Mercurial > hg > camir-aes2014
annotate toolboxes/FullBNT-1.0.7/KPMtools/pca_netlab.m @ 0:e9a9cd732c1e tip
first hg version after svn
| author | wolffd |
|---|---|
| date | Tue, 10 Feb 2015 15:05:51 +0000 |
| parents | |
| children |
| rev | line source |
|---|---|
| wolffd@0 | 1 function [PCcoeff, PCvec] = pca(data, N) |
| wolffd@0 | 2 %PCA Principal Components Analysis |
| wolffd@0 | 3 % |
| wolffd@0 | 4 % Description |
| wolffd@0 | 5 % PCCOEFF = PCA(DATA) computes the eigenvalues of the covariance |
| wolffd@0 | 6 % matrix of the dataset DATA and returns them as PCCOEFF. These |
| wolffd@0 | 7 % coefficients give the variance of DATA along the corresponding |
| wolffd@0 | 8 % principal components. |
| wolffd@0 | 9 % |
| wolffd@0 | 10 % PCCOEFF = PCA(DATA, N) returns the largest N eigenvalues. |
| wolffd@0 | 11 % |
| wolffd@0 | 12 % [PCCOEFF, PCVEC] = PCA(DATA) returns the principal components as well |
| wolffd@0 | 13 % as the coefficients. This is considerably more computationally |
| wolffd@0 | 14 % demanding than just computing the eigenvalues. |
| wolffd@0 | 15 % |
| wolffd@0 | 16 % See also |
| wolffd@0 | 17 % EIGDEC, GTMINIT, PPCA |
| wolffd@0 | 18 % |
| wolffd@0 | 19 |
| wolffd@0 | 20 % Copyright (c) Ian T Nabney (1996-2001) |
| wolffd@0 | 21 |
| wolffd@0 | 22 if nargin == 1 |
| wolffd@0 | 23 N = size(data, 2); |
| wolffd@0 | 24 end |
| wolffd@0 | 25 |
| wolffd@0 | 26 if nargout == 1 |
| wolffd@0 | 27 evals_only = logical(1); |
| wolffd@0 | 28 else |
| wolffd@0 | 29 evals_only = logical(0); |
| wolffd@0 | 30 end |
| wolffd@0 | 31 |
| wolffd@0 | 32 if N ~= round(N) | N < 1 | N > size(data, 2) |
| wolffd@0 | 33 error('Number of PCs must be integer, >0, < dim'); |
| wolffd@0 | 34 end |
| wolffd@0 | 35 |
| wolffd@0 | 36 % Find the sorted eigenvalues of the data covariance matrix |
| wolffd@0 | 37 if evals_only |
| wolffd@0 | 38 PCcoeff = eigdec(cov(data), N); |
| wolffd@0 | 39 else |
| wolffd@0 | 40 [PCcoeff, PCvec] = eigdec(cov(data), N); |
| wolffd@0 | 41 end |
| wolffd@0 | 42 |