genpca.m

计量工具箱

function [inertia,casecoord,varcoord]=genpca(data,varargin)
% PURPOSE : Generalized pca implementation
%---------------------------------------------------------------------------
% USAGE : [inertia,casecoord,varcoord]=genpca(data,weight,metric,center,reduc)
%    where : data : (n x p) data matrix
%            weight : optional (n x 1) weight vector. The default is weight=(1/n,...,1/n)'.
%            metric : optional (p x p) matrix. The default is the identity matrix.
%            center : optional boolean: if center=1 : the pca is centered (default).
%                                       if center=0 : the pca is not centered.
%            reduc : optional boolean: if reduc=1 : the pca is standardized (default).
%                                      if reduc=0 : the pca is not standardized.
%----------------------------------------------------------------------------- 
% OUTPUTS :  Inertia : (p x 1) vector of inertia
%            casecoord : (n x p) matrix containing the case coordinates (principal components)
%            varcoord : (p x p) matrix containing the variables coordinates
%-----------------------------------------------------------------------------   
% used in pcamap.m
%------------------------------------------------------------------------
% Anne Ruiz-Gazen, Julien Moutel
% June 2003
% Universit