fastICAhttp://mloss.orgUpdates and additions to fastICAenThu, 28 Feb 2013 06:30:20 -0000fastICA 0.1<html><p>Description: The open-source C-package fastICA implements the fastICA algorithm of Aapo Hyvarinen et al. &lt;URL:; to perform Independent Component Analysis (ICA) and Projection Pursuit. fastICA is released under the GNU Public License (GPL). </p> <p>Project homepage: </p> <p>Dependencies: gcc compiler, glibc library, Rmath library </p> <p>Installation: fastICA can be installed in two easy steps: </p> <ul> <li><p>extract files from "fastICA.tar.bz2": </p> </li> <li><p>compile files running the makefile: </p> </li> <li><p>make </p> </li> </ul> <p>To check the integrity of the package, run the command: </p> <blockquote><p>make check </p> </blockquote><p>To remove the installed files, use the command: </p> <blockquote><p>make clean </p> </blockquote><p>Usage: $./run_fastICA -n &lt;int&gt; -p &lt;int&gt; -# &lt;int&gt; -a -s -O &lt;dir&gt; -i &lt;file&gt; -v -h </p> <p>OPTIONS </p> <p>run_fastICA has the following options: </p> <pre><code>-n &lt;int&gt; number of observations -p &lt;int&gt; number of dimensions -# &lt;int&gt; desired number of independent components -a &lt;real&gt; alpha constant in (1, 2) for logcosh neg-entropy (default: 1.0) -s if symmetric fastICA should be performed (default: deflation) -O &lt;dir&gt; working directory for results (default: OUTPUT) -i &lt;file&gt; file with the n (row) observations of p (column) attributes -v verbose output, (default: no verbosity) </code></pre><p>EXAMPLE: </p> <p>./run_fastICA -n326900 -p3 -#3 -a1.25 -s -iTEST/karbagan.dat </p> <p>Go to TEST/README.R for information on how to validate results. </p> <p>References: </p> <pre><code> A. Hyvarinen and E. Oja (2000) Independent Component Analysis: Algorithms and Applications, _Neural Networks_, *13(4-5)*:411-430 </code></pre></html>ranjan maitraThu, 28 Feb 2013 06:30:20 -0000 component analysis