ensemble-community is a small project that provides C and R-project
interfaces for community detection with ensemble-based algorithms.
These methods are intended to interface with the igraph package.

The algorithm included in this initial commit is a variant of the 
implementation described in Massen and Doye's paper:  

CP Massen and JPK Doye. "Thermodynamics of community structure".
http://arxiv.org/abs/cond-mat/0610077.

In the effdQ.pdf file, there is some basic algebra that allows each
Monte-Carlo step to be performed quite fast with the appropriate data
structures.  Each step scales on the order of the community size.

Versions:

2009-03-17. This particular implementation is a stripped-down lightweight
            version of code I wrote a while back.  It can get optimal
            communities on sparse networks of 10K nodes in hours on a
            2.4 GHz Intel processor. 

Goals:

- The next release should have an initial igraph interface as well
- Create a suite of ensemble and set-based utilites that will aid in
  analysis. I have created a few in the past, but it will be nice to
  have ones I can use directly within R.
- At some point, try to contribute the code/add it to the igraph release.