org.apache.spark.mllib.regression

Class IsotonicRegressionModel

java.lang.Object

org.apache.spark.mllib.regression.IsotonicRegressionModel

All Implemented Interfaces:

java.io.Serializable, Saveable

public class IsotonicRegressionModel
extends java.lang.Object
implements java.io.Serializable, Saveable

Regression model for isotonic regression.

param: boundaries Array of boundaries for which predictions are known. Boundaries must be sorted in increasing order. param: predictions Array of predictions associated to the boundaries at the same index. Results of isotonic regression and therefore monotone. param: isotonic indicates whether this is isotonic or antitonic.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
IsotonicRegressionModel(double[] boundaries, double[] predictions, boolean isotonic)
IsotonicRegressionModel(java.lang.Iterable<java.lang.Object> boundaries, java.lang.Iterable<java.lang.Object> predictions, java.lang.Boolean isotonic) A Java-friendly constructor that takes two Iterable parameters and one Boolean parameter.

Method Summary

Methods

Modifier and Type	Method and Description
double[]	boundaries()
protected java.lang.String	formatVersion() Current version of model save/load format.
boolean	isotonic()
static IsotonicRegressionModel	load(SparkContext sc, java.lang.String path)
double	predict(double testData) Predict a single label.
JavaDoubleRDD	predict(JavaDoubleRDD testData) Predict labels for provided features.
RDD<java.lang.Object>	predict(RDD<java.lang.Object> testData) Predict labels for provided features.
double[]	predictions()
void	save(SparkContext sc, java.lang.String path) Save this model to the given path.

Methods inherited from class java.lang.Object

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

Constructor Detail

IsotonicRegressionModel

public IsotonicRegressionModel(double[] boundaries,
                       double[] predictions,
                       boolean isotonic)

IsotonicRegressionModel

public IsotonicRegressionModel(java.lang.Iterable<java.lang.Object> boundaries,
                       java.lang.Iterable<java.lang.Object> predictions,
                       java.lang.Boolean isotonic)

A Java-friendly constructor that takes two Iterable parameters and one Boolean parameter.

Parameters:

boundaries - (undocumented)

predictions - (undocumented)

isotonic - (undocumented)

Method Detail

load

public static IsotonicRegressionModel load(SparkContext sc,
                           java.lang.String path)

boundaries

public double[] boundaries()

predictions

public double[] predictions()

isotonic

public boolean isotonic()

predict

public RDD<java.lang.Object> predict(RDD<java.lang.Object> testData)

Predict labels for provided features. Using a piecewise linear function.

Parameters:

testData - Features to be labeled.

Returns:

Predicted labels.

predict

public JavaDoubleRDD predict(JavaDoubleRDD testData)

Predict labels for provided features. Using a piecewise linear function.

Parameters:

testData - Features to be labeled.

Returns:

Predicted labels.

predict

public double predict(double testData)

Predict a single label. Using a piecewise linear function.

Parameters:

testData - Feature to be labeled.

Returns:

Predicted label. 1) If testData exactly matches a boundary then associated prediction is returned. In case there are multiple predictions with the same boundary then one of them is returned. Which one is undefined (same as java.util.Arrays.binarySearch). 2) If testData is lower or higher than all boundaries then first or last prediction is returned respectively. In case there are multiple predictions with the same boundary then the lowest or highest is returned respectively. 3) If testData falls between two values in boundary array then prediction is treated as piecewise linear function and interpolated value is returned. In case there are multiple values with the same boundary then the same rules as in 2) are used.

save

public void save(SparkContext sc,
        java.lang.String path)

Description copied from interface: Saveable

Save this model to the given path.

This saves: - human-readable (JSON) model metadata to path/metadata/ - Parquet formatted data to path/data/

The model may be loaded using Loader.load.

Specified by:

save in interface Saveable

Parameters:

sc - Spark context used to save model data.

path - Path specifying the directory in which to save this model. If the directory already exists, this method throws an exception.

formatVersion

protected java.lang.String formatVersion()

Description copied from interface: Saveable

Current version of model save/load format.

Specified by:

formatVersion in interface Saveable