ClusterPack Matlab / Octave Toolbox (UPDATED 2011)
ClusterPack is a collection of Matlab / Octave functions for computational cluster analysis. It consists of the three modules ClusterVisual for cluster visualization, ClusterBasics for cluster analysis & evaluation, and ClusterEnsemble for cluster ensembles as described in the following. They are a selection out of my personal codebase for machine learning research. They contain general clustering algorithms as well as special algorithms developed in
my research as indicated in the README files.
Please feel free to download NEW VERSION 2.0 of
ClusterPack Matlab / Octave Toolbox (.zip).
Copyright and licensing information is included in the package. The software is provided as is. Please note that commercial use is not allowed without prior written permission. Please contact me if you would like to use any of the code for commercial purposes. The code is sparsely documented and for ease of use functions have been stripped
of optional parameters.
Also, comprehensive error checks have been excluded to keep the number of functions low and make the package easy to use. The full codebase is also available on request. The graph-based clustering module uses Metis, a separate graph-partitioning package by George Karypis from the University of Minnesota. Linux ELF and Win32 binaries of pmetis and shmetis are
included in some of the packages. However, dynamical libaries can change so depending on your system you might have to get new binaries of pmetis and shmetis or recompile them for your system.
If you run into compatability troubles, the original version 1.0 might be worth trying, so here are
ClusterPack,
ClusterVisual,
ClusterBasics and
ClusterEnsemble.
