Package org.apache.spark.mllib.evaluation

Class BinaryClassificationMetrics

Object

org.apache.spark.mllib.evaluation.BinaryClassificationMetrics

All Implemented Interfaces:

org.apache.spark.internal.Logging

public class BinaryClassificationMetrics
extends Object
implements org.apache.spark.internal.Logging

Evaluator for binary classification.

param: scoreAndLabels an RDD of (score, label) or (score, label, weight) tuples. param: numBins if greater than 0, then the curves (ROC curve, PR curve) computed internally will be down-sampled to this many "bins". If 0, no down-sampling will occur. This is useful because the curve contains a point for each distinct score in the input, and this could be as large as the input itself -- millions of points or more, when thousands may be entirely sufficient to summarize the curve. After down-sampling, the curves will instead be made of approximately numBins points instead. Points are made from bins of equal numbers of consecutive points. The size of each bin is floor(scoreAndLabels.count() / numBins), which means the resulting number of bins may not exactly equal numBins. The last bin in each partition may be smaller as a result, meaning there may be an extra sample at partition boundaries.

Nested Class Summary

Nested classes/interfaces inherited from interface org.apache.spark.internal.Logging

org.apache.spark.internal.Logging.LogStringContext, org.apache.spark.internal.Logging.SparkShellLoggingFilter

Constructor Summary

Constructors

Constructor	Description
BinaryClassificationMetrics(RDD<? extends scala.Product> scoreAndLabels, int numBins)
BinaryClassificationMetrics(RDD<scala.Tuple2<Object,Object>> scoreAndLabels)	Defaults numBins to 0.

Method Summary

Modifier and Type	Method	Description
double	areaUnderPR()	Computes the area under the precision-recall curve.
double	areaUnderROC()	Computes the area under the receiver operating characteristic (ROC) curve.
RDD<scala.Tuple2<Object,Object>>	fMeasureByThreshold()	Returns the (threshold, F-Measure) curve with beta = 1.0.
RDD<scala.Tuple2<Object,Object>>	fMeasureByThreshold(double beta)	Returns the (threshold, F-Measure) curve.
int	numBins()
RDD<scala.Tuple2<Object,Object>>	pr()	Returns the precision-recall curve, which is an RDD of (recall, precision), NOT (precision, recall), with (0.0, p) prepended to it, where p is the precision associated with the lowest recall on the curve.
RDD<scala.Tuple2<Object,Object>>	precisionByThreshold()	Returns the (threshold, precision) curve.
RDD<scala.Tuple2<Object,Object>>	recallByThreshold()	Returns the (threshold, recall) curve.
RDD<scala.Tuple2<Object,Object>>	roc()	Returns the receiver operating characteristic (ROC) curve, which is an RDD of (false positive rate, true positive rate) with (0.0, 0.0) prepended and (1.0, 1.0) appended to it.
RDD<? extends scala.Product>	scoreAndLabels()
RDD<Object>	thresholds()	Returns thresholds in descending order.
void	unpersist()	Unpersist intermediate RDDs used in the computation.

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.apache.spark.internal.Logging

initializeForcefully, initializeLogIfNecessary, initializeLogIfNecessary, initializeLogIfNecessary$default$2, isTraceEnabled, log, logBasedOnLevel, logDebug, logDebug, logDebug, logDebug, logError, logError, logError, logError, logInfo, logInfo, logInfo, logInfo, logName, LogStringContext, logTrace, logTrace, logTrace, logTrace, logWarning, logWarning, logWarning, logWarning, MDC, org$apache$spark$internal$Logging$$log_, org$apache$spark$internal$Logging$$log__$eq, withLogContext

Constructor Details

BinaryClassificationMetrics

public BinaryClassificationMetrics(RDD<? extends scala.Product> scoreAndLabels,
 int numBins)

BinaryClassificationMetrics

public BinaryClassificationMetrics(RDD<scala.Tuple2<Object,Object>> scoreAndLabels)

Defaults numBins to 0.

Parameters:

scoreAndLabels - (undocumented)

Method Details

scoreAndLabels

public RDD<? extends scala.Product> scoreAndLabels()

numBins

public int numBins()

unpersist

public void unpersist()

Unpersist intermediate RDDs used in the computation.

thresholds

public RDD<Object> thresholds()

Returns thresholds in descending order.

Returns:

(undocumented)

roc

public RDD<scala.Tuple2<Object,Object>> roc()

Returns the receiver operating characteristic (ROC) curve, which is an RDD of (false positive rate, true positive rate) with (0.0, 0.0) prepended and (1.0, 1.0) appended to it.

Returns:

(undocumented)

See Also:

• Receiver operating characteristic (Wikipedia)

areaUnderROC

public double areaUnderROC()

Computes the area under the receiver operating characteristic (ROC) curve.

Returns:

(undocumented)

pr

public RDD<scala.Tuple2<Object,Object>> pr()

Returns the precision-recall curve, which is an RDD of (recall, precision), NOT (precision, recall), with (0.0, p) prepended to it, where p is the precision associated with the lowest recall on the curve.

Returns:

(undocumented)

See Also:

• Precision and recall (Wikipedia)

areaUnderPR

public double areaUnderPR()

Computes the area under the precision-recall curve.

Returns:

(undocumented)

fMeasureByThreshold

public RDD<scala.Tuple2<Object,Object>> fMeasureByThreshold(double beta)

Returns the (threshold, F-Measure) curve.

Parameters:

beta - the beta factor in F-Measure computation.

Returns:

an RDD of (threshold, F-Measure) pairs.

See Also:

• F1 score (Wikipedia)

fMeasureByThreshold

public RDD<scala.Tuple2<Object,Object>> fMeasureByThreshold()

Returns the (threshold, F-Measure) curve with beta = 1.0.

Returns:

(undocumented)

precisionByThreshold

public RDD<scala.Tuple2<Object,Object>> precisionByThreshold()

Returns the (threshold, precision) curve.

Returns:

(undocumented)

recallByThreshold

public RDD<scala.Tuple2<Object,Object>> recallByThreshold()

Returns the (threshold, recall) curve.

Returns:

(undocumented)