Worksheet 40: Graph Representations Name:
An Active Learning Approach to Data Structures using C
1
Worksheet 40: Graph Representations
In preparation: Reach Chapter 13 to learn more about graphs and graph algorithms.
As you learned in Chapter 13, a graph is composed of vertices, connected by edges.
Vertices can carry information, often just a name. Edges can either be directed, which
indicates an asymmetric relationship, or undirected. In addition, edges can carry a value,
termed a weight. The following is an example graph with directed and weighted edges.
There are two common ways of representing a graph as a data structure. These are as an
adjacency matrix, and as an edge list. To form an adjacency matrix the vertices are
assigned a number. For example, we could number the vertices of the graph above by
listing the cities in alphabetical order: 0-
Pendleton, 1-Pensacola, 2-Peoria, 3-Phoenix,
4-Pierre, 5-Pittsburgh, 6-Princeton, and 7-
Pueblo. An 8 by 8 matrix is then constructed.
Each (i,j) entry describes the association
between vertex i and vertex j. In an unweighted
graph this entry is either 0 if there is no
connection, or 1 if there is a connection
between the two cities. In a weighted matrix
the value is the weight on the arc linking the
two cities, or the value infinity if there is no
connection. In the space at right enter the edge list representation for the graph shown
above. An adjacency matrix requires O(v2) space, where v is the number of vertices. This
is true regardless of the number of edges.
The alternative representation, the edge list, stores only the edges of a graph. This is
advantageous if the graph is relatively sparse. Fundamentally, the edge list uses a map
indexed by the vertex (or vertex label). For an unweighted graph the value is a set of
vertices that represent the neighbors of the key vertex. In a weighted graph the value is
itself represented by another map. In this map the key is the neighboring vertex, and the
0
1
2
3
4
5
6
7
0
1
2
3
4
5
6
7
