weka.classifiers.functions
Class LinearRegression

java.lang.Object
  weka.classifiers.Classifier
      weka.classifiers.functions.LinearRegression

All Implemented Interfaces:

java.lang.Cloneable, OptionHandler, java.io.Serializable, WeightedInstancesHandler

public class LinearRegression

extends Classifier

implements OptionHandler, WeightedInstancesHandler

Class for using linear regression for prediction. Uses the Akaike criterion for model selection, and is able to deal with weighted instances.

Valid options are:

-D
Produce debugging output.

-S num
Set the attriute selection method to use. 1 = None, 2 = Greedy (default 0 = M5' method)

-C
Do not try to eliminate colinear attributes

-R num
The ridge parameter (default 1.0e-8)

Version:

$Revision: 1.19 $

Author:

Eibe Frank (eibe@cs.waikato.ac.nz), Len Trigg (trigg@cs.waikato.ac.nz)

See Also:

Serialized Form

Field Summary

private boolean	b_Debug True if debug output will be printed
private int	m_AttributeSelection The current attribute selection method
private boolean	m_checksTurnedOff Turn off all checks and conversions?
private int	m_ClassIndex The index of the class attribute
private double	m_ClassMean The mean of the class attribute
private double	m_ClassStdDev The standard deviations of the class attribute
private double[]	m_Coefficients Array for storing coefficients of linear regression.
private boolean	m_EliminateColinearAttributes Try to eliminate correlated attributes?
private double[]	m_Means The attributes means
private ReplaceMissingValues	m_MissingFilter The filter for removing missing values.
private double	m_Ridge The ridge parameter
private boolean[]	m_SelectedAttributes Which attributes are relevant?
private double[]	m_StdDevs The attribute standard deviations
private Instances	m_TransformedData Variable for storing transformed training data.
private NominalToBinary	m_TransformFilter The filter storing the transformation from nominal to binary attributes.
static int	SELECTION_GREEDY
static int	SELECTION_M5
static int	SELECTION_NONE
static Tag[]	TAGS_SELECTION

Fields inherited from class weka.classifiers.Classifier

m_Debug

Constructor Summary

LinearRegression()

Method Summary

java.lang.String	attributeSelectionMethodTipText() Returns the tip text for this property
void	buildClassifier(Instances data) Builds a regression model for the given data.
private double	calculateSE(boolean[] selectedAttributes, double[] coefficients) Calculate the squared error of a regression model on the training data
double	classifyInstance(Instance instance) Classifies the given instance using the linear regression function.
double[]	coefficients() Returns the coefficients for this linear model.
java.lang.String	debugTipText() Returns the tip text for this property
private boolean	deselectColinearAttributes(boolean[] selectedAttributes, double[] coefficients) Removes the attribute with the highest standardised coefficient greater than 1.5 from the selected attributes.
private double[]	doRegression(boolean[] selectedAttributes) Calculate a linear regression using the selected attributes
java.lang.String	eliminateColinearAttributesTipText() Returns the tip text for this property
private void	findBestModel() Performs a greedy search for the best regression model using Akaike's criterion.
SelectedTag	getAttributeSelectionMethod() Gets the method used to select attributes for use in the linear regression.
boolean	getDebug() Controls whether debugging output will be printed
boolean	getEliminateColinearAttributes() Get the value of EliminateColinearAttributes.
java.lang.String[]	getOptions() Gets the current settings of the classifier.
double	getRidge() Get the value of Ridge.
java.lang.String	globalInfo() Returns a string describing this classifier
java.util.Enumeration	listOptions() Returns an enumeration describing the available options.
static void	main(java.lang.String[] argv) Generates a linear regression function predictor.
int	numParameters() Get the number of coefficients used in the model
private double	regressionPrediction(Instance transformedInstance, boolean[] selectedAttributes, double[] coefficients) Calculate the dependent value for a given instance for a given regression model.
java.lang.String	ridgeTipText() Returns the tip text for this property
void	setAttributeSelectionMethod(SelectedTag method) Sets the method used to select attributes for use in the linear regression.
void	setDebug(boolean debug) Controls whether debugging output will be printed
void	setEliminateColinearAttributes(boolean newEliminateColinearAttributes) Set the value of EliminateColinearAttributes.
void	setOptions(java.lang.String[] options) Parses a given list of options.
void	setRidge(double newRidge) Set the value of Ridge.
java.lang.String	toString() Outputs the linear regression model as a string.
void	turnChecksOff() Turns off checks for missing values, etc.
void	turnChecksOn() Turns on checks for missing values, etc.

Methods inherited from class weka.classifiers.Classifier

distributionForInstance, forName, makeCopies

Methods inherited from class java.lang.Object

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

Field Detail

m_Coefficients

private double[] m_Coefficients

Array for storing coefficients of linear regression.

m_SelectedAttributes

private boolean[] m_SelectedAttributes

Which attributes are relevant?

m_TransformedData

private Instances m_TransformedData

Variable for storing transformed training data.

m_MissingFilter

private ReplaceMissingValues m_MissingFilter

The filter for removing missing values.

m_TransformFilter

private NominalToBinary m_TransformFilter

The filter storing the transformation from nominal to binary attributes.

m_ClassStdDev

private double m_ClassStdDev

The standard deviations of the class attribute

m_ClassMean

private double m_ClassMean

The mean of the class attribute

m_ClassIndex

private int m_ClassIndex

The index of the class attribute

m_Means

private double[] m_Means

The attributes means

m_StdDevs

private double[] m_StdDevs

The attribute standard deviations

b_Debug

private boolean b_Debug

True if debug output will be printed

m_AttributeSelection

private int m_AttributeSelection

The current attribute selection method

SELECTION_M5

public static final int SELECTION_M5

See Also:

Constant Field Values

SELECTION_NONE

public static final int SELECTION_NONE

See Also:

Constant Field Values

SELECTION_GREEDY

public static final int SELECTION_GREEDY

See Also:

Constant Field Values

TAGS_SELECTION

public static final Tag[] TAGS_SELECTION

m_EliminateColinearAttributes

private boolean m_EliminateColinearAttributes

Try to eliminate correlated attributes?

m_checksTurnedOff

private boolean m_checksTurnedOff

Turn off all checks and conversions?

m_Ridge

private double m_Ridge

The ridge parameter

Constructor Detail

LinearRegression

public LinearRegression()

Method Detail

turnChecksOff

public void turnChecksOff()

Turns off checks for missing values, etc. Use with caution. Also turns off scaling.

turnChecksOn

public void turnChecksOn()

Turns on checks for missing values, etc. Also turns on scaling.

globalInfo

public java.lang.String globalInfo()

Returns a string describing this classifier

Returns:

a description of the classifier suitable for displaying in the explorer/experimenter gui

buildClassifier

public void buildClassifier(Instances data)
                     throws java.lang.Exception

Builds a regression model for the given data.

Specified by:

buildClassifier in class Classifier

Parameters:

data - the training data to be used for generating the linear regression function

Throws:

java.lang.Exception - if the classifier could not be built successfully

classifyInstance

public double classifyInstance(Instance instance)
                        throws java.lang.Exception

Classifies the given instance using the linear regression function.

Overrides:

classifyInstance in class Classifier

Parameters:

instance - the test instance

Returns:

the classification

Throws:

java.lang.Exception - if classification can't be done successfully

toString

public java.lang.String toString()

Outputs the linear regression model as a string.

listOptions

public java.util.Enumeration listOptions()

Returns an enumeration describing the available options.

Specified by:

listOptions in interface OptionHandler

Overrides:

listOptions in class Classifier

Returns:

an enumeration of all the available options.

setOptions

public void setOptions(java.lang.String[] options)
                throws java.lang.Exception

Parses a given list of options. Valid options are:

-D
Produce debugging output.

-S num
Set the attriute selection method to use. 1 = None, 2 = Greedy (default 0 = M5' method)

-C
Do not try to eliminate colinear attributes

-R num
The ridge parameter (default 1.0e-8)

Specified by:

setOptions in interface OptionHandler

Overrides:

setOptions in class Classifier

Parameters:

options - the list of options as an array of strings

Throws:

java.lang.Exception - if an option is not supported

coefficients

public double[] coefficients()

Returns the coefficients for this linear model.

getOptions

public java.lang.String[] getOptions()

Gets the current settings of the classifier.

Specified by:

getOptions in interface OptionHandler

Overrides:

getOptions in class Classifier

Returns:

an array of strings suitable for passing to setOptions

ridgeTipText

public java.lang.String ridgeTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getRidge

public double getRidge()

Get the value of Ridge.

Returns:

Value of Ridge.

setRidge

public void setRidge(double newRidge)

Set the value of Ridge.

Parameters:

newRidge - Value to assign to Ridge.

eliminateColinearAttributesTipText

public java.lang.String eliminateColinearAttributesTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getEliminateColinearAttributes

public boolean getEliminateColinearAttributes()

Get the value of EliminateColinearAttributes.

Returns:

Value of EliminateColinearAttributes.

setEliminateColinearAttributes

public void setEliminateColinearAttributes(boolean newEliminateColinearAttributes)

Set the value of EliminateColinearAttributes.

Parameters:

newEliminateColinearAttributes - Value to assign to EliminateColinearAttributes.

numParameters

public int numParameters()

Get the number of coefficients used in the model

Returns:

the number of coefficients

attributeSelectionMethodTipText

public java.lang.String attributeSelectionMethodTipText()

Returns the tip text for this property

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

setAttributeSelectionMethod

public void setAttributeSelectionMethod(SelectedTag method)

Sets the method used to select attributes for use in the linear regression.

Parameters:

method - the attribute selection method to use.

getAttributeSelectionMethod

public SelectedTag getAttributeSelectionMethod()

Gets the method used to select attributes for use in the linear regression.

Returns:

the method to use.

debugTipText

public java.lang.String debugTipText()

Returns the tip text for this property

Overrides:

debugTipText in class Classifier

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

setDebug

public void setDebug(boolean debug)

Controls whether debugging output will be printed

Overrides:

setDebug in class Classifier

Parameters:

debug - true if debugging output should be printed

getDebug

public boolean getDebug()

Controls whether debugging output will be printed

Overrides:

getDebug in class Classifier

Returns:

true if debugging output is on

deselectColinearAttributes

private boolean deselectColinearAttributes(boolean[] selectedAttributes,
                                           double[] coefficients)

Removes the attribute with the highest standardised coefficient greater than 1.5 from the selected attributes.

Parameters:

selectedAttributes - an array of flags indicating which attributes are included in the regression model

coefficients - an array of coefficients for the regression model

Returns:

true if an attribute was removed

findBestModel

private void findBestModel()
                    throws java.lang.Exception

Performs a greedy search for the best regression model using Akaike's criterion.

Throws:

java.lang.Exception - if regression can't be done

calculateSE

private double calculateSE(boolean[] selectedAttributes,
                           double[] coefficients)
                    throws java.lang.Exception

Calculate the squared error of a regression model on the training data

Parameters:

selectedAttributes - an array of flags indicating which attributes are included in the regression model

coefficients - an array of coefficients for the regression model

Returns:

the mean squared error on the training data

Throws:

java.lang.Exception - if there is a missing class value in the training data

regressionPrediction

private double regressionPrediction(Instance transformedInstance,
                                    boolean[] selectedAttributes,
                                    double[] coefficients)
                             throws java.lang.Exception

Calculate the dependent value for a given instance for a given regression model.

Parameters:

transformedInstance - the input instance

selectedAttributes - an array of flags indicating which attributes are included in the regression model

coefficients - an array of coefficients for the regression model

Returns:

the regression value for the instance.

Throws:

java.lang.Exception - if the class attribute of the input instance is not assigned

doRegression

private double[] doRegression(boolean[] selectedAttributes)
                       throws java.lang.Exception

Calculate a linear regression using the selected attributes

Parameters:

selectedAttributes - an array of booleans where each element is true if the corresponding attribute should be included in the regression.

Returns:

an array of coefficients for the linear regression model.

Throws:

java.lang.Exception - if an error occurred during the regression.

main

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

Generates a linear regression function predictor.