org.cicirello.math.la

Class JacobiDiagonalization

java.lang.Object

org.cicirello.math.la.JacobiDiagonalization

public class JacobiDiagonalization
extends Object

This class uses Jacobi iteration to compute the eigenvalues and eigenvectors of a symmetric matrix, provided as a 2-D Java array.

Since:

1.3

Field Summary

Fields

Modifier and Type	Field and Description
static double	EPSILON Default precision level for convergence check.
static int	MAX_ITERATIONS Default max number of iterations.

Constructor Summary

Constructors

Constructor and Description
JacobiDiagonalization(double[][] matrix) Initializes a Jacobi iteration for finding eigenvalues and eigenvectors of a matrix.
JacobiDiagonalization(int[][] matrix) Initializes a Jacobi iteration for finding eigenvalues and eigenvectors of a matrix.

Method Summary

Modifier and Type	Method and Description
boolean	compute() Computes the eigenvalues and eigenvectors using Jacobi Iteration.
boolean	compute(double epsilon) Computes the eigenvalues and eigenvectors using Jacobi Iteration.
boolean	compute(double epsilon, int maxIters) Computes the eigenvalues and eigenvectors using Jacobi Iteration.
boolean	compute(int maxIters) Computes the eigenvalues and eigenvectors using Jacobi Iteration.
double[]	eigenvalues() Gets the computed eigenvalues.
double[][]	eigenvectors() Gets the computed eigenvectors.

Methods inherited from class java.lang.Object

clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Detail

EPSILON

public static final double EPSILON

Default precision level for convergence check.

See Also:

Constant Field Values

MAX_ITERATIONS

public static final int MAX_ITERATIONS

Default max number of iterations.

See Also:

Constant Field Values

Constructor Detail

JacobiDiagonalization

public JacobiDiagonalization(int[][] matrix)

Initializes a Jacobi iteration for finding eigenvalues and eigenvectors of a matrix.

Parameters:

matrix - a square matrix

JacobiDiagonalization

public JacobiDiagonalization(double[][] matrix)

Initializes a Jacobi iteration for finding eigenvalues and eigenvectors of a matrix.

Parameters:

matrix - a square matrix

Method Detail

eigenvectors

public double[][] eigenvectors()

Gets the computed eigenvectors.

Returns:

The eigenvectors are the columns of the returned matrix. Returns null if not yet computed.

eigenvalues

public double[] eigenvalues()

Gets the computed eigenvalues.

Returns:

The eigenvalues. Returns null if not yet computed.

compute

public boolean compute()

Computes the eigenvalues and eigenvectors using Jacobi Iteration. Uses the default precision level JacobiDiagonalization.EPSILON where off-diagonal elements less than JacobiDiagonalization.EPSILON in absolute value are set to 0. Uses the default max number of iterations JacobiDiagonalization.MAX_ITERATIONS.

Returns:

true if Jacobi method converged, and false if terminated due to maximum number of iterations

compute

public boolean compute(double epsilon)

Computes the eigenvalues and eigenvectors using Jacobi Iteration. Uses the default max number of iterations JacobiDiagonalization.MAX_ITERATIONS.

Parameters:

epsilon - Precision level, where off-diagonal elements less than epsilon in absolute value are set to 0.

Returns:

true if Jacobi method converged, and false if terminated due to maximum number of iterations

compute

public boolean compute(int maxIters)

Computes the eigenvalues and eigenvectors using Jacobi Iteration. Uses the default precision level JacobiDiagonalization.EPSILON where off-diagonal elements less than JacobiDiagonalization.EPSILON in absolute value are set to 0.

Parameters:

maxIters - Maximum number of iterations.

Returns:

true if Jacobi method converged, and false if terminated due to maximum number of iterations

compute

public boolean compute(double epsilon,
                       int maxIters)

Computes the eigenvalues and eigenvectors using Jacobi Iteration.

Parameters:

epsilon - Precision level, where off-diagonal elements less than epsilon in absolute value are set to 0.

maxIters - Maximum number of iterations.

Returns:

true if Jacobi method converged, and false if terminated due to maximum number of iterations