Representing Graphs

On
Computer
Discrete Math 2
 Content
• Represent the graph with the adjacency matrix
– Properties of path on adjacency matrices
– Order Matrix and Induction Matrix
• Representing the graph with the association matrix
• Representing a graph using an edge list
• Representing a graph using an adjacency list

2
Represent the graph with the adjacency
matrix
The adjacency matrix of an undirected graph
• Consider a single undirected graph G =<V, E>, with a set of
vertices V = {1, 2, . . ., n}, edge set E = {e 1 , e 2 ,.., e m }. We call
the adjacency matrix of the graph G a matrix whose elements
are either 0 or 1 according to the following rules:

4
Adjacent matrix properties for undirected graphs

5
The adjacency matrix of a directed graph
• The definition is very similar to that of an undirected
graph
– Attention should be paid to the direction of the edge
– The adjacency matrix of a directed graph is asymmetric

6
Properties of adjacency matrices of directed graphs

7
Properties of the path on the graph using the
adjacency matrix

8
Properties of paths on graphs using adjacency
matrices (example)

9
Weight Matrix

10
Advantages & disadvantages of adjacency
matrices
• Advantage
– Simple, easy to install on the computer
– Use a two-dimensional array to represent the adjacency matrix
– The earrings still show
– It is easy to check two vertices u ,v are adjacent to each other?
• Exactly one comparison ( a[u][v]≠0?)

• Disadvantage
– Waste of memory: no matter how many or few edges we need n2
memory units for representation
– Cannot be represented with graphs with a large number of vertices
– To see what adjacent vertices u has, it takes n comparisons, even if u is
an isolated vertex or a suspended vertex.
– Can't represent parallel edges

11
 Degree Matrix
• -order matrix of the graph G is an nxn square
matrix defined as follows:

12
 Induction Matrix
• Induction matrix (Laplace matrix or Kirchhoff matrix)
• Given a graph G with n vertices, the induction matrix is
defined as follows: L=DA

Property: Every row and every column sums to 0.

13
Adjacent matrix storage format convention
• The first line records the number of vertices of the graph
• The next n lines record the graph's adjacency matrix.
– Two different elements of the adjacency matrix are written a few spaces
apart.

14
Code to represent graph using adjacency matrix

15
Representing the graph with the association
matrix
Membership Matrix: Undirected Graph

17
Membership Matrix: Directed Graph

18
What are the advantages and disadvantages of
association matrices?
• Advantage
– Looking at the matrix we know the number of edges, the
number of vertices, degrees and semi-degrees of that
vertex.
– Know the direction of an edge from which vertex goes to
which vertex (directed graph).
• Disadvantage
– Complicated representation if the graph has a
large number of edges => difficult to represent.

19
Representing a graph using an edge list
 Undirected graph using an edge list
• Only needs to list the edges (u,v) but not the edges (v,u).
• Edges should be listed in ascending order of the first vertex of each
edge.
• The edge list property of an undirected graph:
– The first vertex is smaller than the last vertex of each edge.
– The number of edges whose value u belongs to both the first and last
vertices of the edge list is the degree of vertex u.

21
 Directed graph representation using edge lists
• Each edge is a tuple taking into account the order of the vertices.
• Pay special attention to the direction of the edges
• The edge list property of a directed graph :
– The leading vertex is not necessarily smaller than the end vertex of each edge.
– The number of edges with value u in the top column of the edges is deg+ (u).
– The number of edges whose u value belongs to the vertex column at the end of the
edges is deg- (u).

22
 Weight graph using edge list

• Add a column that is the weight of each edge

23
 Edge list (arcs)
• In the case of graphs with few edges (graphs with the number of
edges m  6n), people often represent the graph as an edge list.
– Stores a list of all edges (arcs) of an undirected (directed) graph.
– Each edge (arc) e(x, y) is corresponding to two variables dau[e], cuoi[e].
– To store the graph, we need 2m memory units.
– If it is a weighted graph, we need m more memory units to store the weights
of the edges (2m+m= 3m memory units in total).

24
 Advantages & disadvantages of edge lists
• Advantages of edge lists:
– In the case of a graph with few edges (m 6n ), representing an edge
list saves memory space
– It is advantageous for some algorithms to only care about the edges
of the graph.
• Disadvantages of edge lists:
– When it is necessary to traverse the vertices adjacent to the vertex u,
it is imperative to traverse all the edges of the graph.
• This makes the algorithm very computationally expensive.

25
 Edge List Storage Format
• The first line records the number n, m corresponding to the number of
vertices and the number of edges of the graph.
– Two numbers are written side by side with a few spaces;
• m lines, each of which represents an edge of the graph
– trailing vertices of each edge are written a few spaces apart.

26
Data structure representing an edge list (1/4)

• Representing a graph's edge list using an array :

// Define an edge of the graph
typedef struct {
int data;
int cuoi;
} Edge;

//List of edges represented in WOOD Array

Edge G[MAX];

27
Data structure representing edge list (2/4)
• For example:

28
Data structure representing an edge list (3/4)

//Define a weighted edge of the graph

typedef struct {
int data;
int cuoi;
int inso;
} Edge;

//List of edges represented in WOOD Array

Edge G[MAX];

29
 Data structure representing edge list (4/4)

• Representing a graph's edge list using a linked list :

typedef struct struct{ //Defines an edge of the graph
int data;
int cuoi;
struct watch *next;
} *Edge;
Edge *G; //Edges are represented by linked list G.

30
Code to represent graph using edge list

31
Representing a graph using an adjacency list
 Adjacent list
• Each vertex u of the graph we store a list of adjacent vertices which
we denote by Ke(u), that is

33
 Advantages & disadvantages of adjacency lists
• Advantages of adjacency list:
– Easily traverse all vertices of an adjacency list;
– Easily traverse the edges of the graph in each adjacency list;
– Optimal representation and storage methods. Especially adjacency lists in arrays.
• Disadvantages of adjacency list:
– Difficult for readers with programming skills weak.
– The implementation of the problem using an adjacency list is relatively longer
than that of an adjacency matrix and an edge list.

34
 Representing an adjacency list using an Array
• array is divided into n segments
– second paragraph i in the array stores the adjacency list of
the i  V
– To know from which element an array segment starts to
which element, we use another array to store the end
positions of each segment .

• Head[ i ] is the end index of the segment managing the adjacent vertex of i-1.
– Or Head[i]+1 is the starting index of the segment that manages the adjacent vertex of i.
 The index of the segment containing the adjacent vertices of i is from Head[i]+1 to Head[i+1]

35
Representing an Adjacent List Using a Linked List

• For each vertex u  V, we represent each vertex's adjacency

list by a linked list List(u).

36
 Adjacent list storage format (using Arrays)
• The first line records the number of vertices of the graph
– WOMEN The next line records the adjacency list of the corresponding vertex in the
format:
• The first element is the end position of the segment , followed by the list of
vertices of the adjacency list
– Elements are written a few spaces apart

37
Graph representation using an Supply List (using Arrays)

38
 Graph representation using Edge Lists (using Arrays)

• Step-by-step description:
– Initially Head[i]=0
– Read each edge (u,v) and increment
Head[u]=Head[u]+1.
– Head[i]=Head[i-1]+Head[i] stack. i=2
→ n+1;
• Note: Head[n+1] is always m if it is a
directed graph.
– Traverse the edges, Adj[Head[u]] = v;
Head[u]=Head[u]-1;

Note: used for directed graphs.

39
 Graph representation using Edge Lists (using Arrays) –
for undirected graphs
• After stacking Head, MMAX (edge) must be doubled.
• If you want to store more edge weights, create an
additional array Wei has the same role as the Adj array,
but this time to store the weights of the edges.

40
 Exercise 1
a) A single undirected graph with n How many edges does a vertex have
at most?
b) During a meeting, some guests shake hands with other guests. Show
that the total number of handshakes of all guests is even.

41
 Exercise 2
Present the undirected graph by:
1) Adjacent Matrix
2) Edge list
3) Adjacent list

❖ Write a program that reads data from the file in the form of
an adjacency matrix and prints to the screen a list of the
edges of the graph, and the degree of each vertex.
42
 Exercise 3
Present the directed graph by:
1) Adjacent Matrix
2) Edge list
3) Adjacent list

❖ Write a program that reads data from the file in the form of
an adjacency list , prints the adjacency matrix and the semi-
in/out degree of each vertex.
43
 Exercise 4
Present the weight graph by:
1) Adjacent Matrix
2) Edge list
3) Adjacent list

❖ Write a program that reads data from the file in the form of a
list of sectors, prints the adjacency matrix and the sum of the
even/odd semi degrees. 44
 Exercise 5
• The chessboard 8 x 8 is numbered as follows:

• Each cell can be considered as a vertex of the graph. Two vertices are considered adjacent if a
king placed in one square can jump to the other square after one move. For example: cell 1 is
adjacent to cell 2, 9, 10, cell 11 is adjacent to 2, 3, 4, 10, 12, 18, 19, 20.
• Write a program to generate the adjacency matrix of the graph, print the results to the file
king.out.
45
 Exercise 6
• The chessboard 8  8 is numbered as follows:

• Each cell can be considered as a vertex of the graph. Two vertices are said to be adjacent if a horse
placed in one square can jump to the other square after a move. For example, cell 1 is adjacent to
11, 18, and cell 11 is adjacent to 1, 5, 17, 21, 26, 28.
• Write a program to generate the adjacency matrix of the graph, the results are written to the file
horse.out.
46