NAME
Statistics::CalinskiHarabasz - Perl extension to the cluster stopping
rule proposed by Calinski and Harabasz (C&H)
SYNOPSIS
use Statistics::CalinskiHarabasz;
&ch(InputFile, "agglo", 10);
Input file is expected in the "dense" format -
Sample Input file:
6 5
1 1 0 0 1
1 0 0 0 0
1 1 0 0 1
1 1 0 0 1
1 0 0 0 1
1 1 0 0 1
DESCRIPTION
C&H use the Variance Ratio Criterion which is analogous to F-Statistics
to estimate the number of clusters a given data naturally falls into.
They minimize Within Cluster/Group Sum of Squares (WGSS) and maximize
Between Cluster/Group Sum of Squares (BGSS)
EXPORT
"ch" function by default.
INPUT
InputFile
The input dataset is expected in "dense" matrix format. The input dense
matrix is expected in a plain text file where the first line in the file
gives the dimensions of the dataset and then the dataset in a matrix
format should follow. The contexts / observations should be along the
rows and the features should be along the column.
eg:
6 5
1 1 0 0 1
1 0 0 0 0
1 1 0 0 1
1 1 0 0 1
1 0 0 0 1
1 1 0 0 1
The first line (6 5) gives the number of rows (observations) and the
number of columns (features) present in the following matrix. Following
each line records the frequency of occurrence of the feature at the
column in the given observation. Thus features1 (1st column) occurs once
in the observation1 and infact once in all the other observations too
while the feature3 does not occur in observation1.
ClusteringMethod
The Clustering Measures that can be used are: 1. rb - Repeated
Bisections [Default] 2. rbr - Repeated Bisections for by k-way
refinement 3. direct - Direct k-way clustering 4. agglo - Agglomerative
clustering 5. graph - Graph partitioning-based clustering 6. bagglo -
Partitional biased Agglomerative clustering
K value
This is an approximate upper bound for the number of clusters that may
be present in the dataset. Thus for a dataset that you expect to be
seperated into 3 clusters this value should be set some integer value
greater than 3.
OUTPUT
A single integer number which is the estimate of number of clusters
present in the input dataset.
PRE-REQUISITES
1. This module uses suite of C programs called CLUTO for clustering
purposes. Thus CLUTO needs to be installed for this module to be
functional. CLUTO can be downloaded from
http://www-users.cs.umn.edu/~karypis/cluto/
SEE ALSO
1. T. Calinski and J. Harabasz. A dendrite method for cluster analysis.
Communications in statistics, 3(1):1--27, 1974.
2.
http://www-users.cs.umn.edu/~karypis/cluto/
AUTHOR
Anagha Kulkarni, University of Minnesota Duluth kulka020 <at> d.umn.edu
Guergana Savova, Mayo Clinic savova.guergana <at> mayo.edu
COPYRIGHT AND LICENSE
Copyright (C) 2005-2006, Guergana Savova and Anagha Kulkarni
This program is free software; you can redistribute it and/or modify it
under the terms of the GNU General Public License as published by the
Free Software Foundation; either version 2 of the License, or (at your
option) any later version. This program is distributed in the hope that
it will be useful, but WITHOUT ANY WARRANTY; without even the implied
warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License along
with this program; if not, write to the Free Software Foundation, Inc.,
59 Temple Place - Suite 330, Boston, MA 02111-1307, USA.
