org.esa.beam.util.math
Class LinEqSysSolver

java.lang.Object
  org.esa.beam.util.math.LinEqSysSolver

public class LinEqSysSolver
extends Object

A gauss-based solver for linear equation systems.

Version:

$Revision$ $Date$

Author:

Norman Fomferra

Constructor Summary

LinEqSysSolver()

Method Summary

static boolean	gauss_debug(double[][] a, double[] c, double[] x, int n, double eps) Same as the gausss method, but prints extra debug information.
static boolean	gauss(double[][] a, double[] c, double[] x, int n, double eps) Solves a linear equation system using the gaussian algorithm.
static boolean	pivotize(double[][] a, double[] c, int k, int n, double eps) Rearranges the matrix a by swapping line k with a suitable pivot line k' > k.

Methods inherited from class java.lang.Object

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

Constructor Detail

LinEqSysSolver

public LinEqSysSolver()

Method Detail

gauss

public static boolean gauss(double[][] a,
                            double[] c,
                            double[] x,
                            int n,
                            double eps)
                     throws SingularMatrixException

Solves a linear equation system using the gaussian algorithm.

Parameters:

a - coefficient matrix [0..n-1][0..n-1] (left side of equation)

c - the constant vector [0..n-1] (right side of equation)

x - the result vector [0..n-1] (right side of equation)

n - number of coefficients

eps - the epsilon value, e.g. 1.0e-09

Throws:

SingularMatrixException - if the matrix a is singular

gauss_debug

public static boolean gauss_debug(double[][] a,
                                  double[] c,
                                  double[] x,
                                  int n,
                                  double eps)

Same as the gausss method, but prints extra debug information.

Parameters:

a - coefficient matrix [0..n-1][0..n-1] (left side of equation)

c - the constant vector [0..n-1] (right side of equation)

x - the result vector [0..n-1] (right side of equation)

n - number of coefficients

eps - the epsilon value, e.g. 1.0e-09

Throws:

SingularMatrixException - if the matrix a is singular

See Also:

gauss(double[][], double[], double[], int, double)

pivotize

public static boolean pivotize(double[][] a,
                               double[] c,
                               int k,
                               int n,
                               double eps)

Rearranges the matrix a by swapping line k with a suitable pivot line k' > k.

Parameters:

a - coefficient matrix [0..n-1][0..n-1] (left side of equation)

c - the constant vector [0..n-1] (right side of equation)

k - the current index of a diagonal element = current line and column

n - matrix size

eps - the epsilon value, e.g. 1.0e-09

Throws:

SingularMatrixException - if the matrix a is singular