Package org.apache.commons.rng.sampling.distribution

Class DirichletSampler

java.lang.Object

org.apache.commons.rng.sampling.distribution.DirichletSampler

All Implemented Interfaces:

ObjectSampler<double[]>, SharedStateObjectSampler<double[]>, SharedStateSampler<SharedStateObjectSampler<double[]>>

Direct Known Subclasses:

DirichletSampler.GeneralDirichletSampler, DirichletSampler.SymmetricDirichletSampler

public abstract class DirichletSampler
extends java.lang.Object
implements SharedStateObjectSampler<double[]>

Sampling from a Dirichlet distribution.

Sampling uses:

• UniformRandomProvider.nextLong()
• UniformRandomProvider.nextDouble()

Since:

1.4

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
private static class	DirichletSampler.GeneralDirichletSampler	Sample from a Dirichlet distribution with different concentration parameters for each category.
private static class	DirichletSampler.SymmetricDirichletSampler	Sample from a symmetric Dirichlet distribution with the same concentration parameter for each category.

Field Summary

Fields

Modifier and Type	Field	Description
private static int	MIN_CATGEORIES	The minimum number of categories.
private UniformRandomProvider	rng	RNG (used for the toString() method).

Constructor Summary

Constructors

Constructor	Description
DirichletSampler(UniformRandomProvider rng)

Method Summary

Modifier and Type	Method	Description
private static SharedStateContinuousSampler	createSampler(UniformRandomProvider rng, double alpha)	Creates a gamma sampler for a category with the given concentration parameter.
protected abstract int	getK()	Gets the number of categories.
private static boolean	isNonZeroPositiveFinite(double x)	Return true if the value is non-zero, positive and finite.
protected abstract double	nextGamma(int category)	Create a gamma sample for the given category.
static DirichletSampler	of(UniformRandomProvider rng, double... alpha)	Creates a new Dirichlet distribution sampler.
double[]	sample()	Create an object sample.
static DirichletSampler	symmetric(UniformRandomProvider rng, int k, double alpha)	Creates a new symmetric Dirichlet distribution sampler using the same concentration parameter for each category.
java.lang.String	toString()
private static void	validateNumberOfCategories(int k)	Validate the number of categories.
abstract DirichletSampler	withUniformRandomProvider(UniformRandomProvider rng)	Create a new instance of the sampler with the same underlying state using the given uniform random provider as the source of randomness.

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.apache.commons.rng.sampling.ObjectSampler

samples, samples

Field Detail

MIN_CATGEORIES

private static final int MIN_CATGEORIES

The minimum number of categories.

See Also:

Constant Field Values

rng

private final UniformRandomProvider rng

RNG (used for the toString() method).

Constructor Detail

DirichletSampler

DirichletSampler(UniformRandomProvider rng)

Parameters:

rng - Generator of uniformly distributed random numbers.

Method Detail

toString

public java.lang.String toString()

Overrides:

toString in class java.lang.Object

sample

public double[] sample()

Create an object sample.

Specified by:

sample in interface ObjectSampler<double[]>

Returns:

a sample.

getK

protected abstract int getK()

Gets the number of categories.

Returns:

k

nextGamma

protected abstract double nextGamma(int category)

Create a gamma sample for the given category.

Parameters:

category - Category.

Returns:

the sample

withUniformRandomProvider

public abstract DirichletSampler withUniformRandomProvider(UniformRandomProvider rng)

Create a new instance of the sampler with the same underlying state using the given uniform random provider as the source of randomness.

Specified by:

withUniformRandomProvider in interface SharedStateSampler<SharedStateObjectSampler<double[]>>

Parameters:

rng - Generator of uniformly distributed random numbers.

Returns:

the sampler

of

public static DirichletSampler of(UniformRandomProvider rng,
                                  double... alpha)

Creates a new Dirichlet distribution sampler.

Parameters:

rng - Generator of uniformly distributed random numbers.

alpha - Concentration parameters.

Returns:

the sampler

Throws:

java.lang.IllegalArgumentException - if the number of concentration parameters is less than 2; or if any concentration parameter is not strictly positive.

symmetric

public static DirichletSampler symmetric(UniformRandomProvider rng,
                                         int k,
                                         double alpha)

Creates a new symmetric Dirichlet distribution sampler using the same concentration parameter for each category.

Parameters:

rng - Generator of uniformly distributed random numbers.

k - Number of categories.

alpha - Concentration parameter.

Returns:

the sampler

Throws:

java.lang.IllegalArgumentException - if the number of categories is less than 2; or if the concentration parameter is not strictly positive.

validateNumberOfCategories

private static void validateNumberOfCategories(int k)

Validate the number of categories.

Parameters:

k - Number of categories.

Throws:

java.lang.IllegalArgumentException - if the number of categories is less than 2.

createSampler

private static SharedStateContinuousSampler createSampler(UniformRandomProvider rng,
                                                          double alpha)

Creates a gamma sampler for a category with the given concentration parameter.

Parameters:

rng - Generator of uniformly distributed random numbers.

alpha - Concentration parameter.

Returns:

the sampler

Throws:

java.lang.IllegalArgumentException - if the concentration parameter is not strictly positive.

isNonZeroPositiveFinite

private static boolean isNonZeroPositiveFinite(double x)

Return true if the value is non-zero, positive and finite.

Parameters:

x - Value.

Returns:

true if non-zero positive finite