no.uib.cipr.matrix.EVD Maven / Gradle / Ivy
Go to download
Show more of this group Show more artifacts with this name
Show all versions of mtj Show documentation
Show all versions of mtj Show documentation
A comprehensive collection of matrix data structures, linear solvers, least squares methods,
eigenvalue, and singular value decompositions.
Forked from: https://github.com/fommil/matrix-toolkits-java
and added support for eigenvalue computation of general matrices
/*
* Copyright (C) 2003-2006 Bjørn-Ove Heimsund
*
* This file is part of MTJ.
*
* This library is free software; you can redistribute it and/or modify it
* under the terms of the GNU Lesser General Public License as published by the
* Free Software Foundation; either version 2.1 of the License, or (at your
* option) any later version.
*
* This library is distributed in the hope that it will be useful, but WITHOUT
* ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or
* FITNESS FOR A PARTICULAR PURPOSE. See the GNU Lesser General Public License
* for more details.
*
* You should have received a copy of the GNU Lesser General Public License
* along with this library; if not, write to the Free Software Foundation,
* Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
package no.uib.cipr.matrix;
import com.github.fommil.netlib.LAPACK;
import org.netlib.util.intW;
/**
* Computes eigenvalue decompositions of general matrices
*/
public class EVD {
/**
* Double work array
*/
private final double[] work;
/**
* Size of the matrix
*/
private final int n;
/**
* Job to do on the left and right eigenvectors
*/
private final JobEig jobLeft, jobRight;
/**
* Contains the real and imaginary parts of the eigenvalues
*/
private final double[] Wr, Wi;
/**
* Contains the left and the right eigenvectors
*/
private final DenseMatrix Vl, Vr;
/**
* Creates an empty eigenvalue decomposition which will compute all the
* eigenvalues and eigenvectors (left and right)
*
* @param n
* Size of the matrix
*/
public EVD(int n) {
this(n, true, true);
}
/**
* Creates an empty eigenvalue decomposition
*
* @param n
* Size of the matrix
* @param left
* Whether to compute the left eigenvectors or not
* @param right
* Whether to compute the right eigenvectors or not
*/
public EVD(int n, boolean left, boolean right) {
this.n = n;
this.jobLeft = left ? JobEig.All : JobEig.Eigenvalues;
this.jobRight = right ? JobEig.All : JobEig.Eigenvalues;
// Allocate space for the decomposition
Wr = new double[n];
Wi = new double[n];
if (left)
Vl = new DenseMatrix(n, n);
else
Vl = null;
if (right)
Vr = new DenseMatrix(n, n);
else
Vr = null;
// Find the needed workspace
double[] worksize = new double[1];
intW info = new intW(0);
LAPACK.getInstance().dgeev(jobLeft.netlib(), jobRight.netlib(), n,
new double[0], Matrices.ld(n), new double[0], new double[0],
new double[0], Matrices.ld(n), new double[0], Matrices.ld(n),
worksize, -1, info);
// Allocate workspace
int lwork = 0;
if (info.val != 0) {
if (jobLeft == JobEig.All || jobRight == JobEig.All)
lwork = 4 * n;
else
lwork = 3 * n;
} else
lwork = (int) worksize[0];
lwork = Math.max(1, lwork);
work = new double[lwork];
}
/**
* Convenience method for computing the complete eigenvalue decomposition of
* the given matrix
*
* @param A
* Matrix to factorize. Not modified
* @return Newly allocated decomposition
* @throws NotConvergedException
*/
public static EVD factorize(Matrix A) throws NotConvergedException {
return new EVD(A.numRows()).factor(new DenseMatrix(A));
}
/**
* Computes the eigenvalue decomposition of the given matrix
*
* @param A
* Matrix to factorize. Overwritten on return
* @return The current decomposition
* @throws NotConvergedException
*/
public EVD factor(DenseMatrix A) throws NotConvergedException {
if (!A.isSquare())
throw new IllegalArgumentException("!A.isSquare()");
else if (A.numRows() != n)
throw new IllegalArgumentException("A.numRows() != n");
intW info = new intW(0);
LAPACK.getInstance().dgeev(jobLeft.netlib(), jobRight.netlib(), n,
A.getData(), Matrices.ld(n), Wr, Wi,
jobLeft == JobEig.All ? Vl.getData() : new double[0],
Matrices.ld(n),
jobRight == JobEig.All ? Vr.getData() : new double[0],
Matrices.ld(n), work, work.length, info);
if (info.val > 0)
throw new NotConvergedException(
NotConvergedException.Reason.Iterations);
else if (info.val < 0)
throw new IllegalArgumentException();
return this;
}
/**
* Gets the left eigenvectors, if available
*/
public DenseMatrix getLeftEigenvectors() {
return Vl;
}
/**
* Gets the right eigenvectors, if available
*/
public DenseMatrix getRightEigenvectors() {
return Vr;
}
/**
* Gets the real part of the eigenvalues
*/
public double[] getRealEigenvalues() {
return Wr;
}
/**
* Gets the imaginary part of the eigenvalues
*/
public double[] getImaginaryEigenvalues() {
return Wi;
}
/**
* True if the left eigenvectors have been computed
*/
public boolean hasLeftEigenvectors() {
return Vl != null;
}
/**
* True if the right eigenvectors have been computed
*/
public boolean hasRightEigenvectors() {
return Vr != null;
}
}