Class DiscreteDistribution

java.lang.Object

random.AbstractDistribution

random.discrete.DiscreteDistribution

All Implemented Interfaces:

StorableDistribution

Direct Known Subclasses:

Bernoulli, Binomial, FixProbability, Poisson, ShiftedGeometric, Uniform, Zipf

public abstract class DiscreteDistribution
extends AbstractDistribution
implements StorableDistribution

Base class of the discrete probability distributions of the JMT random package.

Field Summary

Fields

Modifier and Type	Field	Description
protected boolean	cached
protected Hashtable<Integer,Double>	CDFList
protected boolean	CDFListCalculated

Fields inherited from class AbstractDistribution

engine

Constructor Summary

Constructors

Constructor	Description
DiscreteDistribution()

Method Summary

Modifier and Type	Method	Description
protected static int	binarySearch(int lower, int upper, double probability, Hashtable<Integer,Double> CDF_List)
abstract double	cdf(int x)
abstract double	cdf(int x, Parameter p)	obtain the cumulative distribution from the distribution
protected Hashtable<Integer,Double>	createdCDFList(int lower, int upper)	Calculate the cumulative distribution value for each discrete random variable.
protected Hashtable<Integer,Double>	createdCDFList(int lower, int upper, Parameter p)
abstract int	getlower()
abstract int	getUpper()
protected int	inverseTransformSampling(int lower, int upper)	Inverse transform sampling Take an uniform distribution - u (0, 1) as the input probability.
abstract int	nextRand()	These functions are used to direct initialize the Discrete Distribution.
abstract int	nextRand(Parameter p)	Generate random variables that follow an probability distribution
void	outdated()	Set all cache flag false, call it when parameters changed.
abstract double	pmf(int x)
abstract double	pmf(int x, Parameter p)	obtain the probability distribution from the distribution
abstract double	theorMean()
abstract double	theorMean(Parameter p)	Calculate the Mean value from the parameter of the distribution
abstract double	theorVariance()
abstract double	theorVariance(Parameter p)	Calculate the Variance from the parameter of the distribution
abstract boolean	updatePar(Parameter p)	Sub-class must implement this method to update parameter.

Methods inherited from class AbstractDistribution

setRandomEngine

Methods inherited from class Object

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

Field Details

cached

protected boolean cached

CDFList

protected Hashtable<Integer,Double> CDFList

CDFListCalculated

protected boolean CDFListCalculated

Constructor Details

DiscreteDistribution

public DiscreteDistribution()

Method Details

outdated

public void outdated()

Set all cache flag false, call it when parameters changed.

updatePar

public abstract boolean updatePar(Parameter p)
                           throws IncorrectDistributionParameterException

Sub-class must implement this method to update parameter.

Throws:

IncorrectDistributionParameterException

getUpper

public abstract int getUpper()

getlower

public abstract int getlower()

inverseTransformSampling

protected int inverseTransformSampling(int lower,
 int upper)

Inverse transform sampling Take an uniform distribution - u (0, 1) as the input probability. Find the CDF section that hold the probability generated by uniform distribution - u (0, 1). the upper bound of target section is the random variable that follow the specific distribution.

In each implementation of discrete distribution, the cumulative probability is calculated through `cdf()` this function will create the CDF List for each unit, because the cumulative probability is monotone increasing, use binary search to find the target section.

createdCDFList

protected Hashtable<Integer,Double> createdCDFList(int lower,
 int upper)

Calculate the cumulative distribution value for each discrete random variable. It depends on the `cdf()` implementation from its subclass. result include lower and upper [lower, upper]

createdCDFList

protected Hashtable<Integer,Double> createdCDFList(int lower,
 int upper,
 Parameter p)
                                            throws IncorrectDistributionParameterException

Throws:

IncorrectDistributionParameterException

binarySearch

protected static int binarySearch(int lower,
 int upper,
 double probability,
 Hashtable<Integer,Double> CDF_List)

nextRand

public abstract int nextRand()

These functions are used to direct initialize the Discrete Distribution. once invoke parameter constructor with the distribution parameter, set cached = true. if not cache the parameter, only method with `Parameter` are available.

pmf

public abstract double pmf(int x)

cdf

public abstract double cdf(int x)

theorMean

public abstract double theorMean()

theorVariance

public abstract double theorVariance()

nextRand

public abstract int nextRand(Parameter p)
                      throws IncorrectDistributionParameterException

Generate random variables that follow an probability distribution

Parameters:

p - parameter of the distribution.

Returns:

int - random variables that follow an probability distribution

Throws:

IncorrectDistributionParameterException

pmf

public abstract double pmf(int x,
 Parameter p)
                    throws IncorrectDistributionParameterException

obtain the probability distribution from the distribution

Parameters:

x - int - random variable indicating where to evaluate the pmf.

p - parameter of the distribution.

Returns:

double - the probability weight evaluated at x.

Throws:

IncorrectDistributionParameterException

cdf

public abstract double cdf(int x,
 Parameter p)
                    throws IncorrectDistributionParameterException

obtain the cumulative distribution from the distribution

Parameters:

x - int - random variable indicating where to evaluate the cdf.

p - parameter of the distribution.

Returns:

double - the probability weight cumulated at x.

Throws:

IncorrectDistributionParameterException

theorMean

public abstract double theorMean(Parameter p)
                          throws IncorrectDistributionParameterException

Calculate the Mean value from the parameter of the distribution

Parameters:

p - parameter of the distribution.

Returns:

double - theoretic mean value of the distribution.

Throws:

IncorrectDistributionParameterException

theorVariance

public abstract double theorVariance(Parameter p)
                              throws IncorrectDistributionParameterException

Calculate the Variance from the parameter of the distribution

Parameters:

p - parameter of the distribution.

Returns:

double - theoretic variance value of the distribution.

Throws:

IncorrectDistributionParameterException