[igraph] Newman's clustering failing [Python]

[Pagliari, Roberto]

I’m testing the leading eigenvector method for community detection.

 

·         I tested it with a bipartite weighted graph.

·         Weights are all real numbers in (0, 1]

·         I created the graph without using the specific function for creating bipartite graphs.

·         The number of nodes and edges is fairly small (order of 100) so I don’t think that should be an issue.

 

I am getting this error, due to arpack not converging. Did anyone else see this issue?

 

/usr/lib/python2.6/site-packages/python_igraph-0.7-py2.6-linux-x86_64.egg/igraph/__init__.py:1047: RuntimeWarning: ARPACK solver failed to converge (3001 iterations, 0/1 eigenvectors converged) at arpack.c:776

  membership, _, q = GraphBase.community_leading_eigenvector(self, clusters, **kwds)

 

Thank you,