Distribution

Overview

Package

Class

Use

Deprecated

Index

Home >> All >> riso >> [ distributions overview ]

PREV CLASS NEXT CLASS

SUMMARY:

JAVADOC |

Sourc code of riso.distributions.Distribution

SOURCE |

Download riso.distributions.Distribution

DOWNLOAD | NESTED | FIELD | CONSTR | METHOD

DETAIL: FIELD | CONSTR | METHOD

riso.distributions
Interface Distribution

All Superinterfaces:: ConditionalDistribution

All Known Subinterfaces:: Delta, LocationScaleDensity, Translatable

All Known Implementing Classes:: AbstractDistribution, DiscreteDelta, Gaussian, GaussianDelta, SplineDensity

public interface Distribution
extends ConditionalDistribution

Interface for all unconditional distribution models. Note that an unconditional distribution is a special kind of conditional distribution.

Method Summary

double cdf(double x)
          Compute the cumulative distribution function.

double[] effective_support(double epsilon)
          Returns the support of this distribution, if it is a finite interval; otherwise returns an interval which contains almost all of the mass.

double expected_value()
          Returns the expected value of this distribution.

MixGaussians initial_mix(double[] support)
          Returns a Gaussian mixture which is a reasonable initial approximation to this distribution.

double log_p(double[] x)
          Compute the logarithm of the density at the point x.

double log_prior()
          Computes the log of the prior probability of the parameters of this distribution, assuming some prior distribution has been established.

int ndimensions()
          Return the number of dimensions in which this distribution lives.

double p(double[] x)
          Compute the density at the point x.

double[] random()
          Return an instance of a random variable from this distribution.

double sqrt_variance()
          Returns the square root of the variance of this distribution.

double update(double[][] x, double[] responsibility, int niter_max, double stopping_criterion)
          Use data to modify the parameters of the distribution.

Methods inherited from interface riso.distributions.ConditionalDistribution

clone, format_string, get_density, get_nstates, ndimensions_child, ndimensions_parent, p, parse_string, random, set_variable

Method Detail

ndimensions

public int ndimensions()

Return the number of dimensions in which this distribution lives.

cdf

public double cdf(double x)
           throws java.lang.Exception

Compute the cumulative distribution function.

p

public double p(double[] x)
         throws java.lang.Exception

Compute the density at the point x.

log_p

public double log_p(double[] x)
             throws java.lang.Exception

Compute the logarithm of the density at the point x.

log_prior

public double log_prior()
                 throws java.lang.Exception

Computes the log of the prior probability of the parameters of this distribution, assuming some prior distribution has been established. This may not be meaningful for all distributions.

random

public double[] random()
                throws java.lang.Exception

Return an instance of a random variable from this distribution.

update

public double update(double[][] x,
                     double[] responsibility,
                     int niter_max,
                     double stopping_criterion)
              throws java.lang.Exception

Use data to modify the parameters of the distribution. Classes which implement this method will typically use maximum likelihood or a similar approach to fit the parameters to the data.

expected_value

public double expected_value()
                      throws java.lang.Exception

Returns the expected value of this distribution.

sqrt_variance

public double sqrt_variance()
                     throws java.lang.Exception

Returns the square root of the variance of this distribution.

effective_support

public double[] effective_support(double epsilon)
                           throws java.lang.Exception

Returns the support of this distribution, if it is a finite interval; otherwise returns an interval which contains almost all of the mass.

initial_mix

public MixGaussians initial_mix(double[] support)
                         throws java.lang.Exception

Returns a Gaussian mixture which is a reasonable initial approximation to this distribution. The initial approximation should be further adjusted before using it to compute probabilities and what-not; the initial mixture can be a very rough approximation.