weka.core.neighboursearch.kdtrees

Class KDTreeNode

java.lang.Object

weka.core.neighboursearch.kdtrees.KDTreeNode

All Implemented Interfaces:

java.io.Serializable, RevisionHandler

public class KDTreeNode
extends java.lang.Object
implements java.io.Serializable, RevisionHandler

A class representing a KDTree node. A node does not explicitly store the instances that it contains. Instead, it only stores the start and end index of a portion in a master index array. Each node is assigned a portion in the master index array that stores the indices of the instances that the node contains. Every time a node is split by the KDTree's contruction method, the instances of its left child are moved to the left and the instances of its right child are moved to the right, in the portion of the master index array belonging to the node. The start and end index in each of its children are then set accordingly within that portion so that each have their own portion which contains their instances. P.S.: The master index array is only stored in KDTree class.

Version:

$Revision: 1.2 $

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 and Description
int	m_End The end index of the portion of the master index array, which stores indices of the instances/points the node contains.
KDTreeNode	m_Left left subtree; contains instances with smaller or equal to split value.
int	m_NodeNumber node number (only for debug).
double[][]	m_NodeRanges lowest and highest value and width (= high - low) for each dimension.
double[][]	m_NodesRectBounds The lo and high bounds of the hyper rectangle described by the node.
KDTreeNode	m_Right right subtree; contains instances with larger than split value.
int	m_SplitDim attribute to split on.
double	m_SplitValue value to split on.
int	m_Start The start index of the portion of the master index array, which stores the indices of the instances/points the node contains.

Constructor Summary

Constructors

Constructor and Description
KDTreeNode() Constructor.
KDTreeNode(int nodeNum, int startidx, int endidx, double[][] nodeRanges) Constructor.
KDTreeNode(int nodeNum, int startidx, int endidx, double[][] nodeRanges, double[][] rectBounds)

Method Summary

Modifier and Type	Method and Description
java.lang.String	getRevision() Returns the revision string.
int	getSplitDim() Gets the splitting dimension.
double	getSplitValue() Gets the splitting value.
boolean	isALeaf() Checks if node is a leaf.
int	numInstances() Returns the number of Instances in the rectangular region defined by this node.

Methods inherited from class java.lang.Object

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

Field Detail

m_NodeNumber

public int m_NodeNumber

node number (only for debug).

m_Left

public KDTreeNode m_Left

left subtree; contains instances with smaller or equal to split value.

m_Right

public KDTreeNode m_Right

right subtree; contains instances with larger than split value.

m_SplitValue

public double m_SplitValue

value to split on.

m_SplitDim

public int m_SplitDim

attribute to split on.

m_NodeRanges

public double[][] m_NodeRanges

lowest and highest value and width (= high - low) for each dimension.

m_NodesRectBounds

public double[][] m_NodesRectBounds

The lo and high bounds of the hyper rectangle described by the node.

m_Start

public int m_Start

The start index of the portion of the master index array, which stores the indices of the instances/points the node contains.

m_End

public int m_End

The end index of the portion of the master index array, which stores indices of the instances/points the node contains.

Constructor Detail

KDTreeNode

public KDTreeNode()

Constructor.

KDTreeNode

public KDTreeNode(int nodeNum,
                  int startidx,
                  int endidx,
                  double[][] nodeRanges)

Constructor.

Parameters:

nodeNum - The node number/id.

startidx - The start index of node's portion in master index array.

endidx - The start index of node's portion in master index array.

nodeRanges - The attribute ranges of the Instances/points contained in this node.

KDTreeNode

public KDTreeNode(int nodeNum,
                  int startidx,
                  int endidx,
                  double[][] nodeRanges,
                  double[][] rectBounds)

Parameters:

nodeNum - The node number/id.

startidx - The start index of node's portion in master index array.

endidx - The start index of node's portion in master index array.

nodeRanges - The attribute ranges of the Instances/points contained in this node.

rectBounds - The range of the rectangular region in the point space that this node represents (points inside this rectangular region can have different range).

Method Detail

getSplitDim

public int getSplitDim()

Gets the splitting dimension.

Returns:

splitting dimension

getSplitValue

public double getSplitValue()

Gets the splitting value.

Returns:

splitting value

isALeaf

public boolean isALeaf()

Checks if node is a leaf.

Returns:

true if it is a leaf

numInstances

public int numInstances()

Returns the number of Instances in the rectangular region defined by this node.

Returns:

The number of instances in this KDTreeNode.

getRevision

public java.lang.String getRevision()

Returns the revision string.

Specified by:

getRevision in interface RevisionHandler

Returns:

the revision