weka.core.neighboursearch.balltrees

Class MiddleOutConstructor

java.lang.Object

weka.core.neighboursearch.balltrees.BallTreeConstructor

weka.core.neighboursearch.balltrees.MiddleOutConstructor

All Implemented Interfaces:

Serializable, OptionHandler, Randomizable, RevisionHandler, TechnicalInformationHandler

public class MiddleOutConstructor
extends BallTreeConstructor
implements Randomizable, TechnicalInformationHandler

The class that builds a BallTree middle out.

For more information see also:

Andrew W. Moore: The Anchors Hierarchy: Using the Triangle Inequality to Survive High Dimensional Data. In: UAI '00: Proceedings of the 16th Conference on Uncertainty in Artificial Intelligence, San Francisco, CA, USA, 397-405, 2000.

Ashraf Masood Kibriya (2007). Fast Algorithms for Nearest Neighbour Search. Hamilton, New Zealand.

BibTeX:

 @inproceedings{Moore2000,
    address = {San Francisco, CA, USA},
    author = {Andrew W. Moore},
    booktitle = {UAI '00: Proceedings of the 16th Conference on Uncertainty in Artificial Intelligence},
    pages = {397-405},
    publisher = {Morgan Kaufmann Publishers Inc.},
    title = {The Anchors Hierarchy: Using the Triangle Inequality to Survive High Dimensional Data},
    year = {2000}
 }
 
 @mastersthesis{Kibriya2007,
    address = {Hamilton, New Zealand},
    author = {Ashraf Masood Kibriya},
    school = {Department of Computer Science, School of Computing and Mathematical Sciences, University of Waikato},
    title = {Fast Algorithms for Nearest Neighbour Search},
    year = {2007}
 }
 

Valid options are:

 -S <num>
  The seed for the random number generator used
  in selecting random anchor.
 (default: 1)

 -R
  Use randomly chosen initial anchors.

Version:

$Revision: 8034 $

Author:

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

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
MiddleOutConstructor() Creates a new instance of MiddleOutConstructor.

Method Summary

Methods

Modifier and Type	Method and Description
int[]	addInstance(BallNode node, Instance inst) Adds an instance to the tree.
BallNode	buildTree() Builds a ball tree middle out.
Instance	calcPivot(weka.core.neighboursearch.balltrees.MiddleOutConstructor.MyIdxList list1, weka.core.neighboursearch.balltrees.MiddleOutConstructor.MyIdxList list2, Instances insts) Calculates the centroid pivot of a node based on the list of points that it contains (tbe two lists of its children are provided).
Instance	calcPivot(weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode node1, weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode node2, Instances insts) /** Calculates the centroid pivot of a node based on its two child nodes (if merging two nodes).
double	calcRadius(weka.core.neighboursearch.balltrees.MiddleOutConstructor.MyIdxList list1, weka.core.neighboursearch.balltrees.MiddleOutConstructor.MyIdxList list2, Instance pivot, Instances insts) Calculates the radius of a node based on the list of points that it contains (the two lists of its children are provided).
double	calcRadius(weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode n1, weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode n2) Calculates the radius of a node based on its two child nodes (if merging two nodes).
void	checkIndicesList(weka.core.neighboursearch.balltrees.MiddleOutConstructor.MyIdxList list, int startidx, int endidx) Checks whether if the points in an index list are in some specified of the master index array.
String[]	getOptions() Gets the current settings of this BallTree MiddleOutConstructor.
String	getRevision() Returns the revision string.
int	getSeed() Returns the seed for random number generator.
TechnicalInformation	getTechnicalInformation() Returns an instance of a TechnicalInformation object, containing detailed information about the technical background of this class, e.g., paper reference or book this class is based on.
String	globalInfo() Returns a string describing this nearest neighbour search algorithm.
String	initialAnchorRandomTipText() Returns the tip text for this property.
boolean	isInitialAnchorRandom() Gets whether if the initial anchor is chosen randomly.
Enumeration	listOptions() Returns an enumeration describing the available options.
String	printInsts(int startIdx, int endIdx) For printing indices in some given portion of the master index array.
String	printList(weka.core.neighboursearch.balltrees.MiddleOutConstructor.MyIdxList points) For printing indices in a given point list.
String	seedTipText() Returns the tip text for this property.
void	setInitialAnchorRandom(boolean randomInitialAnchor) Sets whether if the initial anchor is chosen randomly.
void	setInstanceList(int[] instList) Sets the master index array that points to instances in m_Instances, so that only this array is manipulated, and m_Instances is left untouched.
void	setInstances(Instances insts) Sets the instances on which the tree is to be built.
void	setInterAnchorDistances(Vector<weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode> anchors, weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode newAnchor, Vector<double[]> anchorDistances) Sets the distances of a supplied new anchor to all the rest of the previous anchor points.
void	setMaxInstancesInLeaf(int num) Sets the maximum number of instances allowed in a leaf.
void	setOptions(String[] options) Parses a given list of options.
void	setPoints(weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode node, int startIdx, int endIdx, int[] indices) Sets the points of an anchor node.
void	setSeed(int seed) Sets the seed for random number generator (that is used for selecting the first anchor point randomly).
boolean	stealPoints(weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode newAnchor, Vector anchors, Vector anchorDistances) Removes points from old anchors that are nearer to the given new anchor and adds them to the list of points of the new anchor.

Methods inherited from class weka.core.neighboursearch.balltrees.BallTreeConstructor

containChildBallsTipText, getContainChildBalls, getMaxDepth, getMaxInstancesInLeaf, getMaxRelativeLeafRadius, getNumLeaves, getNumNodes, maxInstancesInLeafTipText, maxRelativeLeafRadiusTipText, setContainChildBalls, setEuclideanDistanceFunction, setMaxRelativeLeafRadius

Methods inherited from class java.lang.Object

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

Constructor Detail

MiddleOutConstructor

public MiddleOutConstructor()

Creates a new instance of MiddleOutConstructor.

Method Detail

globalInfo

public String globalInfo()

Returns a string describing this nearest neighbour search algorithm.

Returns:

a description of the algorithm for displaying in the explorer/experimenter gui

getTechnicalInformation

public TechnicalInformation getTechnicalInformation()

Returns an instance of a TechnicalInformation object, containing detailed information about the technical background of this class, e.g., paper reference or book this class is based on.

Specified by:

getTechnicalInformation in interface TechnicalInformationHandler

Returns:

the technical information about this class

buildTree

public BallNode buildTree()
                   throws Exception

Builds a ball tree middle out.

Specified by:

buildTree in class BallTreeConstructor

Returns:

The root node of the tree.

Throws:

Exception - If there is problem building the tree.

setPoints

public void setPoints(weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode node,
             int startIdx,
             int endIdx,
             int[] indices)

Sets the points of an anchor node. It takes the indices of points from the given portion of an index array and stores those indices, together with their distances to the given anchor node, in the point index list of the anchor node.

Parameters:

node - The node in which the points are needed to be set.

startIdx - The start of the portion in the given index array (the master index array).

endIdx - The end of the portion in the given index array.

indices - The index array.

setInterAnchorDistances

public void setInterAnchorDistances(Vector<weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode> anchors,
                           weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode newAnchor,
                           Vector<double[]> anchorDistances)
                             throws Exception

Sets the distances of a supplied new anchor to all the rest of the previous anchor points.

Parameters:

anchors - The old anchor points.

newAnchor - The new anchor point.

anchorDistances - The vector to store the distances of newAnchor to each of the old anchors.

Throws:

Exception - If there is some problem in calculating the distances.

stealPoints

public boolean stealPoints(weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode newAnchor,
                  Vector anchors,
                  Vector anchorDistances)

Removes points from old anchors that are nearer to the given new anchor and adds them to the list of points of the new anchor.

Parameters:

newAnchor - The new anchor.

anchors - The old anchors.

anchorDistances - The distances of new anchor to each of the old anchors.

Returns:

true if any points are removed from the old anchors

calcPivot

public Instance calcPivot(weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode node1,
                 weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode node2,
                 Instances insts)

/** Calculates the centroid pivot of a node based on its two child nodes (if merging two nodes).

Parameters:

node1 - The first child.

node2 - The second child.

insts - The set of instances on which the tree is being built (as dataset header information is required).

Returns:

The centroid pivot of a node.

calcPivot

public Instance calcPivot(weka.core.neighboursearch.balltrees.MiddleOutConstructor.MyIdxList list1,
                 weka.core.neighboursearch.balltrees.MiddleOutConstructor.MyIdxList list2,
                 Instances insts)

Calculates the centroid pivot of a node based on the list of points that it contains (tbe two lists of its children are provided).

Parameters:

list1 - The point index list of first child.

list2 - The point index list of second child.

insts - The insts object on which the tree is being built (for header information).

Returns:

The centroid pivot of the node.

calcRadius

public double calcRadius(weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode n1,
                weka.core.neighboursearch.balltrees.MiddleOutConstructor.TempNode n2)

Calculates the radius of a node based on its two child nodes (if merging two nodes).

Parameters:

n1 - The first child of the node.

n2 - The second child of the node.

Returns:

The radius of the node.

Throws:

Exception

calcRadius

public double calcRadius(weka.core.neighboursearch.balltrees.MiddleOutConstructor.MyIdxList list1,
                weka.core.neighboursearch.balltrees.MiddleOutConstructor.MyIdxList list2,
                Instance pivot,
                Instances insts)

Calculates the radius of a node based on the list of points that it contains (the two lists of its children are provided).

Parameters:

list1 - The point index list of first child.

list2 - The point index list of second child.

pivot - The centre/pivot of the node.

insts - The instances on which the tree is being built (for header info).

Returns:

The radius of the node.

addInstance

public int[] addInstance(BallNode node,
                Instance inst)
                  throws Exception

Adds an instance to the tree. This implementation of MiddleOutConstructor doesn't support addition of instances to already built tree, hence it always throws an exception.

Specified by:

addInstance in class BallTreeConstructor

Parameters:

node - The root of the tree to which the instance is to be added.

inst - The instance to add to the tree.

Returns:

The updated master index array after adding the instance.

Throws:

Exception - Always as this implementation of MiddleOutConstructor doesn't support addition of instances after batch construction of the tree.

setMaxInstancesInLeaf

public void setMaxInstancesInLeaf(int num)
                           throws Exception

Sets the maximum number of instances allowed in a leaf.

Overrides:

setMaxInstancesInLeaf in class BallTreeConstructor

Parameters:

num - The maximum number of instances allowed in a leaf.

Throws:

Exception - If the num is < 2, as the method cannot work for < 2 instances.

setInstances

public void setInstances(Instances insts)

Sets the instances on which the tree is to be built.

Overrides:

setInstances in class BallTreeConstructor

Parameters:

insts - The instances on which to build the ball tree.

setInstanceList

public void setInstanceList(int[] instList)

Sets the master index array that points to instances in m_Instances, so that only this array is manipulated, and m_Instances is left untouched.

Overrides:

setInstanceList in class BallTreeConstructor

Parameters:

instList - The master index array.

initialAnchorRandomTipText

public String initialAnchorRandomTipText()

Returns the tip text for this property.

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

isInitialAnchorRandom

public boolean isInitialAnchorRandom()

Gets whether if the initial anchor is chosen randomly.

Returns:

true if the initial anchor is a random one.

setInitialAnchorRandom

public void setInitialAnchorRandom(boolean randomInitialAnchor)

Sets whether if the initial anchor is chosen randomly. If not then if it is the furthest point from the mean/centroid.

Parameters:

randomInitialAnchor - Should be true if the first anchor is to be chosen randomly.

seedTipText

public String seedTipText()

Returns the tip text for this property.

Returns:

tip text for this property suitable for displaying in the explorer/experimenter gui

getSeed

public int getSeed()

Returns the seed for random number generator.

Specified by:

getSeed in interface Randomizable

Returns:

The random number seed.

setSeed

public void setSeed(int seed)

Sets the seed for random number generator (that is used for selecting the first anchor point randomly).

Specified by:

setSeed in interface Randomizable

Parameters:

seed - The seed.

listOptions

public Enumeration listOptions()

Returns an enumeration describing the available options.

Specified by:

listOptions in interface OptionHandler

Overrides:

listOptions in class BallTreeConstructor

Returns:

an enumeration of all the available options.

setOptions

public void setOptions(String[] options)
                throws Exception

Parses a given list of options. Valid options are:

 -S <num>
  The seed for the random number generator used
  in selecting random anchor.
 (default: 1)

 -R
  Use randomly chosen initial anchors.

Specified by:

setOptions in interface OptionHandler

Overrides:

setOptions in class BallTreeConstructor

Parameters:

options - the list of options as an array of strings

Throws:

Exception - if an option is not supported

getOptions

public String[] getOptions()

Gets the current settings of this BallTree MiddleOutConstructor.

Specified by:

getOptions in interface OptionHandler

Overrides:

getOptions in class BallTreeConstructor

Returns:

an array of strings suitable for passing to setOptions

checkIndicesList

public void checkIndicesList(weka.core.neighboursearch.balltrees.MiddleOutConstructor.MyIdxList list,
                    int startidx,
                    int endidx)
                      throws Exception

Checks whether if the points in an index list are in some specified of the master index array.

Parameters:

list - The point list.

startidx - The start of the portion in master index array.

endidx - The end of the portion in master index array.

Throws:

Exception - If some point in the point list is not in the specified portion of master index array.

printInsts

public String printInsts(int startIdx,
                int endIdx)

For printing indices in some given portion of the master index array.

Parameters:

startIdx - The start of the portion in master index array.

endIdx - The end of the portion in master index array.

Returns:

The string containing the indices in specified portion of the master index array.

printList

public String printList(weka.core.neighboursearch.balltrees.MiddleOutConstructor.MyIdxList points)

For printing indices in a given point list.

Parameters:

points - The point list.

Returns:

String containing indices of the points in the point list.

getRevision

public String getRevision()

Returns the revision string.

Specified by:

getRevision in interface RevisionHandler

Overrides:

getRevision in class BallTreeConstructor

Returns:

the revision