edu.udo.cs.yale.example
Class AttributeWeights

java.lang.Object
  edu.udo.cs.yale.operator.AbstractIOObject
      edu.udo.cs.yale.operator.ResultObjectAdapter
          edu.udo.cs.yale.tools.math.AverageVector
              edu.udo.cs.yale.example.AttributeWeights

All Implemented Interfaces:

IOObject, ResultObject, Saveable, java.lang.Comparable

public class AttributeWeights
extends AverageVector
implements Saveable

AttributeWeights holds the information about the weights of attributes of an example set. It is delivered by several feature weighting algorithms or learning schemes. The use of a linked hash map ensures that the added features are stored in the same sequence they were added.

Version:

$Id: AttributeWeights.java,v 2.25 2006/04/05 08:57:22 ingomierswa Exp $

Author:

Ingo Mierswa

Nested Class Summary

private class

AttributeWeights.WeightComparator This comparator sorts the names of attributes according to their weights.

Field Summary

static int	ABSOLUTE_WEIGHTS Indicates that the the absolute weights should be used for sorting.
static int	DECREASING Indicates that the weights should be sorted in descending order.
static int	INCREASING Indicates that the weights should be sorted in ascending order.
static int	NO_SORTING Indicates that the weights should not be sorted at all.
static int	ORIGINAL_WEIGHTS Indicates that the the actual weights should be used for sorting.
private static java.lang.String[]	SORTING_ORDER_NAMES The names for the sorting orders.
private static java.lang.String[]	SORTING_TYPE_NAMES The names for the sorting value type.
private int	sortType Indicates the type of sorting.
private java.util.Map<java.lang.String,AttributeWeight>	weightMap Maps the name of an attribute to the corresponding attribute weight.
private int	weightType Indicates if absolute or actualy weights should be used for sorting.

Constructor Summary

	AttributeWeights() Creates a new empty attribute weights object.
private	AttributeWeights(AttributeWeights weights) Clone constructor.
	AttributeWeights(ExampleSet exampleSet) Creates a new attribute weights object containing a weight of 1 for each of the given input attributes.

Method Summary

java.lang.Object	clone() Returns a deep clone of the attribute weights which provides the same sequence of attribute names.
int	compareTo(java.lang.Object o) Since this average vector cannot be compared this method always returns 0.
private DataTable	createSortedDataTable()
java.util.Set<java.lang.String>	getAttributeNames() Returns an set of attribute names in this map ordered by their insertion time.
java.lang.String	getExtension() Returns the default extension in file choosers.
java.lang.String	getFileDescription() Returns the file description used in file choosers.
java.lang.String	getName() Returns the name of this AverageVector.
private java.util.Iterator<AttributeWeight>	getSortedWeights() Returns an iterator over all AttributeWeight objects according to the current sorting settings.
int	getSortingType() Returns the currently used sorting type.
java.awt.Component	getVisualisationComponent(IOContainer container) Returns a visualisation component which allows sorting of the attribute weights and several weight plots.
double	getWeight(java.lang.String name) Returns the weight for the attribute with the given name.
int	getWeightType() Returns the currently used weight type.
static AttributeWeights	load(java.io.File file) Loads a new AttributeWeights object from the given file.
void	save(java.io.File file) Saves the attribute weights into a file.
void	setWeight(java.lang.String name, double weight) Sets the weight for the attribute with the given name.
int	size() Returns the number of features in this map.
void	sortByWeight(java.lang.String[] attributeNames, int direction, int value) Sorts the given array of attribute names according to their weight, the sorting direction (ascending or descending), and with respect to the fact if original or absolute weights should be used.
java.lang.String	toString() Returns a string representation of this object.

Methods inherited from class edu.udo.cs.yale.tools.math.AverageVector

addAveragable, buildAverages, getAveragable, getAveragable, getSize, removeAveragable, toHTML, toResultString

Methods inherited from class edu.udo.cs.yale.operator.ResultObjectAdapter

addAction, getActions, toHTML

Methods inherited from class edu.udo.cs.yale.operator.AbstractIOObject

copy, read, read, read, write, write

Methods inherited from class java.lang.Object

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

Methods inherited from interface edu.udo.cs.yale.operator.IOObject

copy, write, write

Field Detail

NO_SORTING

public static final int NO_SORTING

Indicates that the weights should not be sorted at all.

See Also:

Constant Field Values

DECREASING

public static final int DECREASING

Indicates that the weights should be sorted in descending order.

See Also:

Constant Field Values

INCREASING

public static final int INCREASING

Indicates that the weights should be sorted in ascending order.

See Also:

Constant Field Values

SORTING_ORDER_NAMES

private static final java.lang.String[] SORTING_ORDER_NAMES

The names for the sorting orders.

ORIGINAL_WEIGHTS

public static final int ORIGINAL_WEIGHTS

Indicates that the the actual weights should be used for sorting.

See Also:

Constant Field Values

ABSOLUTE_WEIGHTS

public static final int ABSOLUTE_WEIGHTS

Indicates that the the absolute weights should be used for sorting.

See Also:

Constant Field Values

SORTING_TYPE_NAMES

private static final java.lang.String[] SORTING_TYPE_NAMES

The names for the sorting value type.

sortType

private int sortType

Indicates the type of sorting.

weightType

private int weightType

Indicates if absolute or actualy weights should be used for sorting.

weightMap

private java.util.Map<java.lang.String,AttributeWeight> weightMap

Maps the name of an attribute to the corresponding attribute weight.

Constructor Detail

AttributeWeights

public AttributeWeights()

Creates a new empty attribute weights object.

AttributeWeights

public AttributeWeights(ExampleSet exampleSet)

Creates a new attribute weights object containing a weight of 1 for each of the given input attributes.

AttributeWeights

private AttributeWeights(AttributeWeights weights)

Clone constructor.

Method Detail

getName

public java.lang.String getName()

Returns the name of this AverageVector.

Specified by:

getName in interface ResultObject

Overrides:

getName in class ResultObjectAdapter

setWeight

public void setWeight(java.lang.String name,
                      double weight)

Sets the weight for the attribute with the given name.

getWeight

public double getWeight(java.lang.String name)

Returns the weight for the attribute with the given name. Returns Double.NaN if the weight for the queried attribute is not known.

getWeightType

public int getWeightType()

Returns the currently used weight type.

getSortingType

public int getSortingType()

Returns the currently used sorting type.

size

public int size()

Returns the number of features in this map.

Overrides:

size in class AverageVector

getAttributeNames

public java.util.Set<java.lang.String> getAttributeNames()

Returns an set of attribute names in this map ordered by their insertion time.

compareTo

public int compareTo(java.lang.Object o)

Since this average vector cannot be compared this method always returns 0.

Specified by:

compareTo in interface java.lang.Comparable

sortByWeight

public void sortByWeight(java.lang.String[] attributeNames,
                         int direction,
                         int value)

Sorts the given array of attribute names according to their weight, the sorting direction (ascending or descending), and with respect to the fact if original or absolute weights should be used.

Parameters:

direction - ASCENDING or DESCENDING

value - WEIGHT or WEIGHT_ABSOLUTE.

save

public void save(java.io.File file)
          throws java.io.IOException

Saves the attribute weights into a file.

Specified by:

save in interface Saveable

Throws:

java.io.IOException

load

public static AttributeWeights load(java.io.File file)
                             throws java.io.IOException

Loads a new AttributeWeights object from the given file.

Throws:

java.io.IOException

getExtension

public java.lang.String getExtension()

Description copied from interface: Saveable

Returns the default extension in file choosers.

Specified by:

getExtension in interface Saveable

getFileDescription

public java.lang.String getFileDescription()

Description copied from interface: Saveable

Returns the file description used in file choosers.

Specified by:

getFileDescription in interface Saveable

toString

public java.lang.String toString()

Returns a string representation of this object.

Overrides:

toString in class AverageVector

clone

public java.lang.Object clone()

Returns a deep clone of the attribute weights which provides the same sequence of attribute names.

Specified by:

clone in class AverageVector

getVisualisationComponent

public java.awt.Component getVisualisationComponent(IOContainer container)

Returns a visualisation component which allows sorting of the attribute weights and several weight plots.

Specified by:

getVisualisationComponent in interface ResultObject

Overrides:

getVisualisationComponent in class AverageVector

createSortedDataTable

private DataTable createSortedDataTable()

getSortedWeights

private java.util.Iterator<AttributeWeight> getSortedWeights()

Returns an iterator over all AttributeWeight objects according to the current sorting settings.