HessenbergSimilarDecomposition.java

/*
 * Copyright (c) 2009-2013, Peter Abeles. All Rights Reserved.
 *
 * This file is part of Efficient Java Matrix Library (EJML).
 *
 * Licensed under the Apache License, Version 2.0 (the "License");
 * you may not use this file except in compliance with the License.
 * You may obtain a copy of the License at
 *
 *   http://www.apache.org/licenses/LICENSE-2.0
 *
 * Unless required by applicable law or agreed to in writing, software
 * distributed under the License is distributed on an "AS IS" BASIS,
 * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
 * See the License for the specific language governing permissions and
 * limitations under the License.
 */

package mikera.matrixx.decompose.impl.hessenberg;

import mikera.matrixx.AMatrix;
import mikera.matrixx.Matrix;
import mikera.matrixx.decompose.impl.qr.QRHelperFunctions;


/**
 * <p>
 * Finds the decomposition of a matrix in the form of:<br>
 * <br>
 * A = QHQ<sup>T</sup><br>
 * <br>
 * where A is an m by m matrix, Q is an orthogonal matrix, and H is an upper Hessenberg matrix.
 * </p>
 *
 * <p>
 * A matrix is upper Hessenberg if a<sup>ij</sup> = 0 for all i > j+1.
 * </p>
 *
 * <p>
 * This decomposition is primarily used as a step for computing the eigenvalue decomposition of a matrix.
 * The basic algorithm comes from David S. Watkins, "Fundamentals of MatrixComputations" Second Edition.
 * </p>
 */
// TODO create a column based one similar to what was done for QR decomposition?
public class HessenbergSimilarDecomposition {
    // A combined matrix that stores te upper Hessenberg matrix and the orthogonal matrix.
    private Matrix QH;
    // number of rows and columns of the matrix being decompose
    private int N;

    // the first element in the orthogonal vectors
    private double gammas[];
    // temporary storage
    private double b[];
    private double u[];

    // a constructor for the static decompose method to use
    private HessenbergSimilarDecomposition() {}

    /**
     * Computes the decomposition of the provided matrix.  If no errors are detected then true is returned,
     * false otherwise.
     * @param A  The matrix that is being decomposed.  Not modified.
     * @return If it detects any errors or not.
     */
    public static HessenbergResult decompose( AMatrix A )
    {
        HessenbergSimilarDecomposition alg = new HessenbergSimilarDecomposition();
        return alg._decompose(A);
    }

    /**
     * The raw QH matrix that is stored internally.
     *
     * @return QH matrix.
     */
    public AMatrix getQH() {
        return QH;
    }

    /**
     * An upper Hessenberg matrix from the decompostion.
     *
     * @return The extracted H matrix.
     */
    private AMatrix getH() {
    	Matrix H = Matrix.create(N,N);

        // copy the first row
        System.arraycopy(QH.data, 0, H.data, 0, N);

        for( int i = 1; i < N; i++ ) {
            for( int j = i-1; j < N; j++ ) {
                H.set(i,j, QH.get(i,j));
            }
        }

        return H;
    }

    /**
     * An orthogonal matrix that has the following property: H = Q<sup>T</sup>AQ
     *
     * @return The extracted Q matrix.
     */
    private AMatrix getQ() {
        Matrix Q = Matrix.createIdentity(N);

        for( int j = N-2; j >= 0; j-- ) {
            u[j+1] = 1;
            for( int i = j+2; i < N; i++ ) {
                u[i] = QH.get(i,j);
            }
            QRHelperFunctions.rank1UpdateMultR(Q,u,gammas[j],j+1,j+1,N,b);
        }

        return Q;
    }

    /**
     * Internal function for computing the decomposition.
     * @param A 
     */
    private HessenbergResult _decompose(AMatrix A) {
    	if( A.rowCount() != A.columnCount() )
            throw new IllegalArgumentException("A must be square.");
    	QH = A.copy().toMatrix();

        N = A.columnCount();

        b = new double[ N ];
        gammas = new double[ N ];
        u = new double[ N ];
        
        double h[] = QH.data;

        for( int k = 0; k < N-2; k++ ) {
            // find the largest value in this column
            // this is used to normalize the column and mitigate overflow/underflow
            double max = 0;

            for( int i = k+1; i < N; i++ ) {
                // copy the householder vector to vector outside of the matrix to reduce caching issues
                // big improvement on larger matrices and a relatively small performance hit on small matrices.
                double val = u[i] = h[i*N+k];
                val = Math.abs(val);
                if( val > max )
                    max = val;
            }

            if( max > 0 ) {
                // -------- set up the reflector Q_k

                double tau = 0;
                // normalize to reduce overflow/underflow
                // and compute tau for the reflector
                for( int i = k+1; i < N; i++ ) {
                    double val = u[i] /= max;
                    tau += val*val;
                }

                tau = Math.sqrt(tau);

                if( u[k+1] < 0 )
                    tau = -tau;

                // write the reflector into the lower left column of the matrix
                double nu = u[k+1] + tau;
                u[k+1] = 1.0;

                for( int i = k+2; i < N; i++ ) {
                    h[i*N+k] = u[i] /= nu;
                }

                double gamma = nu/tau;
                gammas[k] = gamma;

                // ---------- multiply on the left by Q_k
                QRHelperFunctions.rank1UpdateMultR(QH,u,gamma,k+1,k+1,N,b);

                // ---------- multiply on the right by Q_k
                QRHelperFunctions.rank1UpdateMultL(QH,u,gamma,0,k+1,N);

                // since the first element in the householder vector is known to be 1
                // store the full upper hessenberg
                h[(k+1)*N+k] = -tau*max;

            } else {
                gammas[k] = 0;
            }

        }

        return new HessenbergResult(getH(), getQ());
    }

    public double[] getGammas() {
        return gammas;
    }
}