weka.classifiers.evaluation
Class TwoClassStats

java.lang.Object
  |
  +--weka.classifiers.evaluation.TwoClassStats

public class TwoClassStats

extends java.lang.Object

Encapsulates performance functions for two-class problems.

Version:

$Revision: 1.7 $

Author:

Len Trigg (len@reeltwo.com)

Constructor Summary

TwoClassStats(double tp, double fp, double tn, double fn)
Creates the TwoClassStats with the given initial performance values.

Method Summary

ConfusionMatrix	getConfusionMatrix() Generates a ConfusionMatrix representing the current two-class statistics, using class names "negative" and "positive".
double	getFallout() Calculate the fallout.
double	getFalseNegative() Gets the number of positive instances predicted as negative
double	getFalsePositive() Gets the number of negative instances predicted as positive
double	getFalsePositiveRate() Calculate the false positive rate.
double	getFMeasure() Calculate the F-Measure.
double	getPrecision() Calculate the precision.
double	getRecall() Calculate the recall.
double	getTrueNegative() Gets the number of negative instances predicted as negative
double	getTruePositive() Gets the number of positive instances predicted as positive
double	getTruePositiveRate() Calculate the true positive rate.
void	setFalseNegative(double fn) Sets the number of positive instances predicted as negative
void	setFalsePositive(double fp) Sets the number of negative instances predicted as positive
void	setTrueNegative(double tn) Sets the number of negative instances predicted as negative
void	setTruePositive(double tp) Sets the number of positive instances predicted as positive
java.lang.String	toString() Returns a string containing the various performance measures for the current object

Methods inherited from class java.lang.Object

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

Constructor Detail

TwoClassStats

public TwoClassStats(double tp,
                     double fp,
                     double tn,
                     double fn)

Creates the TwoClassStats with the given initial performance values.

Parameters:

tp - the number of correctly classified positives

fp - the number of incorrectly classified negatives

tn - the number of correctly classified negatives

fn - the number of incorrectly classified positives

Method Detail

setTruePositive

public void setTruePositive(double tp)

Sets the number of positive instances predicted as positive

setFalsePositive

public void setFalsePositive(double fp)

Sets the number of negative instances predicted as positive

setTrueNegative

public void setTrueNegative(double tn)

Sets the number of negative instances predicted as negative

setFalseNegative

public void setFalseNegative(double fn)

Sets the number of positive instances predicted as negative

getTruePositive

public double getTruePositive()

Gets the number of positive instances predicted as positive

getFalsePositive

public double getFalsePositive()

Gets the number of negative instances predicted as positive

getTrueNegative

public double getTrueNegative()

Gets the number of negative instances predicted as negative

getFalseNegative

public double getFalseNegative()

Gets the number of positive instances predicted as negative

getTruePositiveRate

public double getTruePositiveRate()

Calculate the true positive rate. This is defined as

 correctly classified positives
 ------------------------------
       total positives
 

Returns:

the true positive rate

getFalsePositiveRate

public double getFalsePositiveRate()

Calculate the false positive rate. This is defined as

 incorrectly classified negatives
 --------------------------------
        total negatives
 

Returns:

the false positive rate

getPrecision

public double getPrecision()

Calculate the precision. This is defined as

 correctly classified positives
 ------------------------------
  total predicted as positive
 

Returns:

the precision

getRecall

public double getRecall()

Calculate the recall. This is defined as

 correctly classified positives
 ------------------------------
       total positives
 

(Which is also the same as the truePositiveRate.)

Returns:

the recall

getFMeasure

public double getFMeasure()

Calculate the F-Measure. This is defined as

 2 * recall * precision
 ----------------------
   recall + precision
 

Returns:

the F-Measure

getFallout

public double getFallout()

Calculate the fallout. This is defined as

 incorrectly classified negatives
 --------------------------------
   total predicted as positive
 

Returns:

the fallout

getConfusionMatrix

public ConfusionMatrix getConfusionMatrix()

Generates a ConfusionMatrix representing the current two-class statistics, using class names "negative" and "positive".

Returns:

a ConfusionMatrix.

toString

public java.lang.String toString()

Returns a string containing the various performance measures for the current object

Overrides:

toString in class java.lang.Object