Re: [igraph] Re: community detection

[Gabor Csardi]

Matthieu,

On Wed, Mar 28, 2007 at 08:37:35PM +0200, Matthieu Latapy wrote:
> Hi all.
> 
> You may also consider the algorithm proposed in
>   http://www.liafa.jussieu.fr/~pons/publi/communities.pdf
> and its implementation:
>   http://www.liafa.jussieu.fr/~pons/index.php?item=prog&item2=walktrap&lang=en

thanks, it looks good, except that this is c++ as well. :) More and more c++ 
code
sneaks in ... i've heard that igraph does not compile on Minix because of 
the c++ code, so this will be a serious problem.... :)

Hmmm, interestingly we had a similar small project in a summer school, here is 
roughly the result: 
http://arxiv.org/abs/cond-mat/0408263
But your algorithm seems much more sophisticated and clearly better. It is also
very nice that -- if i see it correctly -- it has a single parameter, the length
of the walks. I'll check the implementation as soon as i'll have some more time 
and
incorporate it into igraph. But you're encouraged to do this yourself
if you have the time and motivation. :)

> See also the implementation at:
>  http://www.cs.unm.edu/~aaron/research/fastmodularity.htm

Thanks, yes, of course i know about this. More c++ code. :)

Gabor

ps. please consider joining the mailing list (or mailing from the
address you used for signing up) otherwise i have to acknowledge your
messages manually. Thanks.

> of the algorithm described in the paper "Finding community structure
> in very large networks" already cited.
> 
> Best,
> ML
> 
> 
> _______________________________________________
> igraph-help mailing list
> address@hidden
> http://lists.nongnu.org/mailman/listinfo/igraph-help

-- 
Csardi Gabor <address@hidden>    MTA RMKI, ELTE TTK