[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [igraph] Please check my procedure of community detection
From: |
Gabor Csardi |
Subject: |
Re: [igraph] Please check my procedure of community detection |
Date: |
Mon, 31 Mar 2008 22:37:59 +0200 |
User-agent: |
Mutt/1.5.13 (2006-08-11) |
[Sorry, i was offline for some time.]
On Fri, Mar 28, 2008 at 08:45:36AM +0100, Tamas Nepusz wrote:
> Looks OK to me - although I'm not really familiar with R.
>
> >I wonder if the edge weights set in step 2 will affect the
> >community results.
> Well, Gabor can tell us if the fastgreedy.community interface in
> R supports edge weights or not. I just peeked into the source
> and it looks like it doesn't (although the C core supports it,
> so probably the R interface has not been updated when edge
> weights were introduced in the C layer).
Hmmm, you're right, this is sad. I'll add it soon.
> Why don't you try
> walktrap.community instead? According to my own experience, it
> usually produces better modularity scores and it supports edge
> weights in the R interface as well.
That's funny, for me usually fastgreedy.community() works
better. :)
G.
> --
> T.
>
>
>
> _______________________________________________
> igraph-help mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/igraph-help
--
Csardi Gabor <address@hidden> UNIL DGM