v.net.centrality computes degree, closeness, betweenness and eigenvector centrality measures.
The module computes various centrality measures for each node and stores them in the given columns of an attribute table, which is created and linked to the output map. For the description of these, please check the following wikipedia article. If the column name is not given for a measure then that measure is not computed. If -a flag is set then points are added on nodes without points. Also, the points for which the output is computed can be specified by cats, layer and where parameters. However, if any of these parameters is present then -a flag is ingored and no new points are added.
Betwenness measure is not normalised. In order to get the normalised values (between 0 and 1), each number needs to be divided by N choose 2=N*(N-1)/2 where N is the number of nodes in the connected component. Computation of eigenvector measure terminates if the given number of iterations is reached or the cummulative squared error between the successive iterations is less than error.
Compute closeness and betweenness centrality measures for each node and produce a map containing not only points already present in the input map but a map with point on every node.
v.net.centrality input=roads output=roads_cent closeness=close betweenness=betw -a