Principal components analysis (PCA)

Principal components analysis (PCA)

Description

Do principal components analysis (PCA) on real-valued data.

Two methods are available: 'eig' and 'svd' which solve the problem by eigenvalue decomposition and singula value decomposition, respectively. Please note that 'svd' is running in 'economy' mode.

• MATLAB release: MATLAB 7.4 (R2007a)

[Dimitri Shvorob]

This earlier submission seems to be more comprehensive: http://www.mathworks.co.uk/matlabcentral/fileexchange/loadFile.do?objectId=17915&objectType=file
Redundant?

[Siqing Wu]

Hi, Dimitri Shvorob

Basically I think the purposes of my code and that toolbox are a little different, e.g. if you have a bunch of features and wanna reduce the dimensionality of the dataset, maybe this single code is a little more straightforward and easy to use.

Regards,

[Siqing Wu]

BTW, I found I forgot to correct one testing line (though it won't do anything wrong)... Since it takes one or two days to make the updated file posted, You may want to change "D = [D./max(D); zeros(N-nc,1)];" to "D = D/max(D);" Sorry about this.