jsat.datatransform

Class RemoveAttributeTransform

java.lang.Object

jsat.datatransform.RemoveAttributeTransform

All Implemented Interfaces:

java.io.Serializable, java.lang.Cloneable, DataTransform

Direct Known Subclasses:

MutualInfoFS, ReliefF, SBS

public class RemoveAttributeTransform
extends java.lang.Object
implements DataTransform

This Data Transform allows the complete removal of specific features from the data set.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
RemoveAttributeTransform(DataSet dataSet, java.util.Set<java.lang.Integer> categoricalToRemove, java.util.Set<java.lang.Integer> numericalToRemove) Creates a new transform for removing specified features from a data set
RemoveAttributeTransform(java.util.Set<java.lang.Integer> categoricalToRemove, java.util.Set<java.lang.Integer> numericalToRemove) Creates a new transform for removing specified features from a data set.

Method Summary

Modifier and Type	Method and Description
RemoveAttributeTransform	clone()
void	consolidate(RemoveAttributeTransform preceding) A serious of Remove Attribute Transforms may be learned and applied sequentially to a single data set.
void	fit(DataSet data) Fits this transform to the given dataset.
java.util.List<java.lang.Integer>	getKeptNominal() Returns an unmodifiable list of the original indices of the nominal attributes that will be kept when this transform is applied.
java.util.List<java.lang.Integer>	getKeptNumeric() Returns an unmodifiable list of the original indices of the numeric attributes that will be kept when this transform is applied.
java.util.Map<java.lang.Integer,java.lang.Integer>	getReverseNominalMap() Returns a mapping from the nominal indices in the transformed space back to their original indices
java.util.Map<java.lang.Integer,java.lang.Integer>	getReverseNumericMap() Returns a mapping from the numeric indices in the transformed space back to their original indices
DataPoint	transform(DataPoint dp) Returns a new data point that is a transformation of the original data point.

Methods inherited from class java.lang.Object

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

Constructor Detail

RemoveAttributeTransform

public RemoveAttributeTransform(java.util.Set<java.lang.Integer> categoricalToRemove,
                                java.util.Set<java.lang.Integer> numericalToRemove)

Creates a new transform for removing specified features from a data set. Needs to still call fit(jsat.DataSet) before ready to be used.

Parameters:

categoricalToRemove - the set of categorical attributes to remove, in the rage of [0, DataSet.getNumCategoricalVars()).

numericalToRemove - the set of numerical attributes to remove, in the rage of [0, DataSet.getNumNumericalVars()).

RemoveAttributeTransform

public RemoveAttributeTransform(DataSet dataSet,
                                java.util.Set<java.lang.Integer> categoricalToRemove,
                                java.util.Set<java.lang.Integer> numericalToRemove)

Creates a new transform for removing specified features from a data set

Parameters:

dataSet - the data set that this transform is meant for

categoricalToRemove - the set of categorical attributes to remove, in the rage of [0, DataSet.getNumCategoricalVars()).

numericalToRemove - the set of numerical attributes to remove, in the rage of [0, DataSet.getNumNumericalVars()).

Method Detail

getKeptNumeric

public java.util.List<java.lang.Integer> getKeptNumeric()

Returns an unmodifiable list of the original indices of the numeric attributes that will be kept when this transform is applied.

Returns:

the numeric indices that are not removed by this transform

getReverseNumericMap

public java.util.Map<java.lang.Integer,java.lang.Integer> getReverseNumericMap()

Returns a mapping from the numeric indices in the transformed space back to their original indices

Returns:

a mapping from the transformed numeric space to the original one

getKeptNominal

public java.util.List<java.lang.Integer> getKeptNominal()

Returns an unmodifiable list of the original indices of the nominal attributes that will be kept when this transform is applied.

Returns:

the nominal indices that are not removed by this transform

getReverseNominalMap

public java.util.Map<java.lang.Integer,java.lang.Integer> getReverseNominalMap()

Returns a mapping from the nominal indices in the transformed space back to their original indices

Returns:

a mapping from the transformed nominal space to the original one

fit

public void fit(DataSet data)

Description copied from interface: DataTransform

Fits this transform to the given dataset. Some transforms can only be learned from classification or regression datasets. If an incompatible dataset type is given, a FailedToFitException exception may be thrown.

Specified by:

fit in interface DataTransform

Parameters:

data - the dataset to fir this transform to

consolidate

public void consolidate(RemoveAttributeTransform preceding)

A serious of Remove Attribute Transforms may be learned and applied sequentially to a single data set. Instead of keeping all the transforms around indefinitely, a sequential series of Remove Attribute Transforms can be consolidated into a single transform object.
This method mutates the this transform by providing it with the transform that would have been applied before this current object. Once complete, this transform can be used two perform both removals in one step.

Example:
An initial set of features A is transformed into A' by transform t₁
A' is transformed into A'' by transform t₂
Instead, you can invoke t₂.consolidate(t₁). You can then transform A into A'' by using only transform t₂

Parameters:

preceding - the DataTransform that immediately precedes this one in a sequential list of transforms

transform

public DataPoint transform(DataPoint dp)

Description copied from interface: DataTransform

Returns a new data point that is a transformation of the original data point. This new data point is a different object, but may contain the same references as the original data point. It is not guaranteed that you can mutate the transformed point without having a side effect on the original point.

Specified by:

transform in interface DataTransform

Parameters:

dp - the data point to apply a transformation to

Returns:

a transformed data point

clone

public RemoveAttributeTransform clone()

Specified by:

clone in interface DataTransform

Overrides:

clone in class java.lang.Object