org.apache.ignite.ml.selection.scoring.evaluator.aggregator

Class ClassificationMetricsAggregator<L extends Serializable>

java.lang.Object

org.apache.ignite.ml.selection.scoring.evaluator.aggregator.ClassificationMetricsAggregator<L>

Type Parameters:

L - Type of label.

All Implemented Interfaces:

Serializable, MetricStatsAggregator<L,EmptyContext<L>,ClassificationMetricsAggregator<L>>

public class ClassificationMetricsAggregator<L extends Serializable>
extends Object
implements MetricStatsAggregator<L,EmptyContext<L>,ClassificationMetricsAggregator<L>>

Class represents aggregations for classification metric (including multi-classification case).

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
ClassificationMetricsAggregator() Creates an instance of ClassificationMetricsAggregator.
ClassificationMetricsAggregator(long validAnswersCnt, long totalNumOfExamples) Creates an instance of ClassificationMetricsAggregator.

Method Summary

Modifier and Type	Method and Description
void	aggregate(IgniteModel<Vector,L> mdl, LabeledVector<L> vector) Aggregates statistics for metric computation given model and vector with answer.
EmptyContext<L>	createInitializedContext() Returns initialized context.
long	getTotalNumberOfExamples() Returns totalNumberOfExamples.
long	getValidAnswersCount() Returns validAnswersCount.
void	initByContext(EmptyContext<L> ctx) Inits this aggtegator by evaluation context.
ClassificationMetricsAggregator<L>	mergeWith(ClassificationMetricsAggregator<L> other) Merges statistics of two aggregators to new aggreagator.

Methods inherited from class java.lang.Object

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

Constructor Detail

ClassificationMetricsAggregator

public ClassificationMetricsAggregator()

Creates an instance of ClassificationMetricsAggregator.

ClassificationMetricsAggregator

public ClassificationMetricsAggregator(long validAnswersCnt,
                                       long totalNumOfExamples)

Creates an instance of ClassificationMetricsAggregator.

Parameters:

validAnswersCnt - Valid answers count.

totalNumOfExamples - Total number of examples.

Method Detail

aggregate

public void aggregate(IgniteModel<Vector,L> mdl,
                      LabeledVector<L> vector)

Aggregates statistics for metric computation given model and vector with answer.

Specified by:

aggregate in interface MetricStatsAggregator<L extends Serializable,EmptyContext<L extends Serializable>,ClassificationMetricsAggregator<L extends Serializable>>

Parameters:

mdl - Model.

vector - Vector.

mergeWith

public ClassificationMetricsAggregator<L> mergeWith(ClassificationMetricsAggregator<L> other)

Merges statistics of two aggregators to new aggreagator.

Specified by:

mergeWith in interface MetricStatsAggregator<L extends Serializable,EmptyContext<L extends Serializable>,ClassificationMetricsAggregator<L extends Serializable>>

Parameters:

other - Other aggregator.

Returns:

New aggregator.

createInitializedContext

public EmptyContext<L> createInitializedContext()

Returns initialized context.

Specified by:

createInitializedContext in interface MetricStatsAggregator<L extends Serializable,EmptyContext<L extends Serializable>,ClassificationMetricsAggregator<L extends Serializable>>

Returns:

initialized evaluation context.

initByContext

public void initByContext(EmptyContext<L> ctx)

Inits this aggtegator by evaluation context.

Specified by:

initByContext in interface MetricStatsAggregator<L extends Serializable,EmptyContext<L extends Serializable>,ClassificationMetricsAggregator<L extends Serializable>>

Parameters:

ctx - Evaluation context.

getValidAnswersCount

public long getValidAnswersCount()

Returns validAnswersCount.

Returns:

validAnswersCount.

getTotalNumberOfExamples

public long getTotalNumberOfExamples()

Returns totalNumberOfExamples.

Returns:

totalNumberOfExamples.