org.tribuo.common.tree

Class AbstractTrainingNode<T extends Output<T>>

java.lang.Object

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

All Implemented Interfaces:

Serializable, Node<T>

Direct Known Subclasses:

ClassifierTrainingNode, JointRegressorTrainingNode, RegressorTrainingNode

public abstract class AbstractTrainingNode<T extends Output<T>>
extends Object
implements Node<T>

Base class for decision tree nodes used at training time.

See Also:

Serialized Form

Nested Class Summary

Nested Classes

Modifier and Type	Class and Description
static class	AbstractTrainingNode.LeafDeterminer Contains parameters needed to determine whether a node is a leaf.

Field Summary

Fields

Modifier and Type	Field and Description
protected static int	DEFAULT_SIZE Default buffer size used in the split operation.
protected int	depth
protected Node<T>	greaterThan
protected double	impurityScore
protected AbstractTrainingNode.LeafDeterminer	leafDeterminer
protected Node<T>	lessThanOrEqual
protected int	numExamples
protected boolean	split
protected int	splitID
protected double	splitValue

Constructor Summary

Constructors

Modifier	Constructor and Description
protected	AbstractTrainingNode(int depth, int numExamples, AbstractTrainingNode.LeafDeterminer leafDeterminer) Builds an abstract training node.

Method Summary

Modifier and Type	Method and Description
abstract List<AbstractTrainingNode<T>>	buildTree(int[] featureIDs, SplittableRandom rng, boolean useRandomSplitPoints) Builds next level of a tree.
abstract Node<T>	convertTree() Converts a tree from a training representation to the final inference time representation.
Node<T>	copy() Copies the node and it's children.
SplitNode<T>	createSplitNode() Transforms an AbstractTrainingNode into a SplitNode
int	getDepth() The depth of this node in the tree.
Node<T>	getNextNode(SparseVector example) Returns the next node in the tree based on the supplied example, or null if it's a leaf.
int	getNumExamples() The number of training examples in this node.
abstract float	getWeightSum() The sum of the weights associated with this node's examples.
boolean	isLeaf() Is it a leaf node?
boolean	shouldMakeLeaf(double impurityScore, float weightSum) Determines whether the node to be created should be a LeafNode.

Methods inherited from class java.lang.Object

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

Methods inherited from interface org.tribuo.common.tree.Node

getImpurity

Field Detail

DEFAULT_SIZE

protected static final int DEFAULT_SIZE

Default buffer size used in the split operation.

See Also:

Constant Field Values

depth

protected final int depth

numExamples

protected final int numExamples

leafDeterminer

protected final AbstractTrainingNode.LeafDeterminer leafDeterminer

split

protected boolean split

splitID

protected int splitID

splitValue

protected double splitValue

impurityScore

protected double impurityScore

greaterThan

protected Node<T extends Output<T>> greaterThan

lessThanOrEqual

protected Node<T extends Output<T>> lessThanOrEqual

Constructor Detail

AbstractTrainingNode

protected AbstractTrainingNode(int depth,
                               int numExamples,
                               AbstractTrainingNode.LeafDeterminer leafDeterminer)

Builds an abstract training node.

Parameters:

depth - The depth of this node.

numExamples - The number of examples in this node.

leafDeterminer - The parameters which determine if the node forms a leaf.

Method Detail

buildTree

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

Builds next level of a tree.

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 abstract Node<T> convertTree()

Converts a tree from a training representation to the final inference time representation.

Returns:

The converted subtree.

getWeightSum

public abstract float getWeightSum()

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

Returns:

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

getDepth

public int getDepth()

The depth of this node in the tree.

Returns:

The depth.

shouldMakeLeaf

public boolean shouldMakeLeaf(double impurityScore,
                              float weightSum)

Determines whether the node to be created should be a LeafNode.

Parameters:

impurityScore - impurity score for the new node.

weightSum - total example weight for the new node.

Returns:

Whether the new node should be a LeafNode.

createSplitNode

public SplitNode<T> createSplitNode()

Transforms an AbstractTrainingNode into a SplitNode

Returns:

A SplitNode

getNextNode

public Node<T> getNextNode(SparseVector example)

Description copied from interface: Node

Returns the next node in the tree based on the supplied example, or null if it's a leaf.

Specified by:

getNextNode in interface Node<T extends Output<T>>

Parameters:

example - The example.

Returns:

The next node down in the tree.

getNumExamples

public int getNumExamples()

The number of training examples in this node.

Returns:

The number of training examples in this node.

isLeaf

public boolean isLeaf()

Description copied from interface: Node

Is it a leaf node?

Specified by:

isLeaf in interface Node<T extends Output<T>>

Returns:

True if it's a leaf node.

copy

public Node<T> copy()

Description copied from interface: Node

Copies the node and it's children.

Specified by:

copy in interface Node<T extends Output<T>>

Returns:

A deep copy.