sdsu.algorithms.data
Class NoInverseDistribution

java.lang.Object
  |
  +--sdsu.algorithms.data.ProbabilityDistribution
        |
        +--sdsu.algorithms.data.DiscreteProbability
              |
              +--sdsu.algorithms.data.NoInverseDistribution

Direct Known Subclasses:

GeneralizedEightyTwenty, Zipf

public abstract class NoInverseDistribution

extends DiscreteProbability

An abstract class handling discrete probability distributions with no inverse or easly computable inverse. The inverse is determined by brute force. Let the sample space contain items s₁, s₂, ..., s_n. Let PD( x) denote the density function of x. Map the interval [0, 1] to the sample space by:
[0, PD(s₁)] <-> s₁,
(PD(s₁), PD(s₁) + PD(s₂) ] <-> s₂,
(PD(s₁) + PD(s₂), PD(s₁) + PD(s₂) + PD(s₃)] <-> s₃,
etc. In this way the cumulative distribution of s₁, s₂, ..., s_n gives the inverse. This class computes cumulative distribution.

Version:

1.0 12 June 1997

Author:

Roger Whitney (whitney@cs.sdsu.edu)

Constructor Summary

NoInverseDistribution()

Method Summary

double

nextElement() Return the next element in a random selection of items in the sample space, using distribution of concrete subclass.

Methods inherited from class sdsu.algorithms.data.DiscreteProbability

distribution

Methods inherited from class sdsu.algorithms.data.ProbabilityDistribution

density, density, mean, variance

Methods inherited from class java.lang.Object

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

Constructor Detail

NoInverseDistribution

public NoInverseDistribution()

Method Detail

nextElement

public double nextElement()

Return the next element in a random selection of items in the sample space, using distribution of concrete subclass. The next element is an integer value from 1 to size of the sample space.

Overrides:

nextElement in class ProbabilityDistribution