com.aliasi.stats

Class BernoulliDistribution

java.lang.Object

com.aliasi.stats.AbstractDiscreteDistribution

com.aliasi.stats.MultivariateDistribution

com.aliasi.stats.BernoulliDistribution

All Implemented Interfaces:

DiscreteDistribution

Direct Known Subclasses:

BernoulliConstant, BernoulliEstimator

public abstract class BernoulliDistribution
extends MultivariateDistribution

A BernoulliDistribution is a multivariate distribution with two outcomes, 0 (labeled "failure") and 1 (labeled "success"). A Bernoulli distribution is the basis of the binomial distribution.

For more information, see:

• Eric W. Weisstein. Bernoulli Distribution. From MathWorld--A Wolfram Web Resource.

Since:

LingPipe2.0

Version:

2.0

Author:

Bob Carpenter

Field Summary

Fields

Modifier and Type	Field and Description
static String	FAILURE_LABEL The label for dimension zero: "failure".
static String	SUCCESS_LABEL The label for dimension one: "success".

Constructor Summary

Constructors

Constructor and Description
BernoulliDistribution() Construct a Bernoulli distribution.

Method Summary

Methods

Modifier and Type	Method and Description
String	label(long outcome) Returns the failure label for outcome zero, and the success label for outcome one.
long	maxOutcome() Returns one, the maximum outcome for a Bernoulli distribution.
int	numDimensions() Returns two, the number of dimensions for a Bernoulli distribution.
double	probability(long outcome) Returns the probability of the specified outcome.
abstract double	successProbability() Returns the success probability for this distribution.
double	variance() Returns the variance of this Bernoulli distribution.

Methods inherited from class com.aliasi.stats.MultivariateDistribution

log2Probability, minOutcome, outcome, probability

Methods inherited from class com.aliasi.stats.AbstractDiscreteDistribution

cumulativeProbability, cumulativeProbabilityGreater, cumulativeProbabilityLess, entropy, log2Probability, mean

Methods inherited from class java.lang.Object

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

Field Detail

FAILURE_LABEL

public static final String FAILURE_LABEL

The label for dimension zero: "failure".

See Also:

Constant Field Values

SUCCESS_LABEL

public static final String SUCCESS_LABEL

The label for dimension one: "success".

See Also:

Constant Field Values

Constructor Detail

BernoulliDistribution

public BernoulliDistribution()

Construct a Bernoulli distribution.

Method Detail

maxOutcome

public long maxOutcome()

Returns one, the maximum outcome for a Bernoulli distribution.

Specified by:

maxOutcome in interface DiscreteDistribution

Overrides:

maxOutcome in class MultivariateDistribution

Returns:

The long integer one.

numDimensions

public int numDimensions()

Returns two, the number of dimensions for a Bernoulli distribution.

Specified by:

numDimensions in class MultivariateDistribution

Returns:

Two, the number of dimensions for a Bernoulli distribution.

variance

public double variance()

Returns the variance of this Bernoulli distribution. Applying the general definition of variance to a Bernoulli distribution yields:

variance = P(success) * (1 - P(success))

Specified by:

variance in interface DiscreteDistribution

Overrides:

variance in class AbstractDiscreteDistribution

Returns:

The variance of this distribution.

probability

public double probability(long outcome)

Returns the probability of the specified outcome. This method is implemented to return one minus the success probability for outcome zero, the success probabilty for outcome one, and zero for all other outcomes.

Specified by:

probability in interface DiscreteDistribution

Specified by:

probability in class MultivariateDistribution

Parameters:

outcome - Outcome whose probability is returned.

Returns:

The probability of the specified outcome.

label

public String label(long outcome)

Returns the failure label for outcome zero, and the success label for outcome one. The label for zero is FAILURE_LABEL and the label for one is SUCCESS_LABEL respectively.

Overrides:

label in class MultivariateDistribution

Parameters:

outcome - Outcome whose label is returned.

Returns:

The label for the outcome.

Throws:

IllegalArgumentException - If the outcome is out of range.

successProbability

public abstract double successProbability()

Returns the success probability for this distribution. This method must be defined by concrete subclasses, and will be used to define probability(long).

Returns:

The probability of success for this distribution.