Hi, > It simply does a breadth-first search form the starting vertex, and > the igraph data structure ensures fast traversal of the graph. >
Thanks. I suppose for order > 2 it might be worth using (sparse) matrix multiplication, depending on the parameters involved.. >> >> [g2.subgraph(g2.neighborhood(v, order = 2)).ecount() for v in g.vs] >> >> Is there a better/faster igraph way? > > I think this is probably the best way. At the C level there is a > function that gives these graphs directly, but it does essentially the > same as you are doing. > OK. Thanks. Raphael _______________________________________________ igraph-help mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/igraph-help
