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 org.openimaj.math.matrix.MatrixUtils;
033import org.openimaj.util.array.ArrayUtils;
034
035import Jama.EigenvalueDecomposition;
036import Jama.Matrix;
037
038
039/**
040 * Naive Principle Component Analysis performed by directly calculating 
041 * the covariance matrix and then performing an Eigen decomposition.
042 * 
043 * This implementation should not be used in general as it is expensive.
044 * The {@link SvdPrincipalComponentAnalysis} and {@link ThinSvdPrincipalComponentAnalysis}
045 * implementations are much faster and more efficient.
046 * 
047 * @author Jonathon Hare (jsh2@ecs.soton.ac.uk)
048 *
049 */
050public class CovarPrincipalComponentAnalysis extends PrincipalComponentAnalysis {
051        int ndims;
052        
053        /**
054         * Construct a {@link CovarPrincipalComponentAnalysis} that
055         * will extract all the eigenvectors.
056         */
057        public CovarPrincipalComponentAnalysis() {
058                this(-1);
059        }
060        
061        /**
062         * Construct a {@link CovarPrincipalComponentAnalysis} that
063         * will extract the n best eigenvectors.
064         * @param ndims the number of eigenvectors to select.
065         */
066        public CovarPrincipalComponentAnalysis(int ndims) {
067                this.ndims = ndims;
068        }
069        
070        @Override
071        protected void learnBasisNorm(Matrix m) {
072                Matrix covar = m.transpose().times(m);
073                
074                EigenvalueDecomposition eig = covar.eig();
075                Matrix all_eigenvectors = eig.getV();
076                
077                //note eigenvalues are in increasing order, so last vec is first pc
078                if (ndims > 0)
079                        basis = all_eigenvectors.getMatrix(0, all_eigenvectors.getRowDimension()-1, Math.max(0, all_eigenvectors.getColumnDimension() - ndims), all_eigenvectors.getColumnDimension()-1);
080                else 
081                        basis = all_eigenvectors;
082                
083                eigenvalues = eig.getRealEigenvalues();
084                double norm = 1.0 / (m.getRowDimension() - 1);
085                for (int i=0; i<eigenvalues.length; i++) eigenvalues[i] *= norm;
086                
087                //swap evecs
088                MatrixUtils.reverseColumnsInplace(basis);
089                
090                //swap evals
091                ArrayUtils.reverse(eigenvalues);
092        }
093}