Re: [igraph] Overlapping community structures

[Tamás Nepusz]

Dear Frederik,

> I found Multi-Membership Stochastical Blockmodels particularly
> interesting. I saw that they are implemented in the R package lda. I
> would like to know if there is anything like this implemented for
> igraph
No, there is no implementation of multi-membership stochastic blockmodels in 
igraph, but if you are using R anyway, you can simply use the lda package in 
conjunction with igraph. If I understood the documentation of the lda package 
correctly, you only need the adjacency matrix of the network to execute 
mmsb.collapsed.gibbs.sampler, and you can readily obtain the adjacency matrix 
of any igraph network using get.adjacency().

-- 
T.