I am trying to use R to conduct community division within my weighted network (based from an association matrix). I tried with igraph but I encountered some problems. I usually use the program Socprog (Whitehead 2009) for my analysis, but as I would like to conduct a community division with Newman modularity (2006) on 1000 bootstraps of my data in order to calculate the comembership matrice resulting from all the community division of the bootstraps. So I am looking for a script that would conduct a community division (Newman 2006) from an association matrice and that would give a list of individuals with their cluster.

Does anyone have a R script that works for this analysis? Any advice to implement this analysis would be much apreciated.

Best regards

  • $\begingroup$ check this out and all other questions asked its author on this exchange. thanks. Weighted Community clustering $\endgroup$ – user3038 Feb 4 '11 at 12:53

Try the Louvain method for modularity optimization. It basically tried to optimize divisions within the network by recursively grouping and splitting off nodes form the bottom up.

In R, this kind of thing is super slow, but you can try the igraph package, combined with sna. The latter has methods for clustering.

If you're open to other tools, I'd actually recommend Pajek, which does this kind of thing much faster than R. It's pretty old school, but works wonderfullyhttp://pajek.imfm.si/doku.php?id=download


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy