weka.estimators
Class KKConditionalEstimator

java.lang.Object
  weka.estimators.KKConditionalEstimator

All Implemented Interfaces:

ConditionalEstimator

public class KKConditionalEstimator

extends java.lang.Object

implements ConditionalEstimator

Conditional probability estimator for a numeric domain conditional upon a numeric domain.

Version:

$Revision: 1.4 $

Author:

Len Trigg (trigg@cs.waikato.ac.nz)

Field Summary

private boolean	m_AllWeightsOne Whether we can optimise the kernel summation
private double[]	m_CondValues Vector containing all of the conditioning values seen
private int	m_NumValues Number of values stored in m_Weights, m_CondValues, and m_Values so far
private double	m_Precision The numeric precision
private double	m_StandardDev Current standard dev
private double	m_SumOfWeights The sum of the weights so far
private double[]	m_Values Vector containing all of the values seen
private double[]	m_Weights Vector containing the associated weights

Constructor Summary

KKConditionalEstimator(double precision)
Constructor

Method Summary

void	addValue(double data, double given, double weight) Add a new data value to the current estimator.
private int	findNearestPair(double key, double secondaryKey) Execute a binary search to locate the nearest data value
Estimator	getEstimator(double given) Get a probability estimator for a value
double	getProbability(double data, double given) Get a probability estimate for a value
static void	main(java.lang.String[] argv) Main method for testing this class.
private double	round(double data) Round a data value using the defined precision for this estimator
java.lang.String	toString() Display a representation of this estimator

Methods inherited from class java.lang.Object

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

Field Detail

m_Values

private double[] m_Values

Vector containing all of the values seen

m_CondValues

private double[] m_CondValues

Vector containing all of the conditioning values seen

m_Weights

private double[] m_Weights

Vector containing the associated weights

m_NumValues

private int m_NumValues

Number of values stored in m_Weights, m_CondValues, and m_Values so far

m_SumOfWeights

private double m_SumOfWeights

The sum of the weights so far

m_StandardDev

private double m_StandardDev

Current standard dev

m_AllWeightsOne

private boolean m_AllWeightsOne

Whether we can optimise the kernel summation

m_Precision

private double m_Precision

The numeric precision

Constructor Detail

KKConditionalEstimator

public KKConditionalEstimator(double precision)

Constructor

Parameters:

precision - the precision to which numeric values are given. For example, if the precision is stated to be 0.1, the values in the interval (0.25,0.35] are all treated as 0.3.

Method Detail

findNearestPair

private int findNearestPair(double key,
                            double secondaryKey)

Execute a binary search to locate the nearest data value

Parameters:

key - the data value to locate

secondaryKey - the data value to locate

Returns:

the index of the nearest data value

round

private double round(double data)

Round a data value using the defined precision for this estimator

Parameters:

data - the value to round

Returns:

the rounded data value

addValue

public void addValue(double data,
                     double given,
                     double weight)

Add a new data value to the current estimator.

Specified by:

addValue in interface ConditionalEstimator

Parameters:

data - the new data value

given - the new value that data is conditional upon

weight - the weight assigned to the data value

getEstimator

public Estimator getEstimator(double given)

Get a probability estimator for a value

Specified by:

getEstimator in interface ConditionalEstimator

Parameters:

given - the new value that data is conditional upon

Returns:

the estimator for the supplied value given the condition

getProbability

public double getProbability(double data,
                             double given)

Get a probability estimate for a value

Specified by:

getProbability in interface ConditionalEstimator

Parameters:

data - the value to estimate the probability of

given - the new value that data is conditional upon

Returns:

the estimated probability of the supplied value

toString

public java.lang.String toString()

Display a representation of this estimator

main

public static void main(java.lang.String[] argv)

Main method for testing this class. Creates some random points in the range 0 - 100, and prints out a distribution conditional on some value

Parameters:

argv - should contain: seed conditional_value numpoints