Package moa.classifiers.lazy.neighboursearch.kdtrees

Class KDTreeNodeSplitter

java.lang.Object

moa.classifiers.lazy.neighboursearch.kdtrees.KDTreeNodeSplitter

All Implemented Interfaces:

java.io.Serializable

Direct Known Subclasses:

KMeansInpiredMethod, MedianOfWidestDimension, MidPointOfWidestDimension, SlidingMidPointOfWidestSide

public abstract class KDTreeNodeSplitter
extends java.lang.Object
implements java.io.Serializable

Class that splits up a KDTreeNode.

Version:

$Revision: 8034 $

Author:

Ashraf M. Kibriya (amk14[at-the-rate]cs[dot]waikato[dot]ac[dot]nz)

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field	Description
protected EuclideanDistance	m_EuclideanDistance	The distance function used for building the tree.
protected Instances	m_Instances	The instances that'll be used for tree construction.
protected int[]	m_InstList	The master index array that'll be reshuffled as nodes are split and the tree is constructed.
protected boolean	m_NormalizeNodeWidth	Stores whether if the width of a KDTree node is normalized or not.
static int	MAX	Index of max value in an array of attributes' range.
static int	MIN	Index of min value in an array of attributes' range.
static int	WIDTH	Index of width value (max-min) in an array of attributes' range.

Constructor Summary

Constructors

Constructor	Description
KDTreeNodeSplitter()	default constructor.
KDTreeNodeSplitter(int[] instList, Instances insts, EuclideanDistance e)	Creates a new instance of KDTreeNodeSplitter.

Method Summary

Modifier and Type	Method	Description
protected void	correctlyInitialized()	Checks whether an object of this class has been correctly initialized.
java.lang.String[]	getOptions()	Gets the current settings of the object.
java.util.Enumeration	listOptions()	Returns an enumeration describing the available options.
void	setEuclideanDistanceFunction(EuclideanDistance func)	Sets the EuclideanDistance object to use for splitting nodes.
void	setInstanceList(int[] instList)	Sets the master index array containing indices of the training instances.
void	setInstances(Instances inst)	Sets the training instances on which the tree is (or is to be) built.
void	setNodeWidthNormalization(boolean normalize)	Sets whether if a nodes region is normalized or not.
void	setOptions(java.lang.String[] options)	Parses a given list of options.
abstract void	splitNode(KDTreeNode node, int numNodesCreated, double[][] nodeRanges, double[][] universe)	Splits a node into two.
protected int	widestDim(double[][] nodeRanges, double[][] universe)	Returns the widest dimension.

Methods inherited from class java.lang.Object

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

Field Detail

m_Instances

protected Instances m_Instances

The instances that'll be used for tree construction.

m_EuclideanDistance

protected EuclideanDistance m_EuclideanDistance

The distance function used for building the tree.

m_InstList

protected int[] m_InstList

The master index array that'll be reshuffled as nodes are split and the tree is constructed.

m_NormalizeNodeWidth

protected boolean m_NormalizeNodeWidth

Stores whether if the width of a KDTree node is normalized or not.

MIN

public static final int MIN

Index of min value in an array of attributes' range.

See Also:

Constant Field Values

MAX

public static final int MAX

Index of max value in an array of attributes' range.

See Also:

Constant Field Values

WIDTH

public static final int WIDTH

Index of width value (max-min) in an array of attributes' range.

See Also:

Constant Field Values

Constructor Detail

KDTreeNodeSplitter

public KDTreeNodeSplitter()

default constructor.

KDTreeNodeSplitter

public KDTreeNodeSplitter(int[] instList,
                          Instances insts,
                          EuclideanDistance e)

Creates a new instance of KDTreeNodeSplitter.

Parameters:

instList - Reference of the master index array.

insts - The set of training instances on which the tree is built.

e - The EuclideanDistance object that is used in tree contruction.

Method Detail

listOptions

public java.util.Enumeration listOptions()

Returns an enumeration describing the available options.

Returns:

an enumeration of all the available options.

setOptions

public void setOptions(java.lang.String[] options)
                throws java.lang.Exception

Parses a given list of options.

Parameters:

options - the list of options as an array of strings

Throws:

java.lang.Exception - if an option is not supported

getOptions

public java.lang.String[] getOptions()

Gets the current settings of the object.

Returns:

an array of strings suitable for passing to setOptions

correctlyInitialized

protected void correctlyInitialized()
                             throws java.lang.Exception

Checks whether an object of this class has been correctly initialized. Performs checks to see if all the necessary things (master index array, training instances, distance function) have been supplied or not.

Throws:

java.lang.Exception - If the object has not been correctly initialized.

splitNode

public abstract void splitNode(KDTreeNode node,
                               int numNodesCreated,
                               double[][] nodeRanges,
                               double[][] universe)
                        throws java.lang.Exception

Splits a node into two. After splitting two new nodes are created and correctly initialised. And, node.left and node.right are set appropriately.

Parameters:

node - The node to split.

numNodesCreated - The number of nodes that so far have been created for the tree, so that the newly created nodes are assigned correct/meaningful node numbers/ids.

nodeRanges - The attributes' range for the points inside the node that is to be split.

universe - The attributes' range for the whole point-space.

Throws:

java.lang.Exception - If there is some problem in splitting the given node.

setInstances

public void setInstances(Instances inst)

Sets the training instances on which the tree is (or is to be) built.

Parameters:

inst - The training instances.

setInstanceList

public void setInstanceList(int[] instList)

Sets the master index array containing indices of the training instances. This array will be rearranged as the tree is built, so that each node is assigned a portion in this array which contain the instances insides the node's region.

Parameters:

instList - The master index array.

setEuclideanDistanceFunction

public void setEuclideanDistanceFunction(EuclideanDistance func)

Sets the EuclideanDistance object to use for splitting nodes.

Parameters:

func - The EuclideanDistance object.

setNodeWidthNormalization

public void setNodeWidthNormalization(boolean normalize)

Sets whether if a nodes region is normalized or not. If set to true then, when selecting the widest attribute/dimension for splitting, the width of each attribute/dimension, of the points inside the node's region, is divided by the width of that attribute/dimension for the whole point-space. Thus, each attribute/dimension of that node is normalized.

Parameters:

normalize - Should be true if normalization is required.

widestDim

protected int widestDim(double[][] nodeRanges,
                        double[][] universe)

Returns the widest dimension. The width of each dimension (for the points inside the node) is normalized, if m_NormalizeNodeWidth is set to true.

Parameters:

nodeRanges - The attributes' range of the points inside the node that is to be split.

universe - The attributes' range for the whole point-space.

Returns:

The index of the attribute/dimension in which the points of the node have widest spread.