DataFormats»Raw Matrices

Raw Matrices

Many network programs can read in data stored as an adjancency matrix where the value of the (i,j)th entry of the matrix gives the value of the directed relation between nodes i and j. Although conceptually elegant and useful for various matrix-based manipulations, the mode of representation becomes very inefficient for the large sparse graphs typical of most social networks. It also does not include labeling or other attribute information. The matrix format is generally very useful for exchaning data with statistical or math-based programs.

A matrix can be used to represent valued graph by using real numbers in the matrix. Self-connections can be stored on the diagonal (i=j) entries. For undirected graphs, the matrix must be symetrical. The UCINET DL format has a number of various ways of specifying input matricies, including using only the upper or lower triangle of the matrix, as well as labeling verticies. Matricies are usualy space-delimited

Brief Example:

0 0 0 0 0
1 0 1 0 0
0 1 0 0 0
1 0 0 0 1
0 0 0 1 0