Stochastic Block Model algorithm

Development
1

It would be nice if there would be an SBM clustering algorithm implemented in igraph. We currently only have SBM sampling functions (igraph_sbm_game and the like).

@tamas, it seems you once developed an implementation of an SBM algorithm. Is there any chance that it could be included in igraph at one point?

There is of course the implementation of SBM in graph-tool by Tiago Peixoto, but I imagine it will be quite some work to translate that implementation to igraph. I don’t know of any other implementations that could be easily integrated into igraph?

2

There is; it is basically a matter of priorities only. I’m quite swamped with other things at the moment, but we can try to fit this into the development roadmap for 0.9 or 1.0, depending on how badly you guys need it :slight_smile:

3

OK, sounds good! I don’t think it is a high priority, 0.9 or 1.0 should be fine. But good to know that it sounds feasible.