Class RegressorTrainingNode

java.lang.Object

org.tribuo.common.tree.AbstractTrainingNode<Regressor>

org.tribuo.regression.rtree.impl.RegressorTrainingNode

All Implemented Interfaces:

Serializable, Node<Regressor>

public class RegressorTrainingNode
extends AbstractTrainingNode<Regressor>

A decision tree node used at training time. Contains a list of the example indices currently found in this node, the current impurity and a bunch of other statistics.

See Also:

• Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class	Description
static class	RegressorTrainingNode.InvertedData	Tuple containing an inverted dataset (i.e., feature-wise not exmaple-wise).

Nested classes/interfaces inherited from class org.tribuo.common.tree.AbstractTrainingNode

AbstractTrainingNode.LeafDeterminer

Field Summary

Fields inherited from class org.tribuo.common.tree.AbstractTrainingNode

DEFAULT_SIZE, depth, greaterThan, impurityScore, leafDeterminer, lessThanOrEqual, numExamples, split, splitID, splitValue

Constructor Summary

Constructors

Constructor	Description
RegressorTrainingNode(RegressorImpurity impurity, RegressorTrainingNode.InvertedData tuple, int dimIndex, String dimName, int numExamples, ImmutableFeatureMap featureIDMap, ImmutableOutputInfo<Regressor> outputInfo, AbstractTrainingNode.LeafDeterminer leafDeterminer)

Method Summary

Modifier and Type	Method	Description
List<AbstractTrainingNode<Regressor>>	buildTree(int[] featureIDs, SplittableRandom rng, boolean useRandomSplitPoints)	Builds a tree according to CART (as it does not do multi-way splits on categorical values like C4.5).
Node<Regressor>	convertTree()	Generates a test time tree (made of SplitNode and LeafNode) from the tree rooted at this node.
double	getImpurity()	The impurity score of this node.
float	getWeightSum()	The sum of the weights associated with this node's examples.
static RegressorTrainingNode.InvertedData	invertData(Dataset<Regressor> examples)	Inverts a training dataset from row major to column major.

Methods inherited from class org.tribuo.common.tree.AbstractTrainingNode

copy, createSplitNode, getDepth, getNextNode, getNumExamples, isLeaf, shouldMakeLeaf

Methods inherited from class java.lang.Object

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

Constructor Details

RegressorTrainingNode

public RegressorTrainingNode(RegressorImpurity impurity,
 RegressorTrainingNode.InvertedData tuple,
 int dimIndex,
 String dimName,
 int numExamples,
 ImmutableFeatureMap featureIDMap,
 ImmutableOutputInfo<Regressor> outputInfo,
 AbstractTrainingNode.LeafDeterminer leafDeterminer)

Method Details

getImpurity

public double getImpurity()

Description copied from interface: Node

The impurity score of this node.

Returns:

The node impurity.

getWeightSum

public float getWeightSum()

Description copied from class: AbstractTrainingNode

The sum of the weights associated with this node's examples.

Specified by:

getWeightSum in class AbstractTrainingNode<Regressor>

Returns:

the sum of the weights associated with this node's examples.

buildTree

public List<AbstractTrainingNode<Regressor>> buildTree(int[] featureIDs,
 SplittableRandom rng,
 boolean useRandomSplitPoints)

Builds a tree according to CART (as it does not do multi-way splits on categorical values like C4.5).

Specified by:

buildTree in class AbstractTrainingNode<Regressor>

Parameters:

featureIDs - Indices of the features available in this split.

rng - Splittable random number generator.

useRandomSplitPoints - Whether to choose split points for features at random.

Returns:

A possibly empty list of TrainingNodes.

convertTree

public Node<Regressor> convertTree()

Generates a test time tree (made of SplitNode and LeafNode) from the tree rooted at this node.

Specified by:

convertTree in class AbstractTrainingNode<Regressor>

Returns:

A subtree using the SplitNode and LeafNode classes.

invertData

public static RegressorTrainingNode.InvertedData invertData(Dataset<Regressor> examples)

Inverts a training dataset from row major to column major. This partially de-sparsifies the dataset so it's very expensive in terms of memory.

Parameters:

examples - An input dataset.

Returns:

A list of TreeFeatures which contain InvertedFeatures.