[igraph] iGraph in Python: Relation between a VertexDendrogram object and VertexClustering object

[Durante, Joseph [USA]]

Hi there,

I’m a relatively new user of iGraph and need some help with understanding a function in iGraph run in Python.

Using iGraph's community detection function community_fastgreedy(), I got a VertexDendrogram object, let's call it V. Then using V.as_clustering() I got a VertexClustering object from the dendrogram. I know that the communities are clustered so that modularity is maximized, but I think the dendogram object keeps merging all the way, so it's tough to see the communities on the dendrogram.

My question is: How does the output for the V.as_clustering() correspond to distance between the communities in the dendrogram?

In other words, The communities each come with a representative number ( Community [0], Community [2] for example), so how does the that number correspond to position on the dendrogram? Are community 0 and Community 1 adjacent to each other when the dendrogram is merged up to the level that maximizes modularity, and is community 12 farther from community 0 than 3 is?

If not, then how is the order (the numbers for each community) of the output decided by the as_clustering function?

Thanks in advance.

-Joe