed Multigraphs
 Like directed graphs, but there may be more than one arc
from a vertex to another.
 When there are m directed
edges from vertices u to v,
we say that is an
edge of multiplicity m.

12
Graph Models: Computer Networks
 build a graph model: use an appropriate type of graph to
capture the important features of the application.
 We illustrate this process using graph models of different
types of computer networks.
 In all these graph models
 vertices represent data centers
 edges represent communication links

13
Computer Networks: Simple Graph
 If we are only concerned whether two data centers are connected by a
communications link (and not how many links there may be), we
use a simple graph.
 all communications links work in both directions.

14
Computer Networks: Multigraph
 If we care about the number of links between data centers,
we use a multigraph.

15
Computer Networks: Pseudograph
 To model a computer network with diagnostic links at data
centers, we use a pseudograph, as loops are needed.

16
Computer Networks: Directed Multigraph
 To model a network with multiple one-way links, we use a directed
multigraph.

 Note that we could use a directed graph without multiple edges if we

only care whether there is at least one link from a data center to
another data center.
17
Graph Terminology: Summary

allows both
18
Other Applications of Graphs
 graph theory can be used in models of:
 Social networks
 Communications networks
 Information networks
 Software design
 Transportation networks
 Biological networks
 It’s a challenge to find a subject to which graph theory has not yet been
applied.

19
Graph Models: Social Networks
 In a social network
 vertices represent individuals or organizations
 edges represent (different kinds of) relationships between them.
 Useful graph models of social networks:
 friendship graphs
 collaboration graphs
 influence graphs

By Daniel' (User:Dannie-walker) - Own work, CC BY-SA 3.0,

https://commons.wikimedia.org/w/index.php?curid=8977072

20
Friendship Graph
 undirected graphs where two people are connected if they
are friends (in the real world, on Facebook, or in a particular
virtual world, and so on.)

21
Collaboration Graphs
 undirected graphs where two people are connected if they
collaborate in a specific way
Jason Statham

The Rock

Vin Diesel

22
Influence Graphs
 directed graphs where there is an edge from one person to
another if the first person can influence the second person

Your Mom Your Sister

Your Dad
You

23
Examples of Collaboration Graphs
 The Hollywood graph models the collaboration of actors in
films.
 represent actors by vertices
 connect two vertices if the actors they represent have appeared in
the same movie.
 We will study the Hollywood Graph in Sec. 10.4 when we
discuss Kevin Bacon numbers.

24
Academic Collaboration Graph
 models the collaboration of researchers who have jointly
written a paper in a particular subject.
 represent researchers in a particular academic discipline using
vertices.
 connect two vertices if the researchers they represent have
coauthored a paper.
 In Sec. 10.4, We will study the academic collaboration graph
for mathematicians when we discuss Erd s numbers

25
Applications to Information Networks
 Graphs can be used to model different types of networks that link
different types of information.
 In a web graph, web pages  vertices, and links  directed edges.
 A web graph models the WWW at a particular time.
 In a citation network:
 Research papers in a particular discipline are represented by vertices.
 When a paper cites a second paper as a reference, there is an edge from the
vertex representing this paper to the vertex representing the second paper.

26
Transportation Graphs
 Graph models are extensively used in the study of transportation
networks.
 Airline networks (directed multigraphs)
 airports  vertices
 each flight  a directed edge
 Road networks
 vertices represent intersections and edges represent roads.
 undirected edges represent two-way roads and directed edges represent one-
way roads.

27
Software Design Applications
 A top-down approach divides a system into modules, each
performing a specific task.
 a module dependency graph represents the dependency between
these modules.
 vertices represent software modules
 there is an edge from one module to another if the second module depends
on the first
 These dependencies need to be understood before coding can be
done.

28
Module Dependency Graph: An Example

The dependencies between the seven

modules in the design of a web browser
29
Precedence Graph
 a directed graph called precedence graph represents which
statements must have already been executed before we execute each
statement.
 Vertices represent statements
in a computer program
 There is a directed edge from
a vertex to a second vertex if
the second vertex cannot be
executed before the first

could run in parallel

30
Biological Applications: Niche Overlap Graph
 Niche overlap graphs model competition between species in an
ecosystem
 Vertices represent species
 an edge connects two vertices when
they represent species who compete
for food resources.

the niche overlap graph for a

forest ecosystem with nine species.

31
Biological Applications: Protein Interaction Network
 vertices represent proteins and vertices are connected by an edge if
the proteins they represent interact.
 Protein interaction graphs can be huge (more than vertices
and more than edges)
 often split into smaller graphs,
called modules, which represent
the interactions between proteins
involved in a particular function.
a module of the protein interaction graph of
proteins that degrade RNA in a human cell.
32
Section

33
§10.2: Graph Terminology and Special Types of
Graphs
 Basic Terminology
 Some Special Types of Graphs
 Bipartite Graphs
 Bipartite Graphs and Matchings
 Some Applications of Special Types of Graphs
 New Graphs from Old

34
Terminology: Adjacency
Let G be an undirected graph. Let e = {u,v} be an edge of G.
Then we say:
e v
 u, v are adjacent (neighbors; connected). u
 Edge e is incident with vertices u and v.
 Edge e connects u and v.
 Vertices u and v are endpoints of edge e.

35
Terminology: Neighborhood
 Let G=(V, E) be an undirected graph, vV a vertex, A is a
subset of V.
 The neighborhood of v, denoted by N(v), is the set of all
neighbors of v
 N(A) is the set of all vertices in G that are adjacent to at least
one vertex in A.

36
Degree of a Vertex
 Let G=(V, E) be an undirected graph, vV a vertex.
 The degree of v, deg(v), is its number of incident edges.
(Except that any self-loops are counted twice.)
 A vertex with degree is isolated.
 A vertex of degree is pendant.

37
Degrees and Neighborhoods: Example 1
What are the degrees and neighborhoods
of the vertices in the graph G?

Solution:
deg(a) = 2, deg(b) = deg(c) = deg(f) = 4,
deg(d) = 1, deg(e) = 3, deg(g) = 0.
N(a) = {b, f }, N(b) = {a, c, e, f }, N(c) = {b, d, e, f }, N(d) = {c}, N(e) = {b, c ,
f}, N(f) = {a, b, c, e}, N(g) =  .

38
Degrees and Neighborhoods: Example 2
What are the degrees and neighborhoods
of the vertices in the graph H?

deg(a) = 4, deg(b) = deg(e) = 6, deg(c) = 1, deg(d) = 5.

N(a) = {b, d, e}, N(b) = {a, b, c, d, e}, N(c) = {b},
N(d) = {a, b, e}, N(e) = {a, b ,d}.

39
Handshaking Theorem
 Let G=(V, E) be an undirected (simple, multi-, or pseudo-)
graph with m edges. Then

 deg(v)  2m
vV
Each edge contributes
1 degree at each end

degree+ degree+

40
Handshaking Theorem: Example 1
How many edges are there in a graph with 10 vertices of
degree six?
Solution:
Because the sum of the degrees of the vertices is 6  10 = 60,
the handshaking theorem tells us that 2m = 60.
So the number of edges m = 30.

41
Handshaking Theorem: Example 2
If a graph has 5 vertices, can each vertex have degree 3?
Solution:
This is not possible by the handshaking theorem, because
the sum of the degrees of the vertices 3  5 = 15 is odd.

42
Theorem 2
Theorem 2: An undirected graph G = (V, E) with m edges has an even
number of vertices of odd degree.
Proof: Let V1 be the vertices of even degree and V2 be the vertices of odd
degree in G. Then

must be even

even
must be even
Because all vertices in V2 are of
odd degrees, there must be an
even number of such vertices.

43
Directed Adjacency
 Let G be a directed (possibly multi-) graph, and let e = (u, v) be an
edge of G.
Then we say: e v
 u is adjacent to v, v is adjacent from u u
 e comes from u, e goes to v.
 e connects u to v, e goes from u to v
 the initial vertex of e is u
 the terminal vertex of e is v

44
Directed Degree
 Let G be a directed graph, v a vertex of G.
 The in-degree of v, deg(v), is the number of edges going to v.
 The out-degree of v, deg(v), is the number of edges coming from v.
 The degree of v, deg(v)deg(v)+deg(v), is the sum of v’s in-
degree and out-degree.

in-degree = 2
out-degree = 3

45
In-Degree/Out-Degree Example

deg−(a) = 2, deg−(b) = 2,
deg−(c) = 3, deg−(d) = 2,
deg−(e) = 3, deg−(f) = 0.

deg+(a) = 4, deg+(b) = 1,
deg+(c) = 2, deg+(d) = 2,
deg+ (e) = 3, deg+(f) = 0.

46
Directed Handshaking Theorem
 Let G=(V, E) be a directed (possibly multi-) graph. Then:

E   deg (v)   deg (v)

 

vV vV

Each edge contributes

one in-degree and one out-degree

47
Special Graph Structures
Special cases of undirected graph structures:
 Complete graphs Kn
 Cycles Cn
 Wheels Wn
 n-Cubes Qn
 Bipartite graphs
 Complete bipartite graphs Km,n

48
Complete Graphs
 For any nN, a complete graph on n vertices, Kn, is a simple graph
with n vertices that contains exactly one edge between each pair of
distinct vertices.

K1 K2 K3 K4 K5 K6

Note that Kn has edges.

49
Cycles
 A cycle Cn for n ≥ consists of n vertices v , v , , vn, and
edges {v , v }, {v , v } , , {vn- , vn}, {vn, v }.

C3 C4 C5 C6 C7 C8

How many edges are there in Cn?

50
Wheels
 For any n  , a wheel Wn, is obtained by
 adding an additional vertex to a cycle Cn and
 connecting this new vertex to each of the n vertices in Cn by new
edges.

W3 W4 W5 W6 W7 W8

How many edges are there in Wn?

51
n-cubes (hypercubes)
 An n-dimensional hypercube, or n-cube, Qn
 vertices represent all bit strings of length n ( vertices in total)
 there is an edge between two vertices that differ in exactly one bit position

Q4
Number of edges: Exercise to try!

52
Bipartite Graphs
A simple graph G=(V, E) is bipartite if we can find a way to partition
V into two disjoint subsets V1 and V2 such that every edge connects a
vertex in V1 and a vertex in V2.
V1 V2
In other words, there are • •
no edges which connect • •
• •
two vertices in V1 or in V2. • •
•

53
It’s Not Always Easy to Identify a Bipartite
Graph
This is a bipartite
graph

54
Bipartite Graph and Coloring
 an equivalent definition of a bipartite graph: we can find a way to assign
either red or blue to each vertex so that no two adjacent vertices are
assigned the same color.

H is not bipartite since if we color

a red, then the adjacent vertices f
and b must both be blue.

G is bipartite

55
Bipartite Graph Example 1
Example: Show that C6 is bipartite.
Solution: We can partition the vertex set into V1 = {v1, v3, v5} and V2
= {v2, v4, v6} so that every edge of C6 connects a vertex in V1 and
another vertex in V2.
v1
v6 v2
v5 v3
v4

C6
56
Bipartite Graph Example 2
Example: Show that C3 is not bipartite.
Proof: If we divide the vertex set of C3 into two
nonempty sets, one of the two must contain two
vertices.
C3
But in C3 every vertex is connected to every other vertex.
Therefore, the two vertices in the same partition are
connected.
Hence, C3 is not bipartite.

57
Complete Bipartite Graphs
A complete bipartite graph Km,n is a graph that has its vertex set
partitioned into two subsets V1 of size m and V2 of size n such that
there is an edge from every vertex in V1 to every vertex in V2.

K2,3 K3,3

K3,5 K2,6
58
Server Pod Used by Facebook
 A spine/leaf topology

K4,48

59
Special Graphs in Computer Network
Architecture
 Some local area networks use a star topology, which is
a complete bipartite graph K1,n .
 All devices are connected to a central control device.
 Other local networks are based on a ring topology,
where each device is connected to exactly two others
using Cn.
 Messages may be sent around the ring.
 Others use a Wn–based topology, combining the
features of a star topology and a ring topology.
60
Special Graph in Parallel Processing System
 in a parallel processing system, processors need to be
interconnected as one processor may need the output
generated by another.
 The n-dimensional hypercube, or n-cube, Qn,
is a common way to connect processors in
parallel, e.g., Intel Hypercube.
 Another common method is the mesh
network, illustrated here for processors.

61
Subgraphs
 A subgraph of a graph G=(V,E) is a graph H=(W,F) where W
 V and F  E.

G H
 A subgraph H of G is a proper subgraph of G if H ≠ G

62
Induced Subgraph
 Let G = (V, E) be a simple graph.
 The subgraph induced by a subset of vertices W  V is a graph (W, F)
such that {u, v} F if and only if {u, v} E and {u, v}  W.
 Here we show K5 and the
subgraph induced by
W = {a, b, c, e}.

63
Subgraph vs. Induced Subgraph
 For all u, vW, if {u, v} E then {u, v} must be in the induced
subgraph (not necessarily in a subgraph)
a a

b b
and
but

G = (V, E) A subgraph (W, F) but

not an induced graph

64
Bipartite Graphs and Matchings
 Bipartite graphs are used to model applications that involve
matching the elements of one set to elements in another.
 Job assignments - vertices represent the jobs and the employees,
edges link employees with those jobs they have been trained to do.
 A common goal is to
match jobs to employees
so that the most jobs are
done. ( matching)

65
Matching Problem is Not Easy
Allen Bob Christ David

only two jobs

are assigned
Job Job Job Job

Allen Bob Christ David

Job Job Job Job

66
Matching Examples
 Marriage - vertices represent the men and the women and
edges link a man and a woman if they are an acceptable
spouse.
Allen Bob Christ David
 We may wish to find
the largest number of
possible marriages.

Taylor Swift Lily Collins Katy Perry Jennifer

Lawrence

67
Graph Unions
 The union of two simple graphs and
is the simple graph .

b c a b c a b c
a

d e d f d e f

68
Section

69
§10.3: Representing Graphs and Graph
Isomorphism
 Adjacency Lists
 Adjacency Matrices Graph Representation

 Incidence Matrices
 Isomorphism of Graphs

70
Adjacency Lists
 A table with row per vertex, listing its adjacent vertices.
b Adjacent
a
Vertex Vertices
a b, c
c d b a, c, e, f
e
c a, b, f
f d
e b
For graphs with no f c, b
multiple edges
71
Directed Adjacency Lists
 row per node, listing the terminal nodes of each edge
incident from that node.
Adjacent
a b Vertex Vertices
a
d b a, e
c e c a, b, f
f d
e
For graphs with no f b
multiple edges
72
Adjacency Matrices
 Matrix A=[aij], where aij is if {vi, vj} is an edge of G,
otherwise. a b c d
a b a 0 1 1 1
b 1 0 1 0

c c 1 1 0 0
d  
d 1 0 0 0
can represent simple graph (here) and directed graph (later)

73
Adjacency Matrix Example

The ordering of
vertices is a, b, c, d.

 The adjacency matrix of a simple graph is

symmetric, i.e., aij = aji

74
Adjacency List vs. Adjacency Matrix
 A graph can be sparse (has few edges relatively to the total
number of possible edges) or dense (on the other hand).
 When a graph is sparse, it is much more efficient to
represent the graph using an adjacency list than an
adjacency matrix.
 For a dense graph, an adjacency matrix is preferable.

75
For Loops and Multiple Edges
 A loop at the vertex vi is represented by a at the (i, i)th position of
the matrix.
 When multiple edges are allowed, the (i, j)th entry equals the
number of edges connecting vi and vj.
 the (i, i)th entry > if multiple loops are present

a pseudograph

76
Directed Adjacency Matrices
 For a directed graph G = (V, E), the matrix has a in its (i, j)th
position if there is an edge from vi to vj.
 When there is an edge from vi to vj, there may not be an edge from vj
to vi.
 Thus the adjacency matrix for a directed graph does not have to be
symmetric.
 To represent directed multigraphs, the value of aij is the number of
edges connecting vi to vj.

77
Incidence Matrices
 An n  m matrix M=[mij], where n = |V| and m = |E|
 mij = 1 if edge ej is incident with vi

e1 e1 e2 e3 e4 edges
a b
a 1 1 1 0
e2 1
e4 e3 b
 0 0 1 
c
d c 0 1 0 1
d
 
0 0 1 0
vertices
78
Incidence Matrix for Simple Graph
e1 e2 … e6
v1
v2
:
v5
+)

79
Incidence Matrix for Pseudograph
e1 e2 … e8
v1
v2
:
v5
+)

80
Isomorphism of Graphs
 Graphs with the same structure

a v1
b v2
have the same
and
structure?
c v4
d v3
a d

c
b
81
Graph Isomorphism: Formal Definition
 Simple graphs G =(V , E ) and G =(V , E ) are isomorphic iff
 a bijection f:V V such that  a, bV , a and b are
adjacent in G iff f(a) and f(b) are adjacent in G .
 f is called an isomorphism.
 Two simple graphs that are not isomorphic are called
nonisomorphic.

82
Graph Isomorphism: Example
u1
 Show that the graphs G =(V, E) u2
and H = (W, F) are isomorphic.
 The function f with f(u1) = v1,
u3 u4
f(u2) = v4, f(u3) = v3, and f(u4) = v2
is a one-to-one correspondence G
between V and W.
v1
 Adjacent vertices in G are u1 and u2, v2
v1 v4
u1 and u3, u2 and u4, and u3 and u4.
v3
v1 v3 v4 v2 v3 v2 H v4
83
Determining Graph Isomorphism
 It is difficult to determine whether two simple graphs are isomorphic
using brute force (n! possible one-to-one correspondences).
 graph invariant is a property that should be preserved by
isomorphism
 We try to find a graph invariant that only one of the two graphs has.
 many different useful graph invariants can be used to distinguish
nonisomorphic graphs (number of vertices, number of edges, and
degree sequence)

84
Some graph Invariants under Isomorphism
Necessary but not sufficient conditions for G =(V , E ) to be
isomorphic to G =(V , E ):
|V |=|V |, |E |=|E |.
The number of vertices with degree n is the same in both
graphs for each n.
For every proper subgraph g of one graph, there is a
proper subgraph of the other graph that is isomorphic to g.

85
Isomorphism Example
 If isomorphic, label the nd graph to show the isomorphism.
Otherwise, identify the difference.
d * Same # of vertices
b
b a
a * Same # of edges
d c
e
f * Same # of vertices of
c
e every degree
f

86
Are These Isomorphic?
 If isomorphic, label the nd graph to show the isomorphism.
Otherwise, identify the difference.
a * Same # of vertices
b
* Same # of edges
d
* Different # of vertices of
c e degree 2! (1 vs. 3)

87
Determining Graph Isomorphism: Example 1
deg =
 Both graphs have
vertices and edges.
 They also both have
vertices of degree and
deg =
of degree .
 deg(a) = in G, a must correspond to t, u, x, or y in H
 each of these vertices is adjacent to another vertex of degree two in H,
which is not true for a in G.
Nonisomorphic!
88
Determining Graph Isomorphism: Example 2
 Both graphs have
vertices and edges.
 They also both have
vertices of degree and
of degree .
 The subgraphs of G and H consisting of all the vertices of degree two
and the edges connecting them are isomorphic.
 So, it is reasonable to try to find an isomorphism f.

89
 The function f with f(u1) = v6, f(u2) = v3, f(u3) = v4, and f(u4) = v5 ,
f(u5) = v1, and f(u6) = v2 is a one-to-one correspondence
between G and H.
 Can use the adjacency matrices of G and H to verify that f is an
isomorphism.
90
Algorithms for Graph Isomorphism
 The best known algorithms have exponential worst-case time
complexity (in the number of vertices).
 Some algorithms have linear average-case time complexity.
 a public domain program, NAUTY, determines in < sec. whether
two graphs with  vertices are isomoprhic.
 Graph isomorphism is a problem of special interest because it is one
of a few NP problems not known to be either tractable or NP-
complete.

91
Applications of Graph Isomorphism
 chemists use molecular graphs
to model chemical compounds.
 Vertices represent atoms and
edges represent chemical bonds.
 When a new compound is synthesized, a database of molecular
graphs is checked to determine whether the graph representing the
new compound is isomorphic to the graph of a compound that this
already known.

92
Electronic Circuits
 Electronic circuits are modeled as graphs in which
 the vertices represent components
 the edges represent connections between components
 Graph isomorphism is the basis for
 verifying whether a particular circuit layout corresponds to the
design’s original schematics.
 determining whether a chip from one vendor includes the
intellectual property of another vendor.

93
Section

94
§10.4: Connectivity
 Paths
 Connectedness in Undirected Graphs
 Vertex Connectivity and Edge Connectivity
 Connectedness in Directed Graphs
 Paths and Isomorphism
 Counting Paths between Vertices

95
Path
 In an undirected graph, a path of length n from u to v is a
sequence of n adjacent edges going from vertex u to vertex v.
 A path is a circuit if u = v.
 A path is simple if it contains no edge more than once.

96
Path Definition
 A path of length n from u to v
e1 e2 en
x0 = u x1 x2 xn-1 xn = v
 When the graph is simple, we denote this path by its vertex sequence
x0, x1, … , xn
 The path or circuit is said to pass through the vertices x1, x2, …, xn-1
and traverse the edges e1, …, en. excluding u and v

97
Circuits and Simple Paths
 This graph is simple

a b b, f, c, a, b is a circuit
b, c, a, b, e is a simple path
c d
e b, c, a, b, c, f is not a
f simple path

98
More Examples on Paths
 a, d, c, f, e is a simple path of
length .
 d, e, c, a is not a path because
e is not connected to c.
 b, c, f, e, b is a circuit of
length .
 a, b, e, d, a, b is a path of length , but it is not a simple
path.
99
Paths in Directed Graphs
 Same as in undirected graphs, but the path must go in the
direction of the arrows.
a b
c, f, b, e is a simple path
c d f, b, c, a is not a path
e
f

100
Degrees of Separation
In an acquaintanceship graph, two
people are neighbors if they know
one another.
There is a path between two
people if there is a chain of people Some have speculated that
almost every pair of people in
linking these people the world are linked by a small
Here is a chain of six people chain of no more than six (Six
Degrees of Separation).
linking Kamini and Ching.
101
Erdős numbers
Paul Erdős

 Consider a graph where vertices are scholars

and an edge connects two persons if they have
coauthored a paper.
 the Erdős number of a person m is the
length of the shortest path between m
and Paul Erdős.
 To learn more about Erdős numbers, visit
http://www.ams.org/mathscinet/collaborationDistance.html
102
Here is my result
In fact, my number
should be

103
Bacon Numbers Kevin Bacon

 Consider a graph where vertices are actors and

an edge connects two actors if they have acted
in the same movie.
 The Bacon number of an actor c is the length
of the shortest path connecting c and
Kevin Bacon.
 The oracle of Bacon web site
http://oracleofbacon.org/how.php
provides a tool for finding Bacon numbers.

104
Connectedness in Undirected Graphs
 An undirected graph is connected iff there is
a path between every pair of distinct
vertices in the graph.

 Theorem: There is a simple path between

any pair of vertices in a connected
undirected graph.

105
Connectedness: Examples

not connected
connected

106
Connected Components
 A connected component: a connected subgraph that is not a proper
subgraph of another connected subgraph. (a maximal connected
subgraph)
 A graph G that is not connected has two or more connected
components
 connected components
are disjoint
 G is the union of these
connected components.
107
How Connected is a Graph?
 Sometimes a graph is connected, but a
removal of one vertex or an edge from
this graph will make it not connected.
 We want to know how robust a graph is
in terms of connectedness.

108
Cut Vertex and Cut Edge
 If removing a vertex (and all incident edges) from a
connected graph produces a subgraph that is not connected,
this vertex is called a cut vertex.

 If removing an edge from a connected graph produces a

subgraph that is not connected, this edge is called a cut edge.

109
Cut Vertex/Edge Example
 Cut vertices:
a d f g
 b, c, and e
 Cut edges:
b c e h
 (a, b) and (c, e)

110
Vertex Connectivity
 No all graphs have cut vertices K5 K4
 Removing any vertex from Kn  Kn-1
 Nonseparable graphs
 May need to remove more vertices/edges
 Using the min. number of needed vertices/ edges as a
measure of graph connectivity

111
 Set of vertices

Vertex Cut and Vertex Connectivity

 A subset V’ of the vertex set V of G = (V, E) is a vertex cut
(separating set) if G − V’ is disconnected.
 vertex connectivity κ(G): the minimum number of vertices
in a vertex cut (κ: kappa)

{b, c, e} is a vertex cut

112
Vertex Connectivity Examples
a d f g

G1 κ(G1) =
b c e h

b c d

G3 {b, g} is a vertex cut

a
g f e κ(G3) =
Removing a single vertex
cannot make G3 disconnected

113
κ(G) for Complete Graphs
 A complete graph should also have a vertex cut
 However, removing any subset from it  still a complete graph
 Redefine κ(G): the minimum number of vertices that must be
removed from G to
 either disconnect G
 or produce a graph with a single vertex
 Therefore, κ(Kn) = n −
 if G has n vertices, ≤ κ(G) ≤ n − (in general)

114
k-Connected Graph
 a graph G is k-connected (or k-vertex-connected), if κ(G) ≥ k
 That means we have to remove at least k vertices to disconnect G (or
reduce it to a graph with vertex)
 if G is a k-connected graph, then G is a j-connected graph for all j
with ≤ j ≤ k.
b c d
-connected
a also -connected
g f e

115
 Set of edges

Edge Cut and Edge Connectivity

 A set of edges E’ is called an edge cut of G if the subgraph G
− E’ is disconnected.
 edge connectivity λ(G): the minimum number of edges in an
edge cut of G.
 λ(G) = if  G is not connected or  G is a graph consisting
of a single vertex

116
Edge Connectivity Examples
a d f g

G1 (G1) =
b c e h

b c d

G3 {(b, c), (f, g)} is an edge cut

a
g f e (G3) =
Removing a single edge
cannot disconnect G3

117
Inequalities of Vertex/Edge Connectivity
 Removing all the neighbors of a vertex will disconnect this vertex
from all other vertices
 So κ(G)  minvV deg(v)
 Removing all the edges incident with a vertex will disconnect this
vertex from all other vertices
 So (G)  minvV deg(v) a d
 Removing one vertex may
remove more than one edge c
b e
 κ(G)  (G)
118
Directed Connectedness
 A directed graph is strongly connected iff there is a directed
path from a to b for any two vertices a and b.
 It is weakly connected iff the underlying undirected graph
(i.e., with edge directions removed) is connected.
 Note strongly implies weakly but not vice-versa.

119
Strongly Connectedness Examples
b
a b Strongly connected?
No! There is no path
c e from e to any others
f
Weakly connected?
Yes! The underlying
undirected graph is
connected
120
Directed Connectedness: Examples

strongly connected not strongly connected

weakly connected weakly connected

121
Strongly Connected Component
 the maximal strongly connected subgraphs are called the
strongly connected components or strong components.
 these subgraphs are
strongly connected but
not contained in larger
strongly connected
subgraphs
the meaning of ‘maximal’

122
Paths & Isomorphism
 Note that  connectedness and  the existence of a circuit
or simple circuit of length k are also graph invariants with
respect to isomorphism.
* Same number of vertices
u1 v1
* Same number of edges
u6 u2 v6 v2
* Same number of vertices
of degrees 2 and 3.
u5 u3 v5 v3 * H has simple circuit of length 3,
G u4 H v4 but G has not

123
Counting Paths with Adjacency Matrices
 Let A be the adjacency matrix of graph G.
 The number of paths of length k from vi to vj is equal to (Ak)i,j. (The
notation (M)i,j denotes mi,j where [mi,j] = M.)

a a 0 1 1 0 8 0 0 8
b 1 1 
d 0
0 0 8 8 0
A A4  
c 1 0 0 1 0 8 8 0
b    
d 0 1 1 0
c  8 0 0 8
The number of paths of length 4 from d to a

124
Understanding Why
a 01 1 0 0 1 1 0 2 0 0 2

b 1 0 0 1 1 0 0 1 0 2 2 0
A  A = A 
2

c 1 0 0 1 1 0 0 1 0 2 2 0
     
d 0 1 1 0 0 1 1 0 2 0 0 2
(da) 0  (aa) 0 = (da) 0 a
(db) 1  (ba) 1 = (da) 1 d
(dc) 1  (ca) 1 = (da) 1
+ (dd) 0  (da) 0 = (da) 0
b
(d)  (a) 2 c
125
Understanding Why (cont)
2 0 0 2 0 1 1 0 0 4 4 0
0 2 2 0  1 0 0 1 4 0 0 4
A2    A = A 
3
0 2 2 0 1 0 0 1 4 0 0 4
     
2 0 0 2 0 1 1 0 0 4 4 0
(d a) 2  (aa) 0 = (da) 0
a
(d b) 0  (ba) 1 = (da) 0 d
(d c) 0  (ca) 1 = (da) 0
+ (d d) 2  (da) 0 = (da) 0 b
(d)  (a) 0 c
of length 2 126
Counting Paths between Vertices: Example
 How many paths of length four are there from a to d in the graph G?
a
d
G A=
b
c adjacency matrix of G

a, b, a, b, d a, b, a, c, d
A4 = a, b, d, b, d a, b, d, c, d
a, c, a, b, d a, c, a, c, d
a, c, d, b, d a, c, d, c, d
127
Section

128
§10.5: Euler and Hamiltonian Graphs
 Euler Paths and Circuits
 Hamilton Paths and Circuits
 Applications of Hamilton Circuits

129
Euler Paths and Circuits
 The town of K nigsberg, Prussia (now Kalingrad, Russia) was
divided into four sections by the branches of the Pregel river.
 In the th century seven bridges
connected these regions.
 People wondered whether
it was possible to follow a
path that crosses each bridge
exactly once and returns to
the starting point. The 7 Bridges of Kӧnigsberg

130
And the answer is … Leonard Euler
(1707-1783)

 The Swiss mathematician Leonard Euler proved that no

such path exists.
 This result is often considered the first theorem ever proved
in graph theory.
C C

Multigraph
Model of the
A A D
D Bridges of
Kӧnigsberg
B B

131
Euler Circuit in a Graph
 A path is an Euler circuit in a graph G if
 It is a simple path (containing no edge more than once)
 It contains every edge of G.
 It is a circuit (beginning with and ending at the same vertex.)

a d
Is a, b, d, c, a an Euler circuit? Why?
Is a, b, d, c, a, d an Euler circuit?
Is there an Euler circuit? b
c
132
 A circuit is also a path.
Euler Path in a Graph An Euler circuit is not an Euler path.

 A path in G is an Euler path in a graph G if

 It is a simple path (containing no edge more than once)
 It contains every edge of G.
 It is not a circuit (beginning with and ending at different
vertices.)
a
d
Is a, b, d, c, a, d an Euler path?
b
c
133
Euler Paths and Circuits: Examples
Which has a Euler circuit? Of those that do not, which has an Euler
path?

an Euler circuit no Euler no Euler circuit

a, e, c, d, e, b, a circuit an Euler path a,
no Euler path no Euler path c, d, e, b, d, a, b

134
Necessary Conditions for Euler Circuit / Euler
Path
Every vertex in G
G has an Euler circuit 
is of even degree

Exactly two vertices in

G has an Euler path 
G are of odd degrees

135
Necessary Condition for Euler Circuit
 Suppose that an Euler circuit begins with a vertex a

a b c x a
deg(a)+ deg(b)+ deg(x)+
deg(b)+ deg(c)+ deg(x)+ deg(a)+
deg(c)+
deg(a) must Each time the circuit passes through a vertex it
be even contributes two (in & out) to the vertex’s degree

the degree of every vertex must be even

136
Necessary Condition for Euler Path
 Suppose that an Euler path begins with a vertex a

a b c x y
deg(a)+ deg(y)+
Each time the path passes through a vertex
it contributes two to the vertex’s degree
deg(a) must deg(y) must
be odd be odd

 a graph with an Euler path has exactly two vertices of odd degree.

137
Sufficient Conditions for Euler Circuit / Euler Path
Every vertex in G is
 G has an Euler circuit
of even degree
We prove it by showing
how to construct one.

Exactly two vertices in

G are of odd degrees  G has an Euler path

138
Sufficient Condition for Euler Circuit (1/5)
 Suppose that G is a connected multigraph with ≥ vertices, all of even
degree.
 Let x0 = a be a vertex of even degree.
 Choose an edge {x0 x1 a
{x0 x1 {x1 x2 {xn-1 xn a b

c d
We reach a vertex for which all f
incident edges are included
e
139
Sufficient Condition for Euler Circuit (2/5)


 a (i.e., a circuit)

 a

 The path may or may not use all the edges

 If all of the edges have been used, an Euler circuit has been constructed.

140
Sufficient Condition for Euler Circuit (3/5)
 If not all of the edges have been used, construct subgraph H from G by
deleting the edges already used and vertices not incident with any
remaining edges
a b
c d
G H
c d
f
e
e

 Because G is connected, H must have at least one vertex (c in this example)

in common with the circuit that has been deleted.

141
Sufficient Condition for Euler Circuit (4/5)
 Every vertex in H must have even degree because
 all the vertices in G have even degree and
 for each vertex in H, either or pairs of edges incident with this vertex have
been deleted. c d
 Beginning with the shared vertex
construct a path (circuit) ending in
the same vertex (as was done before).
e
 Then splice this new circuit into
the original circuit.
In the example, we end up with the circuit a, f, c, d, e, c, b, a.
142
Sufficient Condition for Euler Circuit (5/5)
 Continue this process until all edges have been used.
 This produces an Euler circuit since every edge is included
and no edge is included more than once.
 Similar reasoning can be used to show that a graph with
exactly two vertices of odd degree must have an Euler path
connecting these two vertices of odd degree

143
Algorithm for Constructing an Euler Circuit
In our proof we developed an algorithm
procedure Euler(G: connected multigraph with all vertices of even degree)
circuit := a circuit in G beginning at an arbitrarily chosen vertex with edges successively added to
form a path that returns to this vertex.
H := G with the edges of this circuit removed
while H has edges
subciruit := a circuit in H beginning at a vertex in H that also is an endpoint of an edge in circuit.
H := H with edges of subciruit and all isolated vertices removed
circuit := circuit with subcircuit inserted at the appropriate vertex.
return circuit{circuit is an Euler circuit}

144
Theorem for Euler Circuit
 A connected multigraph has an Euler circuit iff each vertex
has even degree.
C
Is there an Euler circuit?

A D

145
Theorem of Euler Path
 A connected multigraph has an Euler path (but not an Euler circuit)
iff it has exactly vertices of odd degree.
C C

A D A D

B B
Is there an How about this? And this?
Euler path?

146
How to Find an Euler Path?
 We have exactly two vertices of odd degree
 One of them must be the starting vertex
 The other must be the ending vertex

147
Euler Circuits and Paths: Examples
a
a b a g f e b
f
c c d g c
d b
e
an Euler path: an Euler path: b, a, g, d
d, a, b, c, d, b. f, e, d, c, g, b, c, f, d No Euler path

148
Applications of Euler Paths and Circuits
 Finding a path or circuit that traverses each
 street in a neighborhood,
 road in a transportation network,
 connection in a utility grid,
 link in a communications network.
 Other applications are found in the
 layout of circuits,
 network multicasting,
 molecular biology, where Euler paths are used in the sequencing of DNA.

149
Hamilton Circuit in a Graph
 A path is a Hamilton circuit in a graph G if
 It is a circuit (beginning with and ending at the same vertex.)
 It traverses each vertex in G exactly once.

a
a, b, d, c is not a Hamilton circuit. Why?
d

b
c

150
Hamilton Path in a Graph
 A path is a Hamilton path in a graph G if
 It is not a circuit (beginning with and ending at different
vertices.)
 It traverses each vertex in G exactly once.

Is a, b, d, c a Hamilton path? a d

b
c
151
Hamilton Circuit: An Example
need not traverse
every edge
William Rowan Hamilton
(1805- 1865)

152
Hamilton Paths and Circuits: Examples
a b a b a b g

e c d c
d d c e f

a, b, c, d, e, a

a, b, c, d

153
Necessary Conditions for Hamilton Circuits
 Do we have the necessary and sufficient conditions for the
existence of a Hamilton circuit?
 No
 There are some useful sufficient conditions.
 Dirac’s Theorem
 Ore’s Theorem

154
Dirac’s Theorem
Gabriel Andrew Dirac
 If (but not only if ) G = (V, E) is (1925-1984)

 connected, yes a
d
 simple, yes
 has n vertices, and yes b
 vV: deg(v)  n/ , yes c
 then G has a Hamilton circuit. 2
How about this?
2

155
Ore’s Theorem
Øysten Ore
(1899-1968)
 If (but not only if ) G = (V, E) is
3 4
 connected, yes
 simple, yes 3

 has n vertices, and yes 3

 (u,v) E deg(u)+deg(v)  n, yes 3

2
3
 then G has a Hamilton circuit.
How about this? 2
3

2
156
Applications of Hamilton Paths and Circuits
 Applications that ask for a path or a circuit that visits
 each intersection of a city,
 each place pipelines intersect in a utility grid, or
 each node in a communications network exactly once
 traveling salesperson problem (TSP) asks for the shortest route a
traveling salesperson should take to visit a set of cities.
 finding a Hamilton circuit such that the total sum of the weights
of its edges is as small as possible.

157
Section

158
§10.6: Shortest-Path Problems
 Introduction
 A Shortest-Path Algorithm
 The Traveling Salesperson Problem

159
Weighted Graph
 a number (weight) is assigned to each edge

160
Shortest-Path Problem
 Let the length of a path be the sum of the weights of the edges of this
path
 What is a shortest path (a path of least length) between two given
vertices?

161
Let’s try an example
 Find the length of the shortest path between a and z.
b 3 c starting b 3 c
4 2 from a 4 2
a z a z
3 3
2 1 a visited 2 1
3 e L(a) = 0 3 e
d d
L(v): the length of the L(b)= choose d
b visited shortest path from a to v L(d)=
L(b) = 4
b 3 c b 3 c
4 2 4 2
a z L(b)= , L(e)= z
3 a 3
2 choose b
1 d visited 2 1
d 3 e L(d) = 2 d 3 e
162
The example continues …
b 3 c L(c)= , La,b,e(e)
4 2
a z = , La,d,e(e)= b 3 c
3 4 2
2 choose a,d,e a z
1 3
d 3 e e visited 2 1
a, b, d visited, L(d) = 2, L(b) = 4 L(e) = 5 d 3 e
L(c)= choose z
L(z)=
b 3 c
4 2
a z
3
z visited 2 1
L(z) = 6 d 3 e
163
Dijkstra’s Algorithm: Notation
 distinguished set: set of vertices that have been visited
 Sk: the distinguished set after k iterations
 S0 = . Sk = Sk-1  {v}, where v  Sk-1 s.t. Lk(v) is min.
 Lk(v) is the length of a shortest path from a to v containing only vertices in
S k.
Lk(v) = min {Lk− 1(v),
Sk-1 v not in Lk− 1(u) + w(u, v) ),
u Sk-1 Lk− 1(w) + w(w, v)}
x
w Lk(x) = min {Lk− 1(x),
Lk− 1(u) + w(u, x) ),
Lk− 1(w) + w(w, x)}
Sk = Sk-1  {v} if Lk(v) < Lk(x)
164
Dijkstra’s Algorithm: from a to z
1. L(a) :=
2. L(x) := w(a, x) for all x  a // will be if x is not adjacent to a
3. S :=
4. while z  S // destination not reached
5. u := a vertex not in S with L(u) minimal
6. S := S ∪ {u}
7. for all vertices v not in S
8. if L(u) + w(u, v) < L(v) then L(v) := L(u) + w(u, v)
9. return L(z)
165
The Traveling Salesperson Problem
 A traveling salesperson wants to visit each of n cities exactly once
and return to his starting point.
 In which order should
he visit these cities to
travel the minimum
total distance?
 Find a Hamilton
circuit with minimal
total weight.

166
All Possible Hamilton Circuits

The best

167
Section

168
§10.7: Planner Graph
 Introduction
 Euler’s Formula
 Kuratowski’s Theorem

169
The Problem
 Is it possible to join these houses and utilities so that none
of the connections cross?
 It’s a complete bipartite
graph K , .
 Can K , be drawn on a
plane so that no two of
its edges cross?

170
Planner Graph
 A graph is called planar if it can be drawn on a plane without any
edges crossing
 Such a drawing is called a planar representation of the graph

K4 drawn with
K4
no crossings.

Q3 A planar
representation of Q3.

171
Is K3,3 Planner? (1/2) The curve splits the plane
into two regions: R1 and R2

 Consider
𝑣 ,𝑣 ,𝑣 ,𝑣 Connecting v1 and v2 to v4
first and v5 forms a closed curve

 Add 𝑣 v3 is either when v3 is in R2 (v3,v4) and (v3,v5)

in R1 or in R2 separate R2 into two
subregions, R21 and R22

172
Is K3,3 Planner? (2/2)

v6 could be in
 Add 𝑣 R1, R21 or R22

A similar argument there is no way to v6 must be

can be used when place v6 without connected to
v3 is in R1 forcing a crossing. v1, v2 and v3

173
Euler’s Formula
 A planar representation of a graph splits the plane into
regions
r=
e=
v=
 Euler’s Formula
 If G is a connected planar simple graph with e edges and v vertices.
 And r is the number of regions in a planar representation of G .
 Then r = e − v + .
174
Degree of a Region
 Degree of a region is the number of edges on the
boundary of R, denoted by deg( )
c
Edge {f, g} appears twice on the
boundary of and thus contributes
b
d to
3
a Edge {b, c} appears twice on the
g 6
e boundary of and thus contributes
7
to
f
175
Corollary 1 of Euler’s Formula
 Corollary
 If is a connected planar simple graph with edges and vertices,
where , then
 Proof
The degree of each region in G is
 First, we have at least


 r=e−v+

176
Example of Corollary 1
 Show that is not planner
 is a connected simple graph. By Corollary ,
if is planner, then .
K5
 However, has vertices and edges
( , ) so
 Therefore, is not planner

Corollary 1 can be used to identify non-planner graph

177
Corollary 2 of Euler’s Formula
 Corollary
 If is a connected planar simple graph, then has a vertex of
degree not exceeding five.
 Proof
 It’s true if has one or two vertices. (easy to check)
 If has at least vertices, by Corollary and thus

 If the degree of every vertex is at least 6, then contradiction

by the handshaking theorem
178
Corollary 3 of Euler’s Formula
 Corollary
 If a connected planar simple graph has edges and vertices with
and no circuits of length three, then .
 Proof The degree of each region in G is at least

 Similar to Corollary but now we have


 r=e−v+

179
Example of Corollary 3
 Show that is not planner
 has no circuits of length three K3,3
 is also a connected simple graph with
 By Corollary , if is planner, then .
 However, has vertices and edges
( , ) so
 Therefore, is not planner

180
Kuratowski’s Theorem
 If a graph contains either or as a subgraph, is not planner
 Kuratowski’s Theorem
 If a graph is not planner, it must contain a subgraph that can be obtained
from or , using certain permitted operations

181
Homeomorphic Graphs
 Elementary subdivision operation
 remove an existing edge and add a new vertex together
with edges and .
 If a graph is planner, any graph obtained by one or more
elementary subdivisions is also planner
 Two graphs and are homeomorphic if
they can be obtained from the same graph by a sequence of
elementary subdivisions.

182
Examples of Homeomorphic Graphs
a b a b a b
Remove h
f f
add f, ,
c c g
c
d e d e d e

a b
i
k
j
c g
d e
183
Kuratowski’s Theorem (Formally)
 A graph is not planar if and only if it contains a subgraph
homeomorphic to or .

homeomorphic
a subgraph to

184
Section

185
The Problem
 Try to color a map such that two regions with a common
border are assigned different colors
an edge connects two
vertices if the
corresponding regions
have a common border

 assign a color to each vertex so that no two adjacent vertices

are colored the same
186
The Chromatic Number χ (chi)
 The chromatic number χ(G) is the least number of colors needed for a
coloring of graph G.
 The Four Color Theorem
 The chromatic number of a planar graph is no greater than four.

χ(G) = χ(G) =
187
Some Facts About χ(G)
 χ(Kn) = n
 χ(Km,n) =
K5 K3,5
 χ(Cn) = if n is even
 χ(Cn) = if n is odd

C8 C7
188
A Question
 We know that K is a planner graph

K4 drawn with
K4
no crossings.

 Is K also a planner graph?

 You can answer this question
based on the fact that χ(Kn) = n. K5

189
Application1 : Scheduling Exam.
 Vertices: final exams.
 An edge connects two exams if they
are common to some students
 What is the minimum number of
time slots needed so that no
students have two exams at the time? χ(G) =

190
Application 2: Freq. Assignment
 Vertices: base stations (BSs).
 An edge connects two BSs if they can
interfere with each other using the
same frequency
 What is the minimum number of
frequencies needed so that no two
χ(G) =
BSs interfere with each other?

191