[Top][All Lists]
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [igraph] community detection significance
From: |
Tamás Nepusz |
Subject: |
Re: [igraph] community detection significance |
Date: |
Thu, 14 Jun 2012 20:58:50 +0200 |
Dear Filipe,
None of the community significance algorithms are implemented in igraph yet and
I am not aware of any such implementations yet. A relatively simple approach is
to take each vertex in the community, calculate their internal and external
degrees (i.e. the number of edges incident on them that lead inside the
community and outside the community, respectively), and then run a Mann-Whitney
U-test. The null hypothesis is that the distribution of internal and external
degrees is the same (in other words, the community is not a real one because it
is just as dense "inside" as "outside"), and the test will reject it if the
community is a significant one. I'm sure that there are far better methods than
this, but this is relatively easy to implement both in R and in Python using
igraph.
Best,--
T.
On Thursday, 14 June 2012 at 20:55, Filipe Alberto wrote:
> Dear list,
>
> I have been using igraph to detect communities in networks. However, I would
> like to use some method to get significance of the derived communities. I see
> that there are several algorithms published, but I don't know if they are
> available in igraph or if some of the users have implemented them using R and
> igraph?
>
>
> Thanks you for your help,
>
> All the best,
> Filipe Alberto
>
> _______________________________________________
> igraph-help mailing list
> address@hidden (mailto:address@hidden)
> https://lists.nongnu.org/mailman/listinfo/igraph-help
Re: [igraph] community detection significance, Filipe Alberto, 2012/06/15