Thank you for the clarification, Tamas. I think I misunderstood what is meant by the "hierarchy". I thought the clusters are recorded from each step of optimizing the modularity, but it appears that it is only from each "pass" of the algorithm. (Following the description of the paper, doi:10.1088/1742-5468/2008/10/P10008.)
Peter > Hi Peter, > > Even if you are using an older version of igraph (as far as I know there was no > release since then), you can access the membership vectors at all levels by > looking at the $memberships member of the result: > > https://lists.nongnu.org/archive/html/igraph-help/2014-10/msg00074.html > > T. > > On 03/26, Peter Carbonetto wrote: > > I am working with the multi-level community detection method implemented in > > the igraph R library, and it is working well on my data. > > > > It appears that the multilevel.community function in the R package was > > recently updated so that it outputs the cluster assignments ("memberships") > > at each level of the cluster hierarchy: > > > > http://github.com/igraph/igraph/issues/755 > > > > Are there any plans to release a new version of the R package on CRAN soon? > > If not, any suggestions for how to obtain an updated (and working) > > implementation of the new multilevel.community function? (I tried to > > install the development version from github using devtools, but I > > encountered problems.) > > > > Thank you, > > > > Peter Carbonetto > > > _______________________________________________ > > igraph-help mailing list > > address@hidden > > https://lists.nongnu.org/mailman/listinfo/igraph-help > > > -- > T.
_______________________________________________ igraph-help mailing list [email protected] https://lists.nongnu.org/mailman/listinfo/igraph-help
