de.jstacs.utils.random

Class DirichletMRGParams

java.lang.Object

de.jstacs.utils.random.MRGParams

de.jstacs.utils.random.DiMRGParams

de.jstacs.utils.random.DirichletMRGParams

public class DirichletMRGParams
extends DiMRGParams

The container for parameters of a Dirichlet random generator.

Author:

Jens Keilwagen

See Also:

DiMRGParams, FastDirichletMRGParams

Constructor Summary

Constructors

Constructor and Description
DirichletMRGParams(double... alpha) Constructor which creates a new hyperparameter vector for a Dirichlet random generator.
DirichletMRGParams(double alpha, int n) Constructor which creates a new hyperparameter vector for a Dirichlet random generator.
DirichletMRGParams(int start, int end, double... alpha) Constructor which creates a new hyperparameter vector for a Dirichlet random generator.

Method Summary

Methods

Modifier and Type	Method and Description
int	getDimension() Returns the dimension of the hyperparameter vector of the underlying Dirichlet distribution and therefore the dimension of the generated random array.
double	getHyperparameter(int i) Returns the value at position i of the hyperparameter vector of the underlying Dirichlet distribution.
double	getSumOfHyperparameter() Returns the sum of the hyperparameters (entries of the hyperparameter vector) of the underlying Dirichlet distribution.

Methods inherited from class java.lang.Object

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

Constructor Detail

DirichletMRGParams

public DirichletMRGParams(double alpha,
                  int n)
                   throws IllegalArgumentException

Constructor which creates a new hyperparameter vector for a Dirichlet random generator. The hyperparameter vector of the underlying Dirichlet distribution consists of the same value at every position.

Parameters:

alpha - the value for the hyperparameter vector

n - the dimension of the hyperparameter vector

Throws:

IllegalArgumentException - if n is chosen incorrectly (has to be greater than 2) or if alpha is chosen incorrectly (has to be positive)

See Also:

FastDirichletMRGParams.FastDirichletMRGParams(double)

DirichletMRGParams

public DirichletMRGParams(double... alpha)
                   throws IllegalArgumentException

Constructor which creates a new hyperparameter vector for a Dirichlet random generator.

Parameters:

alpha - the hyperparameter vector

Throws:

IllegalArgumentException - if at least one of the hyperparameters is not positive

DirichletMRGParams

public DirichletMRGParams(int start,
                  int end,
                  double... alpha)
                   throws IllegalArgumentException

Constructor which creates a new hyperparameter vector for a Dirichlet random generator.

Parameters:

start - start index (inclusive)

end - end index (exclusive)

alpha - the hyperparameter vector

Throws:

IllegalArgumentException - if at least one of the hyperparameters is not positive

Method Detail

getDimension

public int getDimension()

Description copied from class: DiMRGParams

Returns the dimension of the hyperparameter vector of the underlying Dirichlet distribution and therefore the dimension of the generated random array.

Specified by:

getDimension in class DiMRGParams

Returns:

the dimension of the hyperparameter vector of the underlying Dirichlet distribution and therefore the dimension of the generated random array

getHyperparameter

public double getHyperparameter(int i)

Description copied from class: DiMRGParams

Returns the value at position i of the hyperparameter vector of the underlying Dirichlet distribution.

Specified by:

getHyperparameter in class DiMRGParams

Parameters:

i - the position of the hyperparameter vector

Returns:

the value at position i of the hyperparameter vector of the underlying Dirichlet distribution

getSumOfHyperparameter

public double getSumOfHyperparameter()

Returns the sum of the hyperparameters (entries of the hyperparameter vector) of the underlying Dirichlet distribution.

Returns:

the sum of the hyperparameters (entries of the hyperparameter vector) of the underlying Dirichlet distribution