jsat.datatransform

Class NumericalToHistogram

java.lang.Object

jsat.datatransform.NumericalToHistogram

All Implemented Interfaces:

Serializable, Cloneable, DataTransform

public class NumericalToHistogram
extends Object
implements DataTransform

This transform converts numerical features into categorical ones via a simple histogram. Bins will be created for each numeric feature of equal sizes. Each numeric feature will be converted to the same number of bins.
This transform will handle missing values by simply ignoring them, and leaving the value missing in the transformed categorical variable.

Author:

Edward Raff

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
NumericalToHistogram() Creates a new transform which will use at most 25 bins when converting numeric features.
NumericalToHistogram(DataSet dataSet) Creates a new transform which will use O(sqrt(n)) bins for each numeric feature, where n is the number of data points in the dataset.
NumericalToHistogram(DataSet dataSet, int n) Creates a new transform which will use the specified number of bins for each numeric feature.
NumericalToHistogram(int n) Creates a new transform which will use at most the specified number of bins

Method Summary

Modifier and Type	Method and Description
NumericalToHistogram	clone()
void	fit(DataSet dataSet) Fits this transform to the given dataset.
int	getNumberOfBins()
static Distribution	guessNumberOfBins(DataSet data) Attempts to guess the number of bins to use
void	setNumberOfBins(int n) Sets the maximum number of histogram bins to use when creating the categorical version of numeric features.
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, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Constructor Detail

NumericalToHistogram

public NumericalToHistogram()

Creates a new transform which will use at most 25 bins when converting numeric features. This may not be optimal for any given dataset

NumericalToHistogram

public NumericalToHistogram(DataSet dataSet)

Creates a new transform which will use O(sqrt(n)) bins for each numeric feature, where n is the number of data points in the dataset.

Parameters:

dataSet - the data set to create the transform from

NumericalToHistogram

public NumericalToHistogram(int n)

Creates a new transform which will use at most the specified number of bins

Parameters:

n - the number of bins to create

NumericalToHistogram

public NumericalToHistogram(DataSet dataSet,
                            int n)

Creates a new transform which will use the specified number of bins for each numeric feature.

Parameters:

dataSet - the data set to create the transform from

n - the number of bins to create

Method Detail

fit

public void fit(DataSet dataSet)

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:

dataSet - the dataset to fir this transform to

setNumberOfBins

public void setNumberOfBins(int n)

Sets the maximum number of histogram bins to use when creating the categorical version of numeric features.

Parameters:

n - the number of bins to create

getNumberOfBins

public int getNumberOfBins()

Returns:

the maximum number of bins to create

guessNumberOfBins

public static Distribution guessNumberOfBins(DataSet data)

Attempts to guess the number of bins to use

Parameters:

data - the dataset to be transforms

Returns:

a distribution of the guess

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 NumericalToHistogram clone()

Specified by:

clone in interface DataTransform

Overrides:

clone in class Object