Re: [igraph] compare community detection with "ground truth"

[Claudio Martella]

Ok, that makes sense. Thanks!

On Tue, Mar 26, 2013 at 3:23 PM, Tamás Nepusz <address@hidden> wrote:

> Do I understand correctly that the 'danan' algorithm will consider the similarity of these clusterings as "low"?

No, it would indicate that the correspondence is perfect:

In [1]: cl1=Clustering([0,0,1,2,2])
In [2]: cl2=Clustering([2,2,0,1,1])
In [3]: compare_communities(cl1, cl2, method="danon")
Out[3]: 1.0

FWIW, all the other comparison methods also work this way.

--
T.
_______________________________________________
igraph-help mailing list
address@hidden
https://lists.nongnu.org/mailman/listinfo/igraph-help

--
   Claudio Martella
   address@hidden