weka.filters.unsupervised.attribute
Class Discretize

java.lang.Object
  weka.filters.Filter
      weka.filters.unsupervised.attribute.PotentialClassIgnorer
          weka.filters.unsupervised.attribute.Discretize

All Implemented Interfaces:

OptionHandler, java.io.Serializable, UnsupervisedFilter, WeightedInstancesHandler

Direct Known Subclasses:

PKIDiscretize

public class Discretize

extends PotentialClassIgnorer

implements UnsupervisedFilter, OptionHandler, WeightedInstancesHandler

An instance filter that discretizes a range of numeric attributes in the dataset into nominal attributes. Discretization is by simple binning. Skips the class attribute if set.

Valid filter-specific options are:

-B num
Specifies the (maximum) number of bins to divide numeric attributes into. Default = 10.

-M num
Specifies the desired weight of instances per bin for equal-frequency binning. If this is set to a positive number then the -B option will be ignored. Default = -1.

-F
Use equal-frequency instead of equal-width discretization if class-based discretisation is turned off.

-O
Optimize the number of bins using a leave-one-out estimate of the entropy (for equal-width binning). If this is set then the -B option will be ignored.

-R col1,col2-col4,...
Specifies list of columns to Discretize. First and last are valid indexes. (default: first-last)

-V
Invert matching sense.

-D
Make binary nominal attributes.

Version:

$Revision: 1.6 $

Author:

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

See Also:

Serialized Form

Field Summary

protected double[][]	m_CutPoints Store the current cutpoints
protected java.lang.String	m_DefaultCols The default columns to discretize
protected double	m_DesiredWeightOfInstancesPerInterval The desired weight of instances per bin
protected Range	m_DiscretizeCols Stores which columns to Discretize
protected boolean	m_FindNumBins Find the number of bins using cross-validated entropy.
protected boolean	m_MakeBinary Output binary attributes for discretized attributes.
protected int	m_NumBins The number of bins to divide the attribute into
protected boolean	m_UseEqualFrequency Use equal-frequency binning if unsupervised discretization turned on

Fields inherited from class weka.filters.unsupervised.attribute.PotentialClassIgnorer

m_ClassIndex, m_IgnoreClass

Fields inherited from class weka.filters.Filter

m_NewBatch

Constructor Summary

Discretize()
Constructor - initialises the filter

Discretize(java.lang.String cols)
Another constructor

Method Summary

java.lang.String	attributeIndicesTipText() Returns the tip text for this property
boolean	batchFinished() Signifies that this batch of input to the filter is finished.
java.lang.String	binsTipText() Returns the tip text for this property
protected void	calculateCutPoints() Generate the cutpoints for each attribute
protected void	calculateCutPointsByEqualFrequencyBinning(int index) Set cutpoints for a single attribute.
protected void	calculateCutPointsByEqualWidthBinning(int index) Set cutpoints for a single attribute.
protected void	convertInstance(Instance instance) Convert a single instance over.
java.lang.String	desiredWeightOfInstancesPerIntervalTipText() Returns the tip text for this property
protected void	findNumBins(int index) Optimizes the number of bins using leave-one-out cross-validation.
java.lang.String	findNumBinsTipText() Returns the tip text for this property
java.lang.String	getAttributeIndices() Gets the current range selection
int	getBins() Gets the number of bins numeric attributes will be divided into
double[]	getCutPoints(int attributeIndex) Gets the cut points for an attribute
double	getDesiredWeightOfInstancesPerInterval() Get the DesiredWeightOfInstancesPerInterval value.
boolean	getFindNumBins() Get the value of FindNumBins.
boolean	getInvertSelection() Gets whether the supplied columns are to be removed or kept
boolean	getMakeBinary() Gets whether binary attributes should be made for discretized ones.
java.lang.String[]	getOptions() Gets the current settings of the filter.
boolean	getUseEqualFrequency() Get the value of UseEqualFrequency.
java.lang.String	globalInfo() Returns a string describing this filter
boolean	input(Instance instance) Input an instance for filtering.
java.lang.String	invertSelectionTipText() Returns the tip text for this property
java.util.Enumeration	listOptions() Gets an enumeration describing the available options.
static void	main(java.lang.String[] argv) Main method for testing this class.
java.lang.String	makeBinaryTipText() Returns the tip text for this property
void	setAttributeIndices(java.lang.String rangeList) Sets which attributes are to be Discretized (only numeric attributes among the selection will be Discretized).
void	setAttributeIndicesArray(int[] attributes) Sets which attributes are to be Discretized (only numeric attributes among the selection will be Discretized).
void	setBins(int numBins) Sets the number of bins to divide each selected numeric attribute into
void	setDesiredWeightOfInstancesPerInterval(double newDesiredNumber) Set the DesiredWeightOfInstancesPerInterval value.
void	setFindNumBins(boolean newFindNumBins) Set the value of FindNumBins.
boolean	setInputFormat(Instances instanceInfo) Sets the format of the input instances.
void	setInvertSelection(boolean invert) Sets whether selected columns should be removed or kept.
void	setMakeBinary(boolean makeBinary) Sets whether binary attributes should be made for discretized ones.
void	setOptions(java.lang.String[] options) Parses the options for this object.
protected void	setOutputFormat() Set the output format.
void	setUseEqualFrequency(boolean newUseEqualFrequency) Set the value of UseEqualFrequency.
java.lang.String	useEqualFrequencyTipText() Returns the tip text for this property

Methods inherited from class weka.filters.unsupervised.attribute.PotentialClassIgnorer

getOutputFormat, setIgnoreClass

Methods inherited from class weka.filters.Filter

batchFilterFile, bufferInput, copyStringValues, copyStringValues, filterFile, flushInput, getInputFormat, getInputStringIndex, getOutputStringIndex, getStringIndices, inputFormat, inputFormatPeek, isOutputFormatDefined, numPendingOutput, output, outputFormat, outputFormatPeek, outputPeek, push, resetQueue, setOutputFormat, useFilter

Methods inherited from class java.lang.Object

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

Field Detail

m_DiscretizeCols

protected Range m_DiscretizeCols

Stores which columns to Discretize

m_NumBins

protected int m_NumBins

The number of bins to divide the attribute into

m_DesiredWeightOfInstancesPerInterval

protected double m_DesiredWeightOfInstancesPerInterval

The desired weight of instances per bin

m_CutPoints

protected double[][] m_CutPoints

Store the current cutpoints

m_MakeBinary

protected boolean m_MakeBinary

Output binary attributes for discretized attributes.

m_FindNumBins

protected boolean m_FindNumBins

Find the number of bins using cross-validated entropy.

m_UseEqualFrequency

protected boolean m_UseEqualFrequency

Use equal-frequency binning if unsupervised discretization turned on

m_DefaultCols

protected java.lang.String m_DefaultCols

The default columns to discretize

Constructor Detail

Discretize

public Discretize()

Constructor - initialises the filter

Discretize

public Discretize(java.lang.String cols)

Another constructor

Method Detail

listOptions

public java.util.Enumeration listOptions()

Gets an enumeration describing the available options.

Specified by:

listOptions in interface OptionHandler

Returns:

an enumeration of all the available options.

setOptions

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

Parses the options for this object. Valid options are:

-B num
Specifies the (maximum) number of bins to divide numeric attributes into. Default = 10.

-M num
Specifies the desired weight of instances per bin for equal-frequency binning. If this is set to a positive number then the -B option will be ignored. Default = -1.

-F
Use equal-frequency instead of equal-width discretization if class-based discretisation is turned off.

-O
Optimize the number of bins using a leave-one-out estimate of the entropy (for equal-width binning). If this is set then the -B option will be ignored.

-R col1,col2-col4,...
Specifies list of columns to Discretize. First and last are valid indexes. (default none)

-V
Invert matching sense.

-D
Make binary nominal attributes.

Specified by:

setOptions in interface OptionHandler

Parameters:

options - the list of options as an array of strings

Throws:

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

getOptions

public java.lang.String[] getOptions()

Gets the current settings of the filter.

Specified by:

getOptions in interface OptionHandler

Returns:

an array of strings suitable for passing to setOptions

setInputFormat

public boolean setInputFormat(Instances instanceInfo)
                       throws java.lang.Exception

Sets the format of the input instances.

Overrides:

setInputFormat in class PotentialClassIgnorer

Parameters:

instanceInfo - an Instances object containing the input instance structure (any instances contained in the object are ignored - only the structure is required).

Returns:

true if the outputFormat may be collected immediately

Throws:

java.lang.Exception - if the input format can't be set successfully

input

public boolean input(Instance instance)

Input an instance for filtering. Ordinarily the instance is processed and made available for output immediately. Some filters require all instances be read before producing output.

Overrides:

input in class Filter

Parameters:

instance - the input instance

Returns:

true if the filtered instance may now be collected with output().

Throws:

java.lang.IllegalStateException - if no input format has been defined.

batchFinished

public boolean batchFinished()

Signifies that this batch of input to the filter is finished. If the filter requires all instances prior to filtering, output() may now be called to retrieve the filtered instances.

Overrides:

batchFinished in class Filter

Returns:

true if there are instances pending output

Throws:

java.lang.IllegalStateException - if no input structure has been defined

globalInfo

public java.lang.String globalInfo()

Returns a string describing this filter

Returns:

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

findNumBinsTipText

public java.lang.String findNumBinsTipText()

Returns the tip text for this property

Returns:

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

getFindNumBins

public boolean getFindNumBins()

Get the value of FindNumBins.

Returns:

Value of FindNumBins.

setFindNumBins

public void setFindNumBins(boolean newFindNumBins)

Set the value of FindNumBins.

Parameters:

newFindNumBins - Value to assign to FindNumBins.

makeBinaryTipText

public java.lang.String makeBinaryTipText()

Returns the tip text for this property

Returns:

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

getMakeBinary

public boolean getMakeBinary()

Gets whether binary attributes should be made for discretized ones.

Returns:

true if attributes will be binarized

setMakeBinary

public void setMakeBinary(boolean makeBinary)

Sets whether binary attributes should be made for discretized ones.

Parameters:

makeBinary - if binary attributes are to be made

desiredWeightOfInstancesPerIntervalTipText

public java.lang.String desiredWeightOfInstancesPerIntervalTipText()

Returns the tip text for this property

Returns:

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

getDesiredWeightOfInstancesPerInterval

public double getDesiredWeightOfInstancesPerInterval()

Get the DesiredWeightOfInstancesPerInterval value.

Returns:

the DesiredWeightOfInstancesPerInterval value.

setDesiredWeightOfInstancesPerInterval

public void setDesiredWeightOfInstancesPerInterval(double newDesiredNumber)

Set the DesiredWeightOfInstancesPerInterval value.

Parameters:

newDesiredNumber - The new DesiredNumber value.

useEqualFrequencyTipText

public java.lang.String useEqualFrequencyTipText()

Returns the tip text for this property

Returns:

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

getUseEqualFrequency

public boolean getUseEqualFrequency()

Get the value of UseEqualFrequency.

Returns:

Value of UseEqualFrequency.

setUseEqualFrequency

public void setUseEqualFrequency(boolean newUseEqualFrequency)

Set the value of UseEqualFrequency.

Parameters:

newUseEqualFrequency - Value to assign to UseEqualFrequency.

binsTipText

public java.lang.String binsTipText()

Returns the tip text for this property

Returns:

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

getBins

public int getBins()

Gets the number of bins numeric attributes will be divided into

Returns:

the number of bins.

setBins

public void setBins(int numBins)

Sets the number of bins to divide each selected numeric attribute into

Parameters:

numBins - the number of bins

invertSelectionTipText

public java.lang.String invertSelectionTipText()

Returns the tip text for this property

Returns:

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

getInvertSelection

public boolean getInvertSelection()

Gets whether the supplied columns are to be removed or kept

Returns:

true if the supplied columns will be kept

setInvertSelection

public void setInvertSelection(boolean invert)

Sets whether selected columns should be removed or kept. If true the selected columns are kept and unselected columns are deleted. If false selected columns are deleted and unselected columns are kept.

Parameters:

invert - the new invert setting

attributeIndicesTipText

public java.lang.String attributeIndicesTipText()

Returns the tip text for this property

Returns:

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

getAttributeIndices

public java.lang.String getAttributeIndices()

Gets the current range selection

Returns:

a string containing a comma separated list of ranges

setAttributeIndices

public void setAttributeIndices(java.lang.String rangeList)

Sets which attributes are to be Discretized (only numeric attributes among the selection will be Discretized).

Parameters:

rangeList - a string representing the list of attributes. Since the string will typically come from a user, attributes are indexed from 1.
eg: first-3,5,6-last

Throws:

java.lang.IllegalArgumentException - if an invalid range list is supplied

setAttributeIndicesArray

public void setAttributeIndicesArray(int[] attributes)

Sets which attributes are to be Discretized (only numeric attributes among the selection will be Discretized).

Parameters:

attributes - an array containing indexes of attributes to Discretize. Since the array will typically come from a program, attributes are indexed from 0.

Throws:

java.lang.IllegalArgumentException - if an invalid set of ranges is supplied

getCutPoints

public double[] getCutPoints(int attributeIndex)

Gets the cut points for an attribute

Returns:

an array containing the cutpoints (or null if the attribute requested has been discretized into only one interval.)

calculateCutPoints

protected void calculateCutPoints()

Generate the cutpoints for each attribute

calculateCutPointsByEqualWidthBinning

protected void calculateCutPointsByEqualWidthBinning(int index)

Set cutpoints for a single attribute.

Parameters:

index - the index of the attribute to set cutpoints for

calculateCutPointsByEqualFrequencyBinning

protected void calculateCutPointsByEqualFrequencyBinning(int index)

Set cutpoints for a single attribute.

Parameters:

index - the index of the attribute to set cutpoints for

findNumBins

protected void findNumBins(int index)

Optimizes the number of bins using leave-one-out cross-validation.

Parameters:

index - the attribute index

setOutputFormat

protected void setOutputFormat()

Set the output format. Takes the currently defined cutpoints and m_InputFormat and calls setOutputFormat(Instances) appropriately.

convertInstance

protected void convertInstance(Instance instance)

Convert a single instance over. The converted instance is added to the end of the output queue.

Parameters:

instance - the instance to convert

main

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

Main method for testing this class.

Parameters:

argv - should contain arguments to the filter: use -h for help