Package weka.knowledgeflow.steps

Class TrainTestSplitMaker

java.lang.Object

weka.knowledgeflow.steps.BaseStep

weka.knowledgeflow.steps.TrainTestSplitMaker

All Implemented Interfaces:

java.io.Serializable, BaseStepExtender, Step

@KFStep(name="TrainTestSplitMaker",
        category="Evaluation",
        toolTipText="A step that randomly splits incoming data into a training and test set",
        iconPath="weka/gui/knowledgeflow/icons/TrainTestSplitMaker.gif")
public class TrainTestSplitMaker
extends BaseStep

A step that creates a random train/test split from an incoming data set.

Version:

$Revision: $

Author:

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

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor	Description
TrainTestSplitMaker()

Method Summary

Modifier and Type	Method	Description
java.util.List<java.lang.String>	getIncomingConnectionTypes()	Get a list of incoming connection types that this step can accept.
java.util.List<java.lang.String>	getOutgoingConnectionTypes()	Get a list of outgoing connection types that this step can produce.
boolean	getPreserveOrder()	Get whether to preserve the order of the instances or not
java.lang.String	getSeed()	Get the random seed to use
java.lang.String	getTrainPercent()	Get the training percentage
Instances	outputStructureForConnectionType(java.lang.String connectionName)	If possible, get the output structure for the named connection type as a header-only set of instances.
void	processIncoming(Data data)	Process an incoming data payload (if the step accepts incoming connections)
void	setPreserveOrder(boolean preserve)	Set whether to preserve the order of the instances or not
void	setSeed(java.lang.String seed)	Set the random seed to use
void	setTrainPercent(java.lang.String percent)	Set the training percentage
void	stepInit()	Initialize the step

Methods inherited from class weka.knowledgeflow.steps.BaseStep

environmentSubstitute, getCustomEditorForStep, getDefaultSettings, getInteractiveViewers, getInteractiveViewersImpls, getName, getStepManager, globalInfo, isResourceIntensive, isStopRequested, outputStructureForConnectionType, setName, setStepIsResourceIntensive, setStepManager, setStepMustRunSingleThreaded, start, stepMustRunSingleThreaded, stop

Methods inherited from class java.lang.Object

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

Constructor Detail

TrainTestSplitMaker

public TrainTestSplitMaker()

Method Detail

setTrainPercent

@OptionMetadata(displayName="Training percentage",
                description="The percentage of data to go into the training set",
                displayOrder=1)
public void setTrainPercent(java.lang.String percent)

Set the training percentage

Parameters:

percent - the training percentage

getTrainPercent

public java.lang.String getTrainPercent()

Get the training percentage

Returns:

the training percentage

setSeed

@OptionMetadata(displayName="Random seed",
                description="The random seed to use when shuffling the data",
                displayOrder=2)
public void setSeed(java.lang.String seed)

Set the random seed to use

Parameters:

seed - the random seed to use

getSeed

public java.lang.String getSeed()

Get the random seed to use

Returns:

the random seed to use

setPreserveOrder

@OptionMetadata(displayName="Preserve instance order",
                description="Preserve the order of the instances rather than randomly shuffling",
                displayOrder=3)
public void setPreserveOrder(boolean preserve)

Set whether to preserve the order of the instances or not

Parameters:

preserve - true to preserve the order rather than randomly shuffling first

getPreserveOrder

public boolean getPreserveOrder()

Get whether to preserve the order of the instances or not

Returns:

true to preserve the order rather than randomly shuffling first

stepInit

public void stepInit()
              throws WekaException

Initialize the step

Throws:

WekaException - if a problem occurs

processIncoming

public void processIncoming(Data data)
                     throws WekaException

Process an incoming data payload (if the step accepts incoming connections)

Specified by:

processIncoming in interface BaseStepExtender

Specified by:

processIncoming in interface Step

Overrides:

processIncoming in class BaseStep

Parameters:

data - the data to process

Throws:

WekaException - if a problem occurs

getIncomingConnectionTypes

public java.util.List<java.lang.String> getIncomingConnectionTypes()

Get a list of incoming connection types that this step can accept. Ideally (and if appropriate), this should take into account the state of the step and any existing incoming connections. E.g. a step might be able to accept one (and only one) incoming batch data connection.

Returns:

a list of incoming connections that this step can accept given its current state

getOutgoingConnectionTypes

public java.util.List<java.lang.String> getOutgoingConnectionTypes()

Get a list of outgoing connection types that this step can produce. Ideally (and if appropriate), this should take into account the state of the step and the incoming connections. E.g. depending on what incoming connection is present, a step might be able to produce a trainingSet output, a testSet output or neither, but not both.

Returns:

a list of outgoing connections that this step can produce

outputStructureForConnectionType

public Instances outputStructureForConnectionType(java.lang.String connectionName)
                                           throws WekaException

If possible, get the output structure for the named connection type as a header-only set of instances. Can return null if the specified connection type is not representable as Instances or cannot be determined at present.

Specified by:

outputStructureForConnectionType in interface Step

Overrides:

outputStructureForConnectionType in class BaseStep

Parameters:

connectionName - the name of the connection type to get the output structure for

Returns:

the output structure as a header-only Instances object

Throws:

WekaException - if a problem occurs