org.apache.spark.mllib.tree.model

Class TreeEnsembleModel

Object

org.apache.spark.mllib.tree.model.TreeEnsembleModel

All Implemented Interfaces:

java.io.Serializable

Direct Known Subclasses:

GradientBoostedTreesModel, RandomForestModel

public class TreeEnsembleModel
extends Object
implements scala.Serializable

Represents a tree ensemble model.

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
TreeEnsembleModel(scala.Enumeration.Value algo, DecisionTreeModel[] trees, double[] treeWeights, scala.Enumeration.Value combiningStrategy)

Method Summary

Methods

Modifier and Type	Method and Description
int	numTrees() Get number of trees in forest.
JavaRDD<Double>	predict(JavaRDD<Vector> features) Java-friendly version of predict(org.apache.spark.mllib.linalg.Vector).
RDD<Object>	predict(RDD<Vector> features) Predict values for the given data set.
double	predict(Vector features) Predict values for a single data point using the model trained.
String	toDebugString() Print the full model to a string.
String	toString() Print a summary of the model.
int	totalNumNodes() Get total number of nodes, summed over all trees in the forest.

Methods inherited from class Object

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

Constructor Detail

TreeEnsembleModel

public TreeEnsembleModel(scala.Enumeration.Value algo,
                 DecisionTreeModel[] trees,
                 double[] treeWeights,
                 scala.Enumeration.Value combiningStrategy)

Method Detail

predict

public double predict(Vector features)

Predict values for a single data point using the model trained.

Parameters:

features - array representing a single data point

Returns:

predicted category from the trained model

predict

public RDD<Object> predict(RDD<Vector> features)

Predict values for the given data set.

Parameters:

features - RDD representing data points to be predicted

Returns:

RDD[Double] where each entry contains the corresponding prediction

predict

public JavaRDD<Double> predict(JavaRDD<Vector> features)

Java-friendly version of predict(org.apache.spark.mllib.linalg.Vector).

toString

public String toString()

Print a summary of the model.

Overrides:

toString in class Object

toDebugString

public String toDebugString()

Print the full model to a string.

numTrees

public int numTrees()

Get number of trees in forest.

totalNumNodes

public int totalNumNodes()

Get total number of nodes, summed over all trees in the forest.