Graphs

COL 106
Slide Courtesy : http://courses.cs.washington.edu/courses/cse373/
Douglas W. Harder, U Waterloo
 What are graphs?

• Yes, this is a graph….

• But we are interested in a different kind of “graph”

12/26/03 Graph Terminology - Lecture 13 2

 Graphs

• Graphs are composed of

– Nodes (vertices)
– Edges (arcs) node

edge

12/26/03 Graph Terminology - Lecture 13 3

 Varieties
• Nodes
– Labeled or unlabeled
• Edges
– Directed or undirected
– Labeled or unlabeled

12/26/03 Graph Terminology - Lecture 13 4

 Motivation for Graphs
node node
• Consider the data structures we have
looked at so far… Value Next Value Next

• Linked list: nodes with 1 incoming edge +

1 outgoing edge 94

• Binary trees/heaps: nodes with 1

incoming edge + 2 outgoing edges
10 97
• B-trees: nodes with 1 incoming edge +
multiple outgoing edges 3 5
96 99

12/26/03 Graph Terminology - Lecture 13 5

 Motivation for Graphs

• How can you generalize these data structures?

• Consider data structures for representing the
following problems…

12/26/03 Graph Terminology - Lecture 13 6

 CSE Course Prerequisites
461
322
373 106
321

100 326
410 415 370 341

413 417
378
421
Nodes = courses
Directed edge = prerequisite 401

12/26/03 Graph Terminology - Lecture 13 7

 Representing a Maze

S
S

E
E

Nodes = rooms
Edge = door or passage

12/26/03 Graph Terminology - Lecture 13 8

 Representing Electrical Circuits
Battery Switch

Nodes = battery, switch, resistor, etc. Resistor

Edges = connections

12/26/03 Graph Terminology - Lecture 13 9

 Program statements

x1 x2

x1=q+y*z + -
x2=y*z-q
Naive:
q * * q

y*z calculated twice y z

x1 x2
common
subexpression + -
eliminated:
q * q
Nodes = symbols/operators y z
Edges = relationships

12/26/03 Graph Terminology - Lecture 13 10

 Precedence

S1 a=0; 6
S2 b=1;
S3 c=a+1
5
S4 d=b+a;
S5 e=d+1;
S6 e=c+d;

4
Which statements must execute before S6? 3
S1, S2, S3, S4

Nodes = statements
Edges = precedence requirements 1 2

12/26/03 Graph Terminology - Lecture 13 11

 Information Transmission in a Computer
Network
56
Delhi Seattle
Mumbai 128

16 181 New York

30 140

L.A.
Sydney
Nodes = computers
Edges = transmission rates

12/26/03 Graph Terminology - Lecture 13 12

 Traffic Flow on Highways

UW
Nodes = cities
Edges = # vehicles on
connecting highway

12/26/03 Graph Terminology - Lecture 13 13

 Graph Definition
• A graph is simply a collection of nodes plus edges
– Linked lists, trees, and heaps are all special cases of graphs
• The nodes are known as vertices (node = “vertex”)
• Formal Definition: A graph G is a pair (V, E) where
– V is a set of vertices or nodes
– E is a set of edges that connect vertices

12/26/03 Graph Terminology - Lecture 13 14

 Graph Example

• Here is a directed graph G = (V, E)

– Each edge is a pair (v1, v2), where v1, v2 are vertices in V
– V = {A, B, C, D, E, F}
E = {(A,B), (A,D), (B,C), (C,D), (C,E), (D,E)}

B
C
A
F
D E
12/26/03 Graph Terminology - Lecture 13 15
 Directed vs Undirected Graphs
• If the order of edge pairs (v1, v2) matters, the graph is directed
(also called a digraph): (v1, v2) ≠(v2, v1)

v1 v2

• If the order of edge pairs (v1, v2) does not matter, the graph is
called an undirected graph: in this case, (v1, v2) = (v2, v1)

v1 v2

12/26/03 Graph Terminology - Lecture 13 16

 Undirected Terminology
• Two vertices u and v are adjacent in an undirected graph
G if {u,v} is an edge in G
• edge e = {u,v} is incident with vertex u and vertex v

• A graph is connected if given any two vertices u and v,

there is a path from u to v

• The degree of a vertex in an undirected graph is the

number of edges incident with it
– a self-loop counts twice (both ends count)
– denoted with deg(v)

12/26/03 Graph Terminology - Lecture 13 17

 Undirected Terminology

B is adjacent to C and C is adjacent to B

(A,B) is incident
to A and to B B
Self-loop
C
A
F
D E Degree = 0
Degree = 3

12/26/03 Graph Terminology - Lecture 13 18

 Directed Terminology

• Vertex u is adjacent to vertex v in a directed graph

G if (u,v) is an edge in G
– vertex u is the initial vertex of (u,v)
• Vertex v is adjacent from vertex u
– vertex v is the terminal (or end) vertex of (u,v)
• Degree
– in-degree is the number of edges with the vertex as
the terminal vertex
– out-degree is the number of edges with the vertex as
the initial vertex

12/26/03 Graph Terminology - Lecture 13 19

 Directed Terminology

B adjacent to C and C adjacent from B

B
C
A
F
D E In-degree = 0
Out-degree = 0
In-degree = 2
Out-degree = 1

12/26/03 Graph Terminology - Lecture 13 20

 More Graph Terminology
• simple path: no repeated vertices a b

bec
c

d e
• cycle: simple path, except that the last vertex is the same as the first
vertex a b

acda
c

d e
Graphs 21
 Even More Terminology
•connected graph: any two vertices are connected by some path

connected not connected

• subgraph: subset of vertices and edges forming a graph
• connected component: maximal connected subgraph. E.g., the
graph below has 3 connected components.

Graphs 22
 Trees from Perspective of Graphs
• (free) tree - connected graph without cycles
• forest - collection of trees
tree

tree

forest

tree

tree

Graphs 23
 Handshaking Theorem
• Let G=(V,E) be an undirected graph with |E|=e edges.
Then

2e   deg(v) Add up the degrees of all vertices.

vV

• Every edge contributes +1 to the degree of each of

the two vertices it is incident with
– number of edges is exactly half the sum of deg(v)
– the sum of the deg(v) values must be even

12/26/03 Graph Terminology - Lecture 13 24

 Connectivity
• Let n = #vertices, and m = #edges
• A complete graph: one in which all pairs of vertices are adjacent
• How many total edges in a complete graph?
– Each of the n vertices is incident to n-1 edges, however, we would have
counted each edge twice!!! Therefore, intuitively, m = n(n -1)/2.
• Therefore, if a graph is not complete, m < n(n -1)/2

n 5
m  (5 

Graphs 25
 Spanning Tree
• A spanning tree of G is a subgraph which is a tree and which
contains all vertices of G

G spanning tree of G

• Failure on any edge disconnects system (least fault tolerant)

Graphs 26
 Graph Representations

• Space and time are analyzed in terms of:

• Number of vertices = |V| and
• Number of edges = |E|
• There are at least two ways of representing
graphs:
• The adjacency matrix representation
• The adjacency list representation
12/26/03 Graph Terminology - Lecture 13 27
 Adjacency Matrix
A B C D E F
B
C A 0 1 0 1 0 0
A
B 1 0 1 0 0 0
F
C 0 1 0 1 1 0
D E
D 1 0 1 0 1 0
E 0 0 1 1 0 0
1 if (v, w) is in E
M(v, w) =
0 otherwise F 0 0 0 0 0 0
Space = |V|2
12/26/03 Graph Terminology - Lecture 13 28
 Adjacency Matrix for a Digraph
A B C D E F
B
C A 0 1 0 1 0 0
A
B 0 0 1 0 0 0
F
C 0 0 0 1 1 0
D E
D 0 0 0 0 1 0
E 0 0 0 0 0 0
1 if (v, w) is in E
M(v, w) =
0 otherwise F 0 0 0 0 0 0
Space = |V|2
12/26/03 Graph Terminology - Lecture 13 29
 Adjacency List
For each v in V, L(v) = list of w such that (v, w) is in E
a b

B A B D
list of
neighbors
C B
A A C

F C B D E
D E D A C E
E C D
F
Space = a |V| + 2 b |E|
12/26/03 Graph Terminology - Lecture 13 30
 Adjacency List for a Digraph
For each v in V, L(v) = list of w such that (v, w) is in E
a b

B A B D
C B
A C

F C D E
D E D E
E
F
Space = a |V| + b |E|
12/26/03 Graph Terminology - Lecture 13 31
 Searching in graphs
• Find Properties of Graphs
– Spanning trees
– Connected components
– Bipartite structure
– Biconnected components
• Applications
– Finding the web graph – used by Google and others
– Garbage collection – used in Java run time system

12/26/03 Graph Searching - Lecture 16 32

 Graph Searching Methodology
Breadth-First Search (BFS)
• Breadth-First Search (BFS)
– Use a queue to explore neighbors of source
vertex, then neighbors of neighbors etc.
– All nodes at a given distance (in number of edges)
are explored before we go further

12/26/03 Graph Searching - Lecture 16 33

 Breadth-First Search
• A Breadth-First Search (BFS) traverses a connected component of a
graph, and in doing so defines a spanning tree with several useful
properties.
• The starting vertex s has level 0, and defines that point as an
“anchor.”
• In the first round, all of the nodes that are only one edge away from
the anchor are visited.
• These nodes are placed into level 1
• In the second round, all the new nodes that can be reached by one
edge from level 1 nodes are visited and placed in level 2.
• This continues until every vertex has been assigned a level.
• The label of any vertex v corresponds to the length of the shortest
path from s to v.
34
 Example
Consider the graph from previous example
 Example
Performing a breadth-first traversal
– Push the first vertex onto the queue

A
 Example
Performing a breadth-first traversal
– Pop A and push B, C and E
A

B C E
 Example
Performing a breadth-first traversal:
– Pop B and push D
A, B

C E D
 Example
Performing a breadth-first traversal:
– Pop C and push F
A, B, C

E D F
 Example
Performing a breadth-first traversal:
– Pop E and push G and H
A, B, C, E

D F G H
 Example
Performing a breadth-first traversal:
– Pop D
A, B, C, E, D

F G H
 Example
Performing a breadth-first traversal:
– Pop F
A, B, C, E, D, F

G H
 Example
Performing a breadth-first traversal:
– Pop G and push I
A, B, C, E, D, F, G

H I
 Example
Performing a breadth-first traversal:
– Pop H
A, B, C, E, D, F, G, H

I
 Example
Performing a breadth-first traversal:
– Pop I
A, B, C, E, D, F, G, H, I
 Example
Performing a breadth-first traversal:
– The queue is empty: we are finished
A, B, C, E, D, F, G, H, I
 Another Example
0 0 1

A B C D A B C D

a) b)
E F G H E F G H

I J K L I J K L

M N O P M N O P
0 1 2
0 1 2 3
A B C D B C D
A
c) d)
E F G H E F G H

I J K L
I J K L

M N O P
47
M N O P
 Example Continued
0 1 2 3 0 1 2 3

A B C D A B C D

E F G H E F G H

4 4
I J K L I J K L

M N O P M N O P 5

Generic DFS and BFS 48

 Breadth-First Search

BFS
Initialize Q to be empty;
Enqueue(Q,1) and mark 1;
while Q is not empty do
i := Dequeue(Q);
for each j adjacent to i do
if j is not marked then
Enqueue(Q,j) and mark j;
end{BFS}

12/26/03 Graph Searching - Lecture 16 49

 BFS Pseudo-Code
Algorithm BFS(s): Input: A vertex s in a graph
Output: A labeling of the edges as “discovery” edges and “cross edges”
initialize container L0 to contain vertex s
i0
while Li is not empty do
create container Li+1 to initially be empty
for each vertex v in Li do
if edge e incident on v do
let w be the other endpoint of e
if vertex w is unexplored then
label e as a discovery edge
insert w into Li+1
else label e as a cross edge
ii+1

50
 Properties of BFS
• Proposition: Let G be an undirected graph on which a a BFS traversal starting at
vertex s has been performed. Then
– The traversal visits all vertices in the connected component of s .
– The discovery-edges form a spanning tree T, which we call the BFS tree, of the
connected component of s
– For each vertex v at level i, the path of the BFS tree T between s and v has i edges, and
any other path of G between s and v has at least i edges.
– If (u, v) is an edge that is not in the BFS tree, then the level numbers of u and v differ by
at most one.
• Proposition: Let G be a graph with n vertices and m edges. A BFS traversal of G
takes time O(n + m). Also, there exist O(n + m) time algorithms based on BFS for
the following problems:
– Testing whether G is connected.
– Computing a spanning tree of G
– Computing the connected components of G
– Computing, for every vertex v of G, the minimum number of edges of any path between
s and v. 52
 BFS Properties
• Proposition: Let G be an undirected graph on which a BFS traversal
starting at a vertex s has been preformed. Then:
1. The traversal visits all vertices in the connected component of s
2. The discovery edges form a spanning tree of the connected component of s

• Justification of 1:
– Let’s use a contradiction argument: suppose there is at least on vertex v not visited and let
w be the first unvisited vertex on some path from s to v.
– Because w was the first unvisited vertex on the path, there is a neighbor u that has been
visited.
– But when we visited u we must have looked at edge(u, w). Therefore w must have been
visited.

• Justification of 2:
– We only mark edges from when we go to unvisited vertices. So we never form a cycle of
discovery edges, i.e. discovery edges form a tree.
– This is a spanning tree because BFS visits each vertex in the connected component of s

53
 Graph Searching Methodology Depth-
First Search (DFS)
• Depth-First Search (DFS)
– Searches down one path as deep as possible
– When no nodes available, it backtracks
– When backtracking, it explores side-paths that
were not taken
– Uses a stack (instead of a queue in BFS)
– Allows an easy recursive implementation

12/26/03 Graph Searching - Lecture 16 54

 Depth First Search Algorithm
• Recursive marking algorithm
• Initially every vertex is unmarked DFS(i)
i

DFS(i: vertex)
mark i; DFS(j)
j
for each j adjacent to i do
if j is unmarked then DFS(j)
end{DFS} k

Marks all vertices reachable from i

12/26/03 Graph Searching - Lecture 16 55

 Depth-First Search
Algorithm DFS(v); Input: A vertex v in a graph
Output: A labeling of the edges as “discovery” edges and
“backedges”
for each edge e incident on v do
if edge e is unexplored then let w be the other endpoint of e
if vertex w is unexplored then label e as a discovery edge
recursively call DFS(w)
else label e as a backedge

DFS 56
 DFS Application: Spanning Tree
• Given a (undirected) connected graph G(V,E) a
spanning tree of G is a graph G’(V’,E’)
– V’ = V, the tree touches all vertices (spans) the
graph
– E’ is a subset of E such that G’ is connected and
there is no cycle in G’

12/26/03 Graph Searching - Lecture 16 57

 Example of DFS: Graph connectivity
and spanning tree
2
1 DFS(1)
7

5
6
4

12/26/03 Graph Searching - Lecture 16 58

 Example Step 2
2
DFS(1)
1 7 DFS(2)

5
6
4

Red links will define the spanning tree if the graph

is connected

12/26/03 Graph Searching - Lecture 16 59

 Example Step 5
2
DFS(1)
1 7 DFS(2)
DFS(3)
3 DFS(4)
5 DFS(5)

6
4

12/26/03 Graph Searching - Lecture 16 60

 Example Steps 6 and 7
2
DFS(1)
1 7 DFS(2)
DFS(3)
3 DFS(4)
DFS(5)
5
DFS(3)
6 DFS(7)
4

12/26/03 Graph Searching - Lecture 16 61

 Example Steps 8 and 9
2
DFS(1)
1 7 DFS(2)
DFS(3)
3 DFS(4)
DFS(5)
5
DFS(7)
6
Now back up.
4

12/26/03 Graph Searching - Lecture 16 62

 Example Step 10 (backtrack)
2
DFS(1)
1 7 DFS(2)
DFS(3)
3 DFS(4)
5 DFS(5)

6 Back to 5,
but it has no
more neighbors.
4

12/26/03 Graph Searching - Lecture 16 63

 Example Step 12
2
DFS(1)
1 7 DFS(2)
DFS(3)
3 DFS(4)
5 DFS(6)

6 Back up to 4.
From 4 we can
4 get to 6.

12/26/03 Graph Searching - Lecture 16 64

 Example Step 13
2
DFS(1)
1 7 DFS(2)
DFS(3)
3 DFS(4)
5 DFS(6)

6 From 6 there is
nowhere new
to go. Back up.
4

12/26/03 Graph Searching - Lecture 16 65

 Example Step 14
2
DFS(1)
1 7 DFS(2)
DFS(3)
3 DFS(4)
5
Back to 4.
6 Keep backing up.

12/26/03 Graph Searching - Lecture 16 66

 Example Step 17
2
DFS(1)
1 7
All the way
3 back to 1.

5 Done.

6
4
All nodes are marked so graph is connected; red links
define a spanning tree

12/26/03 Graph Searching - Lecture 16 67

 Finding Connected Components using DFS

2
1 3 7
10 4
9
5
6
8
11

3 connected components

12/26/03 Graph Searching - Lecture 16 68

 Connected Components

2
1 3 7
10 4
9
5
6
8
11

3 connected components are labeled

12/26/03 Graph Searching - Lecture 16 69

 Running Time Analysis
• Remember:
-DFS is called on each vertex exactly once.
-Every edge is examined exactly twice, once from each of its vertices
• For ns vertices and ms edges in the connected component of the
vertex s, a DFS starting at s runs in O(ns +ms) time if the graph is
represented in a data structure, like the adjacency list, where vertex
and edge methods take constant time.
• Marking a vertex as explored, and testing to see if a vertex has been
explored, takes O(degree)
• By marking visited nodes, we can systematically consider the edges
incident on the current vertex so we do not examine the same edge
more than once.

DFS 70
 Marking Vertices
• Let’s look at ways to mark vertices in a way that satisfies the above
condition.
• Extend vertex positions to store a variable for marking

Before After
Position Position

Element Element isMarked

Use a hash table mechanism which satisfies the above condition is the probabilistic sense,
because is supports the mark and test operations in O(1) expected time

DFS 71
 Performance DFS
• n vertices and m edges
• Storage complexity O(n + m)
• Time complexity O(n + m)
• Linear Time!

12/26/03 Graph Searching - Lecture 16 72

 DFS Properties
• Proposition: Let G be an undirected graph on which a DFS traversal
starting at a vertex s has been preformed. Then:
1. The traversal visits all vertices in the connected component of s
2. The discovery edges form a spanning tree of the connected component of s

• Justification of 1:
– Let’s use a contradiction argument: suppose there is at least on vertex v not visited and let
w be the first unvisited vertex on some path from s to v.
– Because w was the first unvisited vertex on the path, there is a neighbor u that has been
visited.
– But when we visited u we must have looked at edge(u, w). Therefore w must have been
visited.

• Justification of 2:
– We only mark edges from when we go to unvisited vertices. So we never form a cycle of
discovery edges, i.e. discovery edges form a tree.
– This is a spanning tree because DFS visits each vertex in the connected component of s

84
 Depth-First vs Breadth-First
• Depth-First
– Stack or recursion
– Many applications
• Breadth-First
– Queue (recursion no help)
– Can be used to find shortest paths from the start vertex

12/26/03 Graph Searching - Lecture 16 85