ubic.basecode.math.distribution

Class Wishart

java.lang.Object

ubic.basecode.math.distribution.Wishart

public class Wishart
extends Object

Wishart distribution, used to simulate covariance matrices.

Based on method in Odell and Feiveson JASA 1966 p.199-203

The interface is modeled after ContinuousDistribution from colt, which unfortunately is designed only for univariate distributions.

Author:

pavlidis

Constructor Summary

Constructors

Constructor and Description
Wishart(double df, cern.colt.matrix.DoubleMatrix2D covariance, cern.jet.random.engine.RandomEngine randomGenerator)

Method Summary

Modifier and Type	Method and Description
cern.colt.matrix.DoubleMatrix2D	nextDoubleMatrix() Based on R code from Francesca Dominici, http ://www.biostat.jhsph.edu/~fdominic/teaching/BM/bm.html Returns w=(RU)'RU where Cov=U'U (U is upper triang) and where upper-tri R is R_ij˜N(0,1), i<j ; (R_ii)ˆ2˜Chisq(nu-s+i)

Methods inherited from class java.lang.Object

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

Constructor Detail

Wishart

public Wishart(double df,
               cern.colt.matrix.DoubleMatrix2D covariance,
               cern.jet.random.engine.RandomEngine randomGenerator)

Parameters:

s -

df -

covariance -

randomGenerator -

Method Detail

nextDoubleMatrix

public cern.colt.matrix.DoubleMatrix2D nextDoubleMatrix()

Based on R code from Francesca Dominici, http ://www.biostat.jhsph.edu/~fdominic/teaching/BM/bm.html

Returns

 w=(RU)'RU
 

where

 Cov=U'U (U is upper triang)
 

and where upper-tri R is

 R_ij˜N(0,1), i<j ; (R_ii)ˆ2˜Chisq(nu-s+i)
 

Parameters:

s -

nu -

covariance -

Returns: