Re: [igraph] Obtaining communities and modularity index for an undirected graph

[Giuseppe G.]

On 21/05/12 10:52, Tamás Nepusz wrote:

> Well, that's up to you ;) The dendrogram starts with N communities (each vertex 
> in its own community). If you take 1 step, then two communities will be merged, 
> yielding N-1 communities. If you take another step, you get N-2 communities, 
> and so on. If you need, say, 4 communities in the end, you must perform N-4 
> merges.
>
> I guess that you would like to select the "optimal" number of communities with 
> the modularity function, i.e. you want to take the division which yields the highest 
> modularity. In this case, you can simply try steps=1, steps=2, steps=3, …, steps=N-1, 
> calculate the modularity score for each,  and keep the best one.


Hi Tamás, thanks for your reply. I'm not really sure if I would like to 
select the optimal number of communities based on the highest modularities.

All I've used so far is only "ready-made" implementations of the Girvan 
Neumann algorithm, and I don't remember having had to set the number of 
steps. All I remember getting was one community assignment presented to 
me as THE GN clustering for that undirected network.

Would you say that most of these implementations would basically output 
one "optimal" cluster assignment based the highest modularity function 
observed? Or based on something else?

Thanks, apologies if this does not make sense.

Best
Giuseppe

--
The University of Edinburgh is a charitable body, registered in
Scotland, with registration number SC005336.