NAME

PDL::PCA - Principal Component Analysis routines for PDL

SYNOPSIS

$a = new PDL::PCA($data,{Averaged => 1,IncludeDims => 2});

$tdata = $a->trans();

$todata = $a->trans($otherdata);

$wdata = $a->wtrans();  # Get whitened data

$tmat = $a->get_trans(); # Get trans matrix

WARNING

This is alpha-state code.

DESCRIPTION

This module implements several different algorithms for doing PCA, both off-line (statistics based) and on-line (neural-type) (XXX on-line not yet implemented).

The Averaged and IncludeDims flags are for use when you absolutely want to conserve space (i.e. $data is HUGE).