org.apache.spark.mllib.evaluation

Class RankingMetrics<T>

java.lang.Object

org.apache.spark.mllib.evaluation.RankingMetrics<T>

All Implemented Interfaces:

java.io.Serializable

public class RankingMetrics<T>
extends java.lang.Object
implements scala.Serializable

::Experimental:: Evaluator for ranking algorithms.

Java users should use RankingMetrics$.of to create a RankingMetrics instance.

param: predictionAndLabels an RDD of (predicted ranking, ground truth set) pairs.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
RankingMetrics(RDD<scala.Tuple2<java.lang.Object,java.lang.Object>> predictionAndLabels, scala.reflect.ClassTag<T> evidence$1)

Method Summary

Methods

Modifier and Type	Method and Description
protected static void	initializeLogIfNecessary(boolean isInterpreter)
protected static boolean	isTraceEnabled()
protected static org.slf4j.Logger	log()
protected static void	logDebug(scala.Function0<java.lang.String> msg)
protected static void	logDebug(scala.Function0<java.lang.String> msg, java.lang.Throwable throwable)
protected static void	logError(scala.Function0<java.lang.String> msg)
protected static void	logError(scala.Function0<java.lang.String> msg, java.lang.Throwable throwable)
protected static void	logInfo(scala.Function0<java.lang.String> msg)
protected static void	logInfo(scala.Function0<java.lang.String> msg, java.lang.Throwable throwable)
protected static java.lang.String	logName()
protected static void	logTrace(scala.Function0<java.lang.String> msg)
protected static void	logTrace(scala.Function0<java.lang.String> msg, java.lang.Throwable throwable)
protected static void	logWarning(scala.Function0<java.lang.String> msg)
protected static void	logWarning(scala.Function0<java.lang.String> msg, java.lang.Throwable throwable)
double	meanAveragePrecision() Returns the mean average precision (MAP) of all the queries.
double	ndcgAt(int k) Compute the average NDCG value of all the queries, truncated at ranking position k.
static <E,T extends java.lang.Iterable<E>> RankingMetrics<E>	of(JavaRDD<scala.Tuple2<T,T>> predictionAndLabels) Creates a RankingMetrics instance (for Java users).
double	precisionAt(int k) Compute the average precision of all the queries, truncated at ranking position k.

Methods inherited from class java.lang.Object

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

Constructor Detail

RankingMetrics

public RankingMetrics(RDD<scala.Tuple2<java.lang.Object,java.lang.Object>> predictionAndLabels,
              scala.reflect.ClassTag<T> evidence$1)

Method Detail

of

public static <E,T extends java.lang.Iterable<E>> RankingMetrics<E> of(JavaRDD<scala.Tuple2<T,T>> predictionAndLabels)

Creates a RankingMetrics instance (for Java users).

Parameters:

predictionAndLabels - a JavaRDD of (predicted ranking, ground truth set) pairs

Returns:

(undocumented)

logName

protected static java.lang.String logName()

log

protected static org.slf4j.Logger log()

logInfo

protected static void logInfo(scala.Function0<java.lang.String> msg)

logDebug

protected static void logDebug(scala.Function0<java.lang.String> msg)

logTrace

protected static void logTrace(scala.Function0<java.lang.String> msg)

logWarning

protected static void logWarning(scala.Function0<java.lang.String> msg)

logError

protected static void logError(scala.Function0<java.lang.String> msg)

logInfo

protected static void logInfo(scala.Function0<java.lang.String> msg,
           java.lang.Throwable throwable)

logDebug

protected static void logDebug(scala.Function0<java.lang.String> msg,
            java.lang.Throwable throwable)

logTrace

protected static void logTrace(scala.Function0<java.lang.String> msg,
            java.lang.Throwable throwable)

logWarning

protected static void logWarning(scala.Function0<java.lang.String> msg,
              java.lang.Throwable throwable)

logError

protected static void logError(scala.Function0<java.lang.String> msg,
            java.lang.Throwable throwable)

isTraceEnabled

protected static boolean isTraceEnabled()

initializeLogIfNecessary

protected static void initializeLogIfNecessary(boolean isInterpreter)

precisionAt

public double precisionAt(int k)

Compute the average precision of all the queries, truncated at ranking position k.

If for a query, the ranking algorithm returns n (n < k) results, the precision value will be computed as #(relevant items retrieved) / k. This formula also applies when the size of the ground truth set is less than k.

If a query has an empty ground truth set, zero will be used as precision together with a log warning.

See the following paper for detail:

IR evaluation methods for retrieving highly relevant documents. K. Jarvelin and J. Kekalainen

Parameters:

k - the position to compute the truncated precision, must be positive

Returns:

the average precision at the first k ranking positions

meanAveragePrecision

public double meanAveragePrecision()

Returns the mean average precision (MAP) of all the queries. If a query has an empty ground truth set, the average precision will be zero and a log warning is generated.

Returns:

(undocumented)

ndcgAt

public double ndcgAt(int k)

Compute the average NDCG value of all the queries, truncated at ranking position k. The discounted cumulative gain at position k is computed as: sum,,i=1,,^k^ (2^{relevance of ''i''th item}^ - 1) / log(i + 1), and the NDCG is obtained by dividing the DCG value on the ground truth set. In the current implementation, the relevance value is binary.

If a query has an empty ground truth set, zero will be used as ndcg together with a log warning.

See the following paper for detail:

IR evaluation methods for retrieving highly relevant documents. K. Jarvelin and J. Kekalainen

Parameters:

k - the position to compute the truncated ndcg, must be positive

Returns:

the average ndcg at the first k ranking positions