Re: [igraph] random walks with directed edges

[Tamás Nepusz]

> the big connected component. Now edge.betweenness.community() reveals
> some meaningful communities. Do you think such an approach is sound?

Well, as long as the communities are meaningful, then I guess this approach is 
sound.
 
> Of course, your suggestion looks really interesting and I will give it
> a try, but I have no knowledge of C++, so I don't know how to manage
> such an analysis... should I use the C++ igraph? or can I call the C++
> code from R in a relatively easy way?

Well, my code compiles to a command line application that takes the network in 
an edge list format, tries to fit a stochastic blockmodel to it, and then 
returns the classification of nodes into groups (where, say, group 1 = core and 
group 2 = periphery) and the connection probabilities within the core, within 
the periphery, and between the core and the periphery. So, there's no need for 
C++ coding once the code itself is compiled. As for R, you can call the 
executable from R using system() and then parse the output to get back the node 
classification as an R vector -- but you have to write this glue code yourself.

Best,
T.