be.ac.ulg.montefiore.run.totem.util.distribution
Class InverseNormalDistribution

java.lang.Object
  be.ac.ulg.montefiore.run.totem.util.distribution.InverseNormalDistribution

All Implemented Interfaces:

Distribution

public class InverseNormalDistribution
extends java.lang.Object
implements Distribution

This class implements an inverse normal distribution. The pdf of the distribution is sqrt(lambda/(2*pi*x^3)) * exp(-lambda/(2*mu^2*x) * (x-mu)^2).

This distribution is also known as the inverse gaussian distribution or the Wald distribution.

The implementation was taken from Zabel's dissertation. In this text, the minus sign in the square root of eq. (4.53) must be replaced by a plus sign.

Creation date: 14-juil.-2005

Author:

Jean Lepropre (lepropre@run.montefiore.ulg.ac.be)

Constructor Summary

InverseNormalDistribution(double mu, double lambda)
Initialises a newly created inverse normal distribution.

Method Summary

double

generate() Returns a double value according to the inverse normal distribution.

Methods inherited from class java.lang.Object

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

Constructor Detail

InverseNormalDistribution

public InverseNormalDistribution(double mu,
                                 double lambda)

Initialises a newly created inverse normal distribution.

Parameters:

mu - The location parameter of the inverse normal distribution.

lambda - The scale parameter of the inverse normal distribution.

Method Detail

generate

public double generate()

Returns a double value according to the inverse normal distribution.

Specified by:

generate in interface Distribution