umontreal.iro.lecuyer.randvarmulti
Class RandomMultivariateGen

java.lang.Object
  umontreal.iro.lecuyer.randvarmulti.RandomMultivariateGen

Direct Known Subclasses:

DirichletGen, IIDMultivariateGen, MultinormalGen

public abstract class RandomMultivariateGen
extends Object

This class is the multivariate counterpart of RandomVariateGen. It is the base class for general random variate generators over the d-dimensional real space R^d. It specifies the signature of the nextPoint method, which is normally called to generate a random vector from a given distribution. Contrary to univariate distributions and generators, here the inversion method is not well defined, so we cannot construct a multivariate generator simply by passing a multivariate distribution and a stream; we must specify a generating method as well. For this reason, this class is abstract. Generators can be constructed only by invoking the constructor of a subclass. This is an important difference with RandomVariateGen.

Constructor Summary

RandomMultivariateGen()

Method Summary

int	getDimension() Returns the dimension of this multivariate generator (the dimension of the random points).
RandomStream	getStream() Returns the RandomStream used by this object.
void	nextArrayOfPoints(double[][] v, int start, int n) Generates n random points.
abstract void	nextPoint(double[] p) Generates a random point p using the the stream contained in this object.
void	setStream(RandomStream stream) Sets the RandomStream used by this object to stream.

Methods inherited from class java.lang.Object

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

Constructor Detail

RandomMultivariateGen

public RandomMultivariateGen()

Method Detail

nextPoint

public abstract void nextPoint(double[] p)

Generates a random point p using the the stream contained in this object.

nextArrayOfPoints

public void nextArrayOfPoints(double[][] v,
                              int start,
                              int n)

Generates n random points. These points are stored in the array v, starting at index start. Thus v[start][i] contains coordinate i of the first generated point. By default, this method calls nextPoint n times, but one can override it in subclasses for better efficiency. The array argument v[][d] must have d elements reserved for each generated point before calling this method.

Parameters:

v - array in which the variates will be stored

start - starting index, in v, of the new variates

n - number of variates to generate

getDimension

public int getDimension()

Returns the dimension of this multivariate generator (the dimension of the random points).

getStream

public RandomStream getStream()

Returns the RandomStream used by this object.

Returns:

the stream associated to this object

setStream

public void setStream(RandomStream stream)

Sets the RandomStream used by this object to stream.