001/** 002 * Copyright (c) 2011, The University of Southampton and the individual contributors. 003 * All rights reserved. 004 * 005 * Redistribution and use in source and binary forms, with or without modification, 006 * are permitted provided that the following conditions are met: 007 * 008 * * Redistributions of source code must retain the above copyright notice, 009 * this list of conditions and the following disclaimer. 010 * 011 * * Redistributions in binary form must reproduce the above copyright notice, 012 * this list of conditions and the following disclaimer in the documentation 013 * and/or other materials provided with the distribution. 014 * 015 * * Neither the name of the University of Southampton nor the names of its 016 * contributors may be used to endorse or promote products derived from this 017 * software without specific prior written permission. 018 * 019 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND 020 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED 021 * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE 022 * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR 023 * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES 024 * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; 025 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON 026 * ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT 027 * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS 028 * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE. 029 */ 030package org.openimaj.math.matrix.algorithm.pca; 031 032import java.util.Arrays; 033 034import no.uib.cipr.matrix.NotConvergedException; 035import Jama.Matrix; 036 037 038/** 039 * Compute the PCA using an SVD without actually constructing 040 * the covariance matrix. This class performs a full SVD extracting all 041 * singular values and vectors. If you know apriori how many principle 042 * components (or have an upper bound on the number), then use a 043 * {@link ThinSvdPrincipalComponentAnalysis} instead. 044 * 045 * @author Jonathon Hare (jsh2@ecs.soton.ac.uk) 046 */ 047public class SvdPrincipalComponentAnalysis extends PrincipalComponentAnalysis { 048 int ndims; 049 050 /** 051 * Construct a {@link SvdPrincipalComponentAnalysis} that 052 * will extract all the eigenvectors. 053 */ 054 public SvdPrincipalComponentAnalysis() { 055 this(-1); 056 } 057 058 /** 059 * Construct a {@link SvdPrincipalComponentAnalysis} that 060 * will extract the n best eigenvectors. 061 * @param ndims the number of eigenvectors to select. 062 */ 063 public SvdPrincipalComponentAnalysis(int ndims) { 064 this.ndims = ndims; 065 } 066 067 @Override 068 public void learnBasisNorm(Matrix norm) { 069 try { 070 no.uib.cipr.matrix.DenseMatrix mjtA = new no.uib.cipr.matrix.DenseMatrix(norm.getArray()); 071 no.uib.cipr.matrix.SVD svd = no.uib.cipr.matrix.SVD.factorize(mjtA); 072 073 no.uib.cipr.matrix.DenseMatrix output = svd.getVt(); 074 075 int dims = ndims < 0 ? svd.getS().length : ndims; 076 077 basis = new Matrix(output.numColumns(), dims); 078 eigenvalues = Arrays.copyOf(svd.getS(), dims); 079 080 double normEig = 1.0 / (norm.getRowDimension() - 1); 081 for (int i=0; i<eigenvalues.length; i++) 082 eigenvalues[i] = eigenvalues[i] * eigenvalues[i] * normEig; 083 084 double[][] basisData = basis.getArray(); 085 for (int j=0; j<output.numColumns(); j++) 086 for (int i=0; i<dims; i++) 087 basisData[j][i] = output.get(i, j); 088 089 } catch (NotConvergedException e) { 090 throw new RuntimeException(e); 091 } 092 } 093}