weka.filters.unsupervised.attribute

Class PreConstructedPCA

java.lang.Object

weka.filters.Filter

weka.filters.unsupervised.attribute.PreConstructedPCA

All Implemented Interfaces:

java.io.Serializable, weka.core.CapabilitiesHandler, weka.core.CapabilitiesIgnorer, weka.core.CommandlineRunnable, weka.core.OptionHandler, Preconstructed, weka.core.RevisionHandler, PreconstructedFilter, weka.filters.StreamableFilter

@KFIgnore
 @GPCIgnore
public class PreConstructedPCA
extends weka.filters.Filter
implements weka.filters.StreamableFilter, java.io.Serializable, PreconstructedFilter, weka.core.OptionHandler, weka.core.CommandlineRunnable

Performs a principal components analysis and transformation of the data.
Dimensionality reduction is accomplished by choosing enough eigenvectors to account for some percentage of the variance in the original data -- default 0.95 (95%).
Based on code of the attribute selection scheme 'PrincipalComponents' by Mark Hall and Gabi Schmidberger.

Valid options are:

 -header <path to ARFF header>
  Path to the ARFF header used when the matrix
  was constructed. Must contain summary attributes.
 

 -matrix <path to matrix file>
  Path to the correlation/covariance matrix.
 

 -covariance
  Matrix is a covariance matrix (rather than correlation).
 

 -keep-class
  Keep the class (if set). Set this if the
  class was retained when computing the matrix (i.e. there is a column
  in the matrix corresponding to the class).
 

 -R <num>
  Retain enough PC attributes to account
  for this proportion of variance in the original data.
  (default: 0.95)
 

 -A <num>
  Maximum number of attributes to include in 
  transformed attribute names.
  (-1 = include all, default: 5)
 

 -M <num>
  Maximum number of PC attributes to retain.
  (-1 = include all, default: -1)
 

Version:

$Revision: 11378 $

Author:

Mark Hall (mhall{[at]}pentaho{[dot]}com)

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
static double	DEFAULT_VARIANCE_COVERED Default proportion of variance covered by PCs
static int	MAX_ATTRIB_NAMES Default number of original attribute names to include in the transformed attribute names

Constructor Summary

Constructors

Constructor and Description
PreConstructedPCA() Default constructor.
PreConstructedPCA(weka.core.Instances headerWithSummaryAtts, weka.core.matrix.Matrix matrix, boolean keepClassIfSet, boolean isCovariance) Construct a new PreConstructedPCA.
PreConstructedPCA(weka.core.Instances header, weka.core.matrix.Matrix matrix, java.util.List<NumericStats> stats, boolean keepClassIfSet, boolean isCovariance) Construct a new PreConstructedPCA.

Method Summary

Modifier and Type	Method and Description
weka.core.Capabilities	getCapabilities() Returns the capabilities of this evaluator.
boolean	getKeepClassIfSet() Get whether the class should be kept
boolean	getMatrixIsCovariance() Get whether the matrix is a covariance rather than correlation one
int	getMaximumAttributeNames() Gets maximum number of attributes to include in transformed attribute names.
int	getMaximumAttributes() Gets maximum number of PC attributes to retain.
java.lang.String[]	getOptions() Gets the current settings of the filter.
java.lang.String	getPathToHeaderWihtSummaryAtts() Get the path to the ARFF header (including summary attributes) used when the matrix was constructed.
java.lang.String	getPathToMatrix() Get the path to the correlation/covariance matrix
double	getVarianceCovered() Gets the proportion of total variance to account for when retaining principal components.
java.lang.String	globalInfo() Returns a string describing this filter.
boolean	input(weka.core.Instance instance)
boolean	isConstructed() Returns true if this Preconstructed instance is initialized and ready to be used
java.lang.String	keepClassIfSet() Tip text for this property
java.util.Enumeration<weka.core.Option>	listOptions() Returns an enumeration describing the available options.
static void	main(java.lang.String[] args) Main method for command line execution
java.lang.String	matrixIsCovarianceTipText() Tip text for this property
java.lang.String	maximumAttributeNamesTipText() Returns the tip text for this property.
java.lang.String	maximumAttributesTipText() Returns the tip text for this property.
java.lang.String	pathToHeaderWithSummaryAttsTipText() Tip text for this property
java.lang.String	pathToMatrixTipText() Tip text for this property
void	resetPreconstructed() Reset.
void	run(java.lang.Object toRun, java.lang.String[] options)
boolean	setInputFormat(weka.core.Instances instancesInfo) Sets the format of the input instances.
void	setKeepClassIfSet(boolean keep) Set whether the class should be kept
void	setMatrixIsCovariance(boolean c) Set whether the matrix is a covariance rather than correlation one
void	setMaximumAttributeNames(int value) Sets maximum number of attributes to include in transformed attribute names.
void	setMaximumAttributes(int value) Sets maximum number of PC attributes to retain.
void	setOptions(java.lang.String[] options) Parses a list of options for this object.
void	setPathToHeaderWithSummaryAtts(java.lang.String path) Set the path to the ARFF header (including summary attributes) used when the matrix was constructed.
void	setPathToMatrix(java.lang.String path) Set the path to the correlation/covariance matrix
void	setVarianceCovered(double value) Sets the amount of variance to account for when retaining principal components.
static void	test(java.lang.String[] args) Method for testing this class
java.lang.String	toString()
java.lang.String	varianceCoveredTipText() Returns the tip text for this property.

Methods inherited from class weka.filters.Filter

batchFilterFile, batchFinished, debugTipText, doNotCheckCapabilitiesTipText, filterFile, getCapabilities, getCopyOfInputFormat, getDebug, getDoNotCheckCapabilities, getOutputFormat, getRevision, isFirstBatchDone, isNewBatch, isOutputFormatDefined, makeCopies, makeCopy, mayRemoveInstanceAfterFirstBatchDone, numPendingOutput, output, outputPeek, postExecution, preExecution, runFilter, setDebug, setDoNotCheckCapabilities, useFilter, wekaStaticWrapper

Methods inherited from class java.lang.Object

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

Methods inherited from interface weka.core.OptionHandler

makeCopy

Methods inherited from interface weka.core.CommandlineRunnable

postExecution, preExecution

Field Detail

MAX_ATTRIB_NAMES

public static final int MAX_ATTRIB_NAMES

Default number of original attribute names to include in the transformed attribute names

See Also:

Constant Field Values

DEFAULT_VARIANCE_COVERED

public static final double DEFAULT_VARIANCE_COVERED

Default proportion of variance covered by PCs

See Also:

Constant Field Values

Constructor Detail

PreConstructedPCA

public PreConstructedPCA()

Default constructor. All settings, including matrix and header, must be set via mutator methods before use

PreConstructedPCA

public PreConstructedPCA(weka.core.Instances headerWithSummaryAtts,
                         weka.core.matrix.Matrix matrix,
                         boolean keepClassIfSet,
                         boolean isCovariance)
                  throws java.lang.Exception

Construct a new PreConstructedPCA.

Parameters:

headerWithSummaryAtts - header of the data used to construct the matrix from. Must contain the meta summary attributes.

matrix - the correlation or covariance matrix

keepClassIfSet - true if the class attribute was retained when computing the matrix (i.e. there is a column in the matrix for the class)

isCovariance - true if the matrix is a covariance matrix rather than correlation matrix

Throws:

java.lang.Exception - if a problem occurs

PreConstructedPCA

public PreConstructedPCA(weka.core.Instances header,
                         weka.core.matrix.Matrix matrix,
                         java.util.List<NumericStats> stats,
                         boolean keepClassIfSet,
                         boolean isCovariance)
                  throws java.lang.Exception

Construct a new PreConstructedPCA.

Parameters:

header - header of the data used to construct the matrix from. Should NOT contain any meta summary attributes.

matrix - the correlation or covariance matrix

stats - a list of NumericStats objects corresponding to the attributes in the header

keepClassIfSet - true if the class attribute was retained when computing the matrix (i.e. there is a column in the matrix for the class)

isCovariance - true if the matrix is a covariance matrix rather than correlation matrix

Throws:

java.lang.Exception - if a problem occurs

Method Detail

getCapabilities

public weka.core.Capabilities getCapabilities()

Returns the capabilities of this evaluator.

Specified by:

getCapabilities in interface weka.core.CapabilitiesHandler

Overrides:

getCapabilities in class weka.filters.Filter

Returns:

the capabilities of this evaluator

See Also:

Capabilities

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

listOptions

public java.util.Enumeration<weka.core.Option> listOptions()

Returns an enumeration describing the available options.

Specified by:

listOptions in interface weka.core.OptionHandler

Overrides:

listOptions in class weka.filters.Filter

Returns:

an enumeration of all the available options.

setOptions

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

Parses a list of options for this object.

Valid options are:

 -header <path to ARFF header>
  Path to the ARFF header used when the matrix
  was constructed. Must contain summary attributes.
 

 -matrix <path to matrix file>
  Path to the correlation/covariance matrix.
 

 -covariance
  Matrix is a covariance matrix (rather than correlation).
 

 -keep-class
  Keep the class (if set). Set this if the
  class was retained when computing the matrix (i.e. there is a column
  in the matrix corresponding to the class).
 

 -R <num>
  Retain enough PC attributes to account
  for this proportion of variance in the original data.
  (default: 0.95)
 

 -A <num>
  Maximum number of attributes to include in 
  transformed attribute names.
  (-1 = include all, default: 5)
 

 -M <num>
  Maximum number of PC attributes to retain.
  (-1 = include all, default: -1)
 

Specified by:

setOptions in interface weka.core.OptionHandler

Overrides:

setOptions in class weka.filters.Filter

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 weka.core.OptionHandler

Overrides:

getOptions in class weka.filters.Filter

Returns:

an array of strings suitable for passing to setOptions

pathToMatrixTipText

public java.lang.String pathToMatrixTipText()

Tip text for this property

Returns:

the tip text for this property

setPathToMatrix

public void setPathToMatrix(java.lang.String path)

Set the path to the correlation/covariance matrix

Parameters:

path - path to the matrix to use

getPathToMatrix

public java.lang.String getPathToMatrix()

Get the path to the correlation/covariance matrix

Returns:

path to the matrix to use

pathToHeaderWithSummaryAttsTipText

public java.lang.String pathToHeaderWithSummaryAttsTipText()

Tip text for this property

Returns:

the tip text for this property

setPathToHeaderWithSummaryAtts

public void setPathToHeaderWithSummaryAtts(java.lang.String path)

Set the path to the ARFF header (including summary attributes) used when the matrix was constructed.

Parameters:

path - the path to the ARFF header.

getPathToHeaderWihtSummaryAtts

public java.lang.String getPathToHeaderWihtSummaryAtts()

Get the path to the ARFF header (including summary attributes) used when the matrix was constructed.

Returns:

the path to the ARFF header.

keepClassIfSet

public java.lang.String keepClassIfSet()

Tip text for this property

Returns:

the tip text for this property

setKeepClassIfSet

public void setKeepClassIfSet(boolean keep)

Set whether the class should be kept

Parameters:

keep - true if the class is to be kept

getKeepClassIfSet

public boolean getKeepClassIfSet()

Get whether the class should be kept

Returns:

true if the class is to be kept

matrixIsCovarianceTipText

public java.lang.String matrixIsCovarianceTipText()

Tip text for this property

Returns:

the tip text for this property

setMatrixIsCovariance

public void setMatrixIsCovariance(boolean c)

Set whether the matrix is a covariance rather than correlation one

Parameters:

c - true if the matrix is a covariance matrix

getMatrixIsCovariance

public boolean getMatrixIsCovariance()

Get whether the matrix is a covariance rather than correlation one

Returns:

true if the matrix is a covariance matrix

varianceCoveredTipText

public java.lang.String varianceCoveredTipText()

Returns the tip text for this property.

Returns:

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

setVarianceCovered

public void setVarianceCovered(double value)

Sets the amount of variance to account for when retaining principal components.

Parameters:

value - the proportion of total variance to account for

getVarianceCovered

public double getVarianceCovered()

Gets the proportion of total variance to account for when retaining principal components.

Returns:

the proportion of variance to account for

maximumAttributeNamesTipText

public java.lang.String maximumAttributeNamesTipText()

Returns the tip text for this property.

Returns:

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

setMaximumAttributeNames

public void setMaximumAttributeNames(int value)

Sets maximum number of attributes to include in transformed attribute names.

Parameters:

value - the maximum number of attributes

getMaximumAttributeNames

public int getMaximumAttributeNames()

Gets maximum number of attributes to include in transformed attribute names.

Returns:

the maximum number of attributes

maximumAttributesTipText

public java.lang.String maximumAttributesTipText()

Returns the tip text for this property.

Returns:

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

setMaximumAttributes

public void setMaximumAttributes(int value)

Sets maximum number of PC attributes to retain.

Parameters:

value - the maximum number of attributes

getMaximumAttributes

public int getMaximumAttributes()

Gets maximum number of PC attributes to retain.

Returns:

the maximum number of attributes

setInputFormat

public boolean setInputFormat(weka.core.Instances instancesInfo)
                       throws java.lang.Exception

Sets the format of the input instances.

Overrides:

setInputFormat in class weka.filters.Filter

Parameters:

instancesInfo - 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(weka.core.Instance instance)
              throws java.lang.Exception

Overrides:

input in class weka.filters.Filter

Throws:

java.lang.Exception

toString

public java.lang.String toString()

Overrides:

toString in class weka.filters.Filter

test

public static void test(java.lang.String[] args)

Method for testing this class

Parameters:

args - arguments to the filter

main

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

Main method for command line execution

Parameters:

args - options to the filter

run

public void run(java.lang.Object toRun,
                java.lang.String[] options)

Specified by:

run in interface weka.core.CommandlineRunnable

Overrides:

run in class weka.filters.Filter

isConstructed

public boolean isConstructed()

Description copied from interface: Preconstructed

Returns true if this Preconstructed instance is initialized and ready to be used

Specified by:

isConstructed in interface Preconstructed

Returns:

true if this instance is initialized and ready to be used

resetPreconstructed

public void resetPreconstructed()

Description copied from interface: Preconstructed

Reset. After calling reset() a call to isConstructed() should return false.

Specified by:

resetPreconstructed in interface Preconstructed