Package org.tribuo.classification.evaluation

Enum LabelMetrics

java.lang.Object

java.lang.Enum<LabelMetrics>

org.tribuo.classification.evaluation.LabelMetrics

All Implemented Interfaces:

Serializable, Comparable<LabelMetrics>

public enum LabelMetrics
extends Enum<LabelMetrics>

An enum of the default LabelMetrics supported by the multi-class classification evaluation package.

Enum Constant Summary

Enum Constants

Enum Constant	Description
ACCURACY	The accuracy.
AUCROC	The area under the receiver-operator curve (ROC).
AVERAGED_PRECISION	The averaged precision.
BALANCED_ERROR_RATE	The balanced error rate, i.e., the mean of the per class recalls.
F1	The F_1 score, i.e., the harmonic mean of the precision and the recall.
FN	The number of false negatives.
FP	The number of false positives.
PRECISION	The precision, i.e., the number of true positives divided by the number of predicted positives.
RECALL	The recall, i.e., the number of true positives divided by the number of ground truth positives.
TN	The number of true negatives.
TP	The number of true positives.

Method Summary

Modifier and Type	Method	Description
static double	AUCROC(Label label, List<Prediction<Label>> predictions)	Area under the ROC curve.
static double	AUCROC(MetricTarget<Label> tgt, List<Prediction<Label>> predictions)	Area under the ROC curve.
static double	averagedPrecision(Label label, List<Prediction<Label>> predictions)
static double	averagedPrecision(MetricTarget<Label> tgt, List<Prediction<Label>> predictions)
LabelMetric	forTarget(MetricTarget<Label> tgt)	Gets the LabelMetric wrapped around the supplied MetricTarget.
ToDoubleBiFunction<MetricTarget<Label>,LabelMetric.Context>	getImpl()	Returns the implementing function for this metric.
static LabelEvaluationUtil.PRCurve	precisionRecallCurve(Label label, List<Prediction<Label>> predictions)
static LabelMetrics	valueOf(String name)	Returns the enum constant of this type with the specified name.
static LabelMetrics[]	values()	Returns an array containing the constants of this enum type, in the order they are declared.

Methods inherited from class java.lang.Enum

clone, compareTo, equals, finalize, getDeclaringClass, hashCode, name, ordinal, toString, valueOf

Methods inherited from class java.lang.Object

getClass, notify, notifyAll, wait, wait, wait

Enum Constant Details

TP

public static final LabelMetrics TP

The number of true positives.

FP

public static final LabelMetrics FP

The number of false positives.

TN

public static final LabelMetrics TN

The number of true negatives.

FN

public static final LabelMetrics FN

The number of false negatives.

PRECISION

public static final LabelMetrics PRECISION

The precision, i.e., the number of true positives divided by the number of predicted positives.

RECALL

public static final LabelMetrics RECALL

The recall, i.e., the number of true positives divided by the number of ground truth positives.

F1

public static final LabelMetrics F1

The F_1 score, i.e., the harmonic mean of the precision and the recall.

ACCURACY

public static final LabelMetrics ACCURACY

The accuracy.

BALANCED_ERROR_RATE

public static final LabelMetrics BALANCED_ERROR_RATE

The balanced error rate, i.e., the mean of the per class recalls.

AUCROC

public static final LabelMetrics AUCROC

The area under the receiver-operator curve (ROC).

AVERAGED_PRECISION

public static final LabelMetrics AVERAGED_PRECISION

The averaged precision.

Method Details

values

public static LabelMetrics[] values()

Returns an array containing the constants of this enum type, in the order they are declared.

Returns:

an array containing the constants of this enum type, in the order they are declared

valueOf

public static LabelMetrics valueOf(String name)

Returns the enum constant of this type with the specified name. The string must match exactly an identifier used to declare an enum constant in this type. (Extraneous whitespace characters are not permitted.)

Parameters:

name - the name of the enum constant to be returned.

Returns:

the enum constant with the specified name

Throws:

IllegalArgumentException - if this enum type has no constant with the specified name

NullPointerException - if the argument is null

getImpl

public ToDoubleBiFunction<MetricTarget<Label>,LabelMetric.Context> getImpl()

Returns the implementing function for this metric.

Returns:

The implementing function.

forTarget

public LabelMetric forTarget(MetricTarget<Label> tgt)

Gets the LabelMetric wrapped around the supplied MetricTarget.

Parameters:

tgt - The metric target.

Returns:

The label metric combining the implementation function with the supplied metric target.

averagedPrecision

public static double averagedPrecision(MetricTarget<Label> tgt,
 List<Prediction<Label>> predictions)

Parameters:

tgt - The metric target to use.

predictions - The predictions to use.

Returns:

The averaged precision for the supplied target with the supplied predictions.

See Also:

• LabelEvaluationUtil.averagedPrecision(boolean[], double[])

averagedPrecision

public static double averagedPrecision(Label label,
 List<Prediction<Label>> predictions)

Parameters:

label - The Label to average across.

predictions - The predictions to use.

Returns:

The averaged precision for the supplied label with the supplied predictions.

See Also:

• LabelEvaluationUtil.averagedPrecision(boolean[], double[])

precisionRecallCurve

public static LabelEvaluationUtil.PRCurve precisionRecallCurve(Label label,
 List<Prediction<Label>> predictions)

Parameters:

label - The Label to calculate precision and recall for.

predictions - The predictions to use.

Returns:

The Precision Recall Curve for the supplied label with the supplied predictions.

See Also:

• LabelEvaluationUtil.generatePRCurve(boolean[], double[])

AUCROC

public static double AUCROC(Label label,
 List<Prediction<Label>> predictions)

Area under the ROC curve.

Parameters:

label - the label corresponding to the "positive" class

predictions - the predictions for which we'll compute the score

Returns:

AUC ROC score

Throws:

UnsupportedOperationException - if a prediction with no probability score, which are required to compute the ROC curve. (See also: Model.generatesProbabilities())

AUCROC

public static double AUCROC(MetricTarget<Label> tgt,
 List<Prediction<Label>> predictions)

Area under the ROC curve.

Parameters:

tgt - The metric target for the positive class.

predictions - the predictions for which we'll compute the score

Returns:

AUC ROC score

Throws:

UnsupportedOperationException - if a prediction with no probability score, which are required to compute the ROC curve. (See also: Model.generatesProbabilities())