JMSLTM Numerical Library 6.1

com.imsl.stat
Class GammaDistribution

java.lang.Object
  extended by com.imsl.stat.GammaDistribution
All Implemented Interfaces:
Distribution, ProbabilityDistribution, Serializable

public class GammaDistribution
extends Object
implements ProbabilityDistribution, Serializable

Evaluates a gamma probability density for a given set of data.

GammaDistribution evaluates the gamma density of a given set of data, xData. If parameters are not supplied, the Eval method fits the gamma probability density function to the data by first calculating the shape and scale parameters using an MLE technique for a best fit. The gamma probability density function is defined as:

f(x)= {x}^{a-1}frac{e^{-frac{x}{b}}} {b^a Gamma(a)} mbox{,},,,x gt 0 ,mbox{,},,a gt 0 ,,,mbox{and},,, b gt 0 ,,mbox{,}

where a and b are the scale and shape parameters.

The DataMining package class NaiveBayesClassifier uses GammaDistribution as a method to train continuous data.

See Also:
Example, Serialized Form

Constructor Summary
GammaDistribution()
           
 
Method Summary
 double[] eval(double[] xData)
          Fits a gamma probability distribution to xData and returns the probability density at each value.
 double[] eval(double[] xData, Object[] parameters)
          Evaluates a gamma probability distribution with a given set of parameters at each point in xData and returns the probability density at each value.
 double eval(double xData, Object[] parameters)
          Evaluates a gamma probability density at a given point xData.
 Object[] getParameters()
          Returns the current parameters of the gamma probability density function.
 double getScaleParameter()
          Returns the maximum-likelihood estimate found for the gamma scale parameter.
 double getShapeParameter()
          Returns the maximum-likelihood estimate found for the gamma shape parameter.
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Constructor Detail

GammaDistribution

public GammaDistribution()
Method Detail

eval

public double[] eval(double[] xData)
Fits a gamma probability distribution to xData and returns the probability density at each value.

Specified by:
eval in interface Distribution
Parameters:
xData - a double array representing the points at which the gamma probability distribution function is to be evaluated
Returns:
a double array representing the gamma probability density at each value of xData

eval

public double[] eval(double[] xData,
                     Object[] parameters)
Evaluates a gamma probability distribution with a given set of parameters at each point in xData and returns the probability density at each value.

Specified by:
eval in interface ProbabilityDistribution
Parameters:
xData - a double array representing the points at which the gamma probability distribution function is to be evaluated
parameters - an Object array representing the parameters used to evaluate the gamma distribution, see method getParameters
Returns:
a double array representing the gamma probability density at each value of xData

eval

public double eval(double xData,
                   Object[] parameters)
Evaluates a gamma probability density at a given point xData.

Specified by:
eval in interface ProbabilityDistribution
Parameters:
xData - a double representing the point at which the gamma probability distribution function is to be evaluated
parameters - an Object array representing the parameters used to evaluate the gamma distribution, see method getParameters
Returns:
a double representing the gamma probability density at xData

getParameters

public Object[] getParameters()
Returns the current parameters of the gamma probability density function.

Specified by:
getParameters in interface ProbabilityDistribution
Returns:
an Object array containing the parameters resulting from the last invocation of the (Distribution) eval method with the following signature, double[] eval(double[] xData). This Object array can be used as input to the eval methods that require an Object array of distribution parameters as input.

getScaleParameter

public double getScaleParameter()
Returns the maximum-likelihood estimate found for the gamma scale parameter.

Returns:
a double representing the maximum-likelihood estimate found for the gamma scale parameter

getShapeParameter

public double getShapeParameter()
Returns the maximum-likelihood estimate found for the gamma shape parameter.

Returns:
a double representing the maximum-likelihood estimate found for the gamma shape parameter

JMSLTM Numerical Library 6.1

Copyright © 1970-2010 Visual Numerics, Inc.
Built July 30 2010.