com.pervasive.datarush.analytics.decisiontree.predictor

Class DecisionTreePredictor

java.lang.Object

com.pervasive.datarush.operators.AbstractLogicalOperator

com.pervasive.datarush.operators.StreamingOperator

com.pervasive.datarush.operators.ExecutableOperator

com.pervasive.datarush.analytics.util.AbstractPredictor

com.pervasive.datarush.analytics.decisiontree.predictor.DecisionTreePredictor

All Implemented Interfaces:

LogicalOperator

public final class DecisionTreePredictor
extends AbstractPredictor

Operator responsible for predicting outcomes based on a Decision Tree PMML model. Supports most of the functionality listed here. That is, a subset of the PMML specification, but a superset of the functionality required for C4.5. Specifically, supports all required elements/attributes as well as the following optional elements/attributes:

1. missingValueStategy ( all strategies supported )
2. missingValuePenalty
3. noTrueChildStategy ( all strategies supported )
4. All predicates: SimplePredicate, CompoundPredicate, SimpleSetPredicate, True, False
5. ScoreDistribution

Ignores the following elements/attributes:

1. EmbeddedModel
2. Partition
3. ModelStats
4. ModelExplanation
5. Targets
6. LocalTransformations
7. ModelVerification
8. splitCharacteristic

Constructor Summary

Constructors

Constructor and Description
DecisionTreePredictor() Creates a decision tree predictor with default settings.

Method Summary

Modifier and Type	Method and Description
protected void	execute(PMMLModel pmml, RecordValued input, ScalarSettable[] predictedFields) Called to perform prediction.
String	getConfidencePrefix() Gets the field name prefix to use for confidence.
RecordPort	getOutput() Returns a record port consisting of the original input plus predicted values.
String	getRecordCountPrefix() Gets the field name prefix to use for record counts.
String	getWinnerField() Gets the name of the winner field to output.
boolean	isAppendConfidence() Returns whether to append confidence information.
boolean	isAppendRecordCount() Returns whether to append record count information.
protected RecordTokenType	predictedType(PMMLModelSpec modelSpec) Given the model spec, returns the predicted type.
void	setAppendConfidence(boolean appendConfidence) Sets whether to append confidence information.
void	setAppendRecordCount(boolean appendRecordCount) Sets whether to append record count information.
void	setConfidencePrefix(String confidencePrefix) Sets the field name prefix to use for confidence.
void	setRecordCountPrefix(String recordCountPrefix) Sets the field name prefix to use for record counts.
void	setWinnerField(String winnerField) Sets the name of the winner field to output.

Methods inherited from class com.pervasive.datarush.analytics.util.AbstractPredictor

computeMetadata, execute, getInput, getModel, pushPrediction, stepNext

Methods inherited from class com.pervasive.datarush.operators.ExecutableOperator

cloneForExecution, getNumInputCopies, getPortSettings, handleInactiveOutput

Methods inherited from class com.pervasive.datarush.operators.AbstractLogicalOperator

disableParallelism, getInputPorts, getOutputPorts, newInput, newInput, newOutput, newRecordInput, newRecordInput, newRecordOutput, notifyError

Methods inherited from class java.lang.Object

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

Constructor Detail

DecisionTreePredictor

public DecisionTreePredictor()

Creates a decision tree predictor with default settings.

Method Detail

getOutput

public RecordPort getOutput()

Returns a record port consisting of the original input plus predicted values. This adds the following additional fields.

1. winner: The predicted value. Generally corresponds to the target with the highest record count; although PMML may specify an alternative winner via the "score" attribute. The name "winner" is the default; this is configurable via the property winnerField.
2. record_count_targetValue (optional): The record count of the named targetValue. The prefix "record_count_" is the default; this is configurable via the property recordCountPrefix.
3. confidence_targetValue (optional): The probability of the named targetValue. The prefix "confidence_" is the default; this is configurable via the property confidencePrefix.

Overrides:

getOutput in class AbstractPredictor

Returns:

a record flow of predicted values and their record counts

isAppendRecordCount

public boolean isAppendRecordCount()

Returns whether to append record count information. This is false by-default.

Returns:

whether to append record count information.

setAppendRecordCount

public void setAppendRecordCount(boolean appendRecordCount)

Sets whether to append record count information. This is false by-default.

Parameters:

appendRecordCount - whether to append record count information.

getWinnerField

public String getWinnerField()

Gets the name of the winner field to output. This is "winner" by-default.

Returns:

the name of the winner field to output.

setWinnerField

public void setWinnerField(String winnerField)

Sets the name of the winner field to output. This is "winner" by-default.

Parameters:

winnerField - the name of the winner field to output.

getRecordCountPrefix

public String getRecordCountPrefix()

Gets the field name prefix to use for record counts. This is "record_count_" by-default.

Returns:

the field name prefix to use for record counts.

setRecordCountPrefix

public void setRecordCountPrefix(String recordCountPrefix)

Sets the field name prefix to use for record counts. This is "record_count_" by-default.

Parameters:

recordCountPrefix - the field name prefix to use for record counts.

getConfidencePrefix

public String getConfidencePrefix()

Gets the field name prefix to use for confidence. This is "confidence_" by-default.

Returns:

the field name prefix to use for confidence.

setConfidencePrefix

public void setConfidencePrefix(String confidencePrefix)

Sets the field name prefix to use for confidence. This is "confidence_" by-default.

Parameters:

confidencePrefix - the field name prefix to use for confidence.

isAppendConfidence

public boolean isAppendConfidence()

Returns whether to append confidence information. This is false by-default.

Returns:

whether to append confidence information.

setAppendConfidence

public void setAppendConfidence(boolean appendConfidence)

Sets whether to append confidence information. This is false by-default.

Parameters:

appendConfidence - whether to append confidence information.

predictedType

protected RecordTokenType predictedType(PMMLModelSpec modelSpec)

Description copied from class: AbstractPredictor

Given the model spec, returns the predicted type. This should not include the input type ( the input is automatically prepended to the type that is returned )

Specified by:

predictedType in class AbstractPredictor

Parameters:

modelSpec - the model metadata

Returns:

the predicted type

execute

protected void execute(PMMLModel pmml,
                       RecordValued input,
                       ScalarSettable[] predictedFields)

Description copied from class: AbstractPredictor

Called to perform prediction. Subclasses are expected to loop over the input by calling AbstractPredictor.stepNext(). For each row of input, subclasses should first set the predicted values in the predictedFields array and then invoke AbstractPredictor.pushPrediction(). Subclasses should not invoke pushEndOfData since that is automatically handled by the base class.

Specified by:

execute in class AbstractPredictor

Parameters:

pmml - The input PMML model

input - The input data

predictedFields - An array of fields that reference the predicted field locations. The array positionally corresponds to the type returned by AbstractPredictor.predictedType(PMMLModelSpec).