weka.core

Class RandomVariates

java.lang.Object

java.util.Random

weka.core.RandomVariates

All Implemented Interfaces:

Serializable, RevisionHandler

public final class RandomVariates
extends Random
implements RevisionHandler

Class implementing some simple random variates generator.

Version:

$Revision: 8034 $

Author:

Xin Xu (xx5@cs.waikato.ac.nz)

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
RandomVariates() Simply the constructor of super class
RandomVariates(long seed) Simply the constructor of super class

Method Summary

Methods

Modifier and Type	Method and Description
String	getRevision() Returns the revision string.
static void	main(String[] ops) Main method for testing this class.
double	nextErlang(int a) Generate a value of a variate following standard Erlang distribution.
double	nextExponential() Generate a value of a variate following standard exponential distribution using simple inverse method.
double	nextGamma(double a) Generate a value of a variate following standard Gamma distribution with shape parameter a.

Methods inherited from class java.util.Random

nextBoolean, nextBytes, nextDouble, nextFloat, nextGaussian, nextInt, nextInt, nextLong, setSeed

Methods inherited from class java.lang.Object

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

Constructor Detail

RandomVariates

public RandomVariates()

Simply the constructor of super class

RandomVariates

public RandomVariates(long seed)

Simply the constructor of super class

Parameters:

seed - the seed in this random object

Method Detail

nextExponential

public double nextExponential()

Generate a value of a variate following standard exponential distribution using simple inverse method.

Variates related to standard Exponential can be generated using simple transformations.

Returns:

a value of the variate

nextErlang

public double nextErlang(int a)
                  throws Exception

Generate a value of a variate following standard Erlang distribution. It can be used when the shape parameter is an integer and not too large, say, <100. When the parameter is not an integer (which is often named Gamma distribution) or is large, use nextGamma(double a) instead.

Parameters:

a - the shape parameter, must be no less than 1

Returns:

a value of the variate

Throws:

Exception - if parameter less than 1

nextGamma

public double nextGamma(double a)
                 throws Exception

Generate a value of a variate following standard Gamma distribution with shape parameter a.

If a>1, it uses a rejection method developed by Minh(1988)"Generating Gamma Variates", ACM Trans. on Math. Software, Vol.14, No.3, pp261-266.
If a<1, it uses the algorithm "GS" developed by Ahrens and Dieter(1974) "COMPUTER METHODS FOR SAMPLING FROM GAMMA, BETA, POISSON AND BINOMIAL DISTRIBUTIONS", COMPUTING, 12 (1974), pp223-246, and further implemented in Fortran by Ahrens, Kohrt and Dieter(1983) "Algorithm 599: sampling from Gamma and Poisson distributions", ACM Trans. on Math. Software, Vol.9 No.2, pp255-257.

Variates related to standard Gamma can be generated using simple transformations.

Parameters:

a - the shape parameter, must be greater than 1

Returns:

a value of the variate

Throws:

Exception - if parameter not greater than 1

getRevision

public String getRevision()

Returns the revision string.

Specified by:

getRevision in interface RevisionHandler

Returns:

the revision

main

public static void main(String[] ops)

Main method for testing this class.

Parameters:

ops - # of variates/seed, default is 10/45