Package org.tribuo.classification.evaluation

Class LabelEvaluator

java.lang.Object

org.tribuo.evaluation.AbstractEvaluator<Label,LabelMetric.Context,LabelEvaluation,LabelMetric>

org.tribuo.classification.evaluation.LabelEvaluator

All Implemented Interfaces:

Evaluator<Label,LabelEvaluation>

public final class LabelEvaluator
extends AbstractEvaluator<Label,LabelMetric.Context,LabelEvaluation,LabelMetric>

An Evaluator for Labels.

The default set of metrics is taken from LabelMetrics. If the supplied model generates probabilities, then it also calculates LabelMetrics.AUCROC and LabelMetrics.AVERAGED_PRECISION.

If the dataset contains an unknown Label (as generated by LabelFactory.getUnknownOutput()) or a valid Label which is outside of the domain of the Model then the evaluate methods will throw IllegalArgumentException with an appropriate message.

Constructor Summary

Constructors

Constructor	Description
LabelEvaluator()

Method Summary

Modifier and Type	Method	Description
protected LabelMetric.Context	createContext(Model<Label> model, List<Prediction<Label>> predictions)	Create the context needed for evaluation.
protected LabelEvaluation	createEvaluation(LabelMetric.Context ctx, Map<MetricID<Label>,Double> results, EvaluationProvenance provenance)	Create an evaluation for the given results
protected Set<LabelMetric>	createMetrics(Model<Label> model)	Creates the appropriate set of metrics for this model, by querying for it's OutputInfo.

Methods inherited from class org.tribuo.evaluation.AbstractEvaluator

computeResults, evaluate, evaluate, evaluate

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.tribuo.evaluation.Evaluator

createOnlineEvaluator, evaluate

Constructor Details

LabelEvaluator

public LabelEvaluator()

Method Details

createMetrics

protected Set<LabelMetric> createMetrics(Model<Label> model)

Description copied from class: AbstractEvaluator

Creates the appropriate set of metrics for this model, by querying for it's OutputInfo.

Specified by:

createMetrics in class AbstractEvaluator<Label,LabelMetric.Context,LabelEvaluation,LabelMetric>

Parameters:

model - The model to inspect.

Returns:

The set of metrics.

createContext

protected LabelMetric.Context createContext(Model<Label> model,
 List<Prediction<Label>> predictions)

Description copied from class: AbstractEvaluator

Create the context needed for evaluation. The context might store global properties or cache computation.

Specified by:

createContext in class AbstractEvaluator<Label,LabelMetric.Context,LabelEvaluation,LabelMetric>

Parameters:

model - the model that will be evaluated

predictions - the predictions that will be evaluated

Returns:

the context for this model and its predictions

createEvaluation

protected LabelEvaluation createEvaluation(LabelMetric.Context ctx,
 Map<MetricID<Label>,Double> results,
 EvaluationProvenance provenance)

Description copied from class: AbstractEvaluator

Create an evaluation for the given results

Specified by:

createEvaluation in class AbstractEvaluator<Label,LabelMetric.Context,LabelEvaluation,LabelMetric>

Parameters:

ctx - the context that was used to compute these results

results - the results

provenance - the provenance of the results (including information about the model and dataset)

Returns:

the evaluation