org.apache.spark.mllib.evaluation

Class RegressionMetrics

Object

org.apache.spark.mllib.evaluation.RegressionMetrics

public class RegressionMetrics
extends Object

Evaluator for regression.

param: predictionAndObservations an RDD of (prediction, observation) pairs param: throughOrigin True if the regression is through the origin. For example, in linear regression, it will be true without fitting intercept.

Constructor Summary

Constructors

Constructor and Description
RegressionMetrics(RDD<scala.Tuple2<Object,Object>> predictionAndObservations)
RegressionMetrics(RDD<scala.Tuple2<Object,Object>> predictionAndObservations, boolean throughOrigin)

Method Summary

Methods

Modifier and Type	Method and Description
double	explainedVariance() Returns the variance explained by regression.
double	meanAbsoluteError() Returns the mean absolute error, which is a risk function corresponding to the expected value of the absolute error loss or l1-norm loss.
double	meanSquaredError() Returns the mean squared error, which is a risk function corresponding to the expected value of the squared error loss or quadratic loss.
double	r2() Returns R^2^, the unadjusted coefficient of determination.
double	rootMeanSquaredError() Returns the root mean squared error, which is defined as the square root of the mean squared error.

Methods inherited from class Object

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

Constructor Detail

RegressionMetrics

public RegressionMetrics(RDD<scala.Tuple2<Object,Object>> predictionAndObservations,
                 boolean throughOrigin)

RegressionMetrics

public RegressionMetrics(RDD<scala.Tuple2<Object,Object>> predictionAndObservations)

Method Detail

explainedVariance

public double explainedVariance()

Returns the variance explained by regression. explainedVariance = \sum_i (\hat{y_i} - \bar{y})^2 / n

Returns:

(undocumented)

See Also:

meanAbsoluteError

public double meanAbsoluteError()

Returns the mean absolute error, which is a risk function corresponding to the expected value of the absolute error loss or l1-norm loss.

Returns:

(undocumented)

meanSquaredError

public double meanSquaredError()

Returns the mean squared error, which is a risk function corresponding to the expected value of the squared error loss or quadratic loss.

Returns:

(undocumented)

rootMeanSquaredError

public double rootMeanSquaredError()

Returns the root mean squared error, which is defined as the square root of the mean squared error.

Returns:

(undocumented)

r2

public double r2()

Returns R^2^, the unadjusted coefficient of determination.

Returns:

(undocumented)

See Also:

In case of regression through the origin, the definition of R^2^ is to be modified., G. Eisenhauer, Regression through the Origin. Teaching Statistics 25, 76-80 (2003) {@link https://online.stat.psu.edu/~ajw13/stat501/SpecialTopics/Reg_thru_origin.pdf}