fr.cnes.sirius.patrius.math.optim.joptimizer.util

Class Utils

java.lang.Object

fr.cnes.sirius.patrius.math.optim.joptimizer.util.Utils

public final class Utils
extends Object

Utility class.

Since:

4.6

Author:

alberto trivellato (alberto.trivellato@gmail.com)

Method Summary

Modifier and Type	Method and Description
static double	calculateScaledResidual(RealMatrix a, RealMatrix x, RealMatrix b) Calculate the scaled residual ||Ax-b||_oo/( ||A||_oo .
static double	calculateScaledResidual(RealMatrix a, RealVector x, RealVector b) Calculate the scaled residual ||Ax-b||_oo/( ||A||_oo .
static double	getDoubleMachineEpsilon() The smallest positive (epsilon) such that 1.0 + epsilon !
static double	max(double a, double b) Returns max(a, b) or NaN if one value is a NaN.
static double	min(double a, double b) Returns min(a, b) or NaN if one value is a NaN.
static RealMatrix	randomValuesMatrix(int rows, int cols, double min, double max, Long seed) Returns matrix filled with random values.
static RealMatrix	randomValuesPositiveMatrix(int rows, int cols, double min, double max, Long seed) Returns matrix filled with random positive values.
static double[]	replaceValues(double[] v, double oldValue, double newValue) Return a new array with all the occurrences of oldValue replaced by newValue.

Methods inherited from class java.lang.Object

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

Method Detail

getDoubleMachineEpsilon

public static final double getDoubleMachineEpsilon()

The smallest positive (epsilon) such that 1.0 + epsilon != 1.0.

Returns:

smallest positive (epsilon) such that 1.0 + epsilon != 1.0 See http://en.wikipedia.org/wiki/Machine_epsilon#Approximation_using_Java

calculateScaledResidual

public static double calculateScaledResidual(RealMatrix a,
                                             RealMatrix x,
                                             RealMatrix b)

Calculate the scaled residual
||Ax-b||_oo/( ||A||_oo . ||x||_oo + ||b||_oo ), with
||x||_oo = max(||x[i]||)

Parameters:

a - A matrix

x - X matrix

b - B matrix

Returns:

scaled residual

calculateScaledResidual

public static double calculateScaledResidual(RealMatrix a,
                                             RealVector x,
                                             RealVector b)

Calculate the scaled residual
||Ax-b||_oo/( ||A||_oo . ||x||_oo + ||b||_oo ), with
||x||_oo = max(||x[i]||)

Parameters:

a - A matrix

x - X matrix

b - B matrix

Returns:

scaled residual

randomValuesMatrix

public static RealMatrix randomValuesMatrix(int rows,
                                            int cols,
                                            double min,
                                            double max,
                                            Long seed)

Returns matrix filled with random values.

Parameters:

rows - number of rows

cols - number of columns

min - min

max - max

seed - seed

Returns:

matrix filled with random values

randomValuesPositiveMatrix

public static RealMatrix randomValuesPositiveMatrix(int rows,
                                                    int cols,
                                                    double min,
                                                    double max,
                                                    Long seed)

Returns matrix filled with random positive values.

Parameters:

rows - number of rows

cols - number of columns

min - min

max - max

seed - seed

Returns:

matrix filled with random positive values

See Also:

"http://mathworld.wolfram.com/PositiveDefiniteMatrix.html"

replaceValues

public static final double[] replaceValues(double[] v,
                                           double oldValue,
                                           double newValue)

Return a new array with all the occurrences of oldValue replaced by newValue.

Parameters:

v - array

oldValue - old value

newValue - new value

Returns:

updated array

max

public static double max(double a,
                         double b)

Returns max(a, b) or NaN if one value is a NaN.

Parameters:

a - a

b - b

Returns:

max(a, b) or NaN if one value is a NaN

min

public static double min(double a,
                         double b)

Returns min(a, b) or NaN if one value is a NaN.

Parameters:

a - a

b - b

Returns:

min(a, b) or NaN if one value is a NaN