If I understood your problem correctly, you are not dealing with a single weighted graph here but an ensemble of _unweighted_ graphs where edges are independent and exist with a certain probability, and you are interested in the "consensus community structure" of the ensemble. In my opinion, the cleanest (but probably very computationally intensive) way of dealing with it would be to generate a large number of graph realisations from the ensemble, cluster each and every one of them, and then for each pair of vertices, count how many times they ended up in the same community.
T. > On 11 Aug 2015, at 22:59, Jozef Balaz <[email protected]> wrote: > > Hi, > > I’m looking for a way to define graph with weighted connection between edges > which will be then used by "Community label propagation” function. > > There are 3 use-cases: > * 2 edges are definitely not connected > * 2 edges are definitely connected > * 2 edges may be connected with certain probability > > What is the best way to define this in igraph so that community label > propagation function will take that in consideration in clustering? > Small example will really help. > > Thanks a lot for your support > Jozef > _______________________________________________ > igraph-help mailing list > [email protected] > https://lists.nongnu.org/mailman/listinfo/igraph-help _______________________________________________ igraph-help mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/igraph-help
