Package org.apache.spark.mllib.regression

Class IsotonicRegression

Object

org.apache.spark.mllib.regression.IsotonicRegression

All Implemented Interfaces:

Serializable

public class IsotonicRegression
extends Object
implements Serializable

Isotonic regression. Currently implemented using parallelized pool adjacent violators algorithm. Only univariate (single feature) algorithm supported.

Sequential PAV implementation based on: Grotzinger, S. J., and C. Witzgall. "Projections onto order simplexes." Applied mathematics and Optimization 12.1 (1984): 247-270.

Sequential PAV parallelization based on: Kearsley, Anthony J., Richard A. Tapia, and Michael W. Trosset. "An approach to parallelizing isotonic regression." Applied Mathematics and Parallel Computing. Physica-Verlag HD, 1996. 141-147. Available from here

See Also:

• Isotonic regression (Wikipedia)
• Serialized Form

Constructor Summary

Constructors

Constructor	Description
IsotonicRegression()	Constructs IsotonicRegression instance with default parameter isotonic = true.

Method Summary

Modifier and Type	Method	Description
IsotonicRegressionModel	run(JavaRDD<scala.Tuple3<Double,Double,Double>> input)	Run pool adjacent violators algorithm to obtain isotonic regression model.
IsotonicRegressionModel	run(RDD<scala.Tuple3<Object,Object,Object>> input)	Run IsotonicRegression algorithm to obtain isotonic regression model.
IsotonicRegression	setIsotonic(boolean isotonic)	Sets the isotonic parameter.

Methods inherited from class java.lang.Object

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

Constructor Details

IsotonicRegression

public IsotonicRegression()

Constructs IsotonicRegression instance with default parameter isotonic = true.

Method Details

run

public IsotonicRegressionModel run(RDD<scala.Tuple3<Object,Object,Object>> input)

Run IsotonicRegression algorithm to obtain isotonic regression model.

Parameters:

input - RDD of tuples (label, feature, weight) where label is dependent variable for which we calculate isotonic regression, feature is independent variable and weight represents number of measures with default 1. If multiple labels share the same feature value then they are aggregated using the weighted average before the algorithm is executed.

Returns:

Isotonic regression model.

run

public IsotonicRegressionModel run(JavaRDD<scala.Tuple3<Double,Double,Double>> input)

Run pool adjacent violators algorithm to obtain isotonic regression model.

Parameters:

input - JavaRDD of tuples (label, feature, weight) where label is dependent variable for which we calculate isotonic regression, feature is independent variable and weight represents number of measures with default 1. If multiple labels share the same feature value then they are aggregated using the weighted average before the algorithm is executed.

Returns:

Isotonic regression model.

setIsotonic

public IsotonicRegression setIsotonic(boolean isotonic)

Sets the isotonic parameter.

Parameters:

isotonic - Isotonic (increasing) or antitonic (decreasing) sequence.

Returns:

This instance of IsotonicRegression.