Package com.bayesserver.data.discovery

Class Clustering

java.lang.Object

com.bayesserver.data.discovery.Clustering

All Implemented Interfaces:

Discretize

public final class Clustering
extends Object
implements Discretize

Discretizes continuous data in bins, using a probabilistic clustering algorithm.

Constructor Summary

Constructors

Constructor	Description
Clustering()

Method Summary

Modifier and Type	Method	Description
List<DiscretizationInfo>	discretize(DataReaderCommand dataReaderCommand, List<DiscretizationColumn> dataColumns, DiscretizationAlgoOptions options)	Discretizes one or more data columns, that may contain missing (null) values.
List<Interval<Double>>	discretize(Iterable<Double> unsortedData, DiscretizationOptions options, String dataColumn)	Discretizes unsorted continuous data that may contain missing (null) values.
List<Interval<Double>>	discretizeWeighted(Iterable<WeightedValue> unsortedData, DiscretizationOptions options, String dataColumn)	Discretizes unsorted weighted continuous data that may contain missing (null) values.
DiscretizeProgress	getProgress()	Gets an instance that receive progress notifications.
void	setProgress(DiscretizeProgress value)	Gets an instance that receive progress notifications.

Methods inherited from class java.lang.Object

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

Constructor Detail

Clustering

public Clustering()

Method Detail

getProgress

public DiscretizeProgress getProgress()

Gets an instance that receive progress notifications.

Specified by:

getProgress in interface Discretize

setProgress

public void setProgress(DiscretizeProgress value)

Gets an instance that receive progress notifications.

Specified by:

setProgress in interface Discretize

discretize

public List<Interval<Double>> discretize(Iterable<Double> unsortedData,
                                         DiscretizationOptions options,
                                         String dataColumn)

Discretizes unsorted continuous data that may contain missing (null) values.

Specified by:

discretize in interface Discretize

Parameters:

unsortedData - The data to discretize.

options - Options that affect how discretization is performed, such as the algorithm to use.

dataColumn - The name of the source column. This is only used for error reporting.

Returns:

A number of bins each identified by an interval.

discretizeWeighted

public List<Interval<Double>> discretizeWeighted(Iterable<WeightedValue> unsortedData,
                                                 DiscretizationOptions options,
                                                 String dataColumn)

Discretizes unsorted weighted continuous data that may contain missing (null) values.

Specified by:

discretizeWeighted in interface Discretize

Parameters:

unsortedData - The weighted data to discretize.

options - Options that affect how discretization is performed, such as the algorithm to use.

dataColumn - The name of the source column. This is only used for error reporting.

Returns:

A number of bins each identified by an interval.

discretize

public List<DiscretizationInfo> discretize(DataReaderCommand dataReaderCommand,
                                           List<DiscretizationColumn> dataColumns,
                                           DiscretizationAlgoOptions options)

Discretizes one or more data columns, that may contain missing (null) values.

Specified by:

discretize in interface Discretize

Parameters:

dataReaderCommand - The data reader command to allow iteration of data.

dataColumns - The data columns that should be discretized and options per column.

options - Options governing the overall discretization algorithm. Each data column also has options.

Returns:

A number of bins each identified by an interval for each data column.