Hi,
In order to find the vertices with maximum PageRank value in unipartite projection, say g[[1]], of a bipartite graph, I'm running following script:
V(g[[1]])$pagerank <- page.rank(g[[1]])$vector
vertex_max_pagerank <- V(g[[1]])[pagerank == max( page.rank(g[[1]])$vector )]$name
However, I am getting most of the sometimes NAs as result of the that script while only a few run are giving meaningful vertex names. Moreover, if I try to run the script repeatedly for a same input (i.e. a specific incidence matrix), in each run I am getting randomly NAs or a meaningful vertex name result. I couldn't figure out the missing point, would you suggest anything for that discrepancy?
İbrahim Mutlay