weka.core.neighboursearch
Class PLSNNSearch

java.lang.Object
  weka.core.neighboursearch.NearestNeighbourSearch
      weka.core.neighboursearch.NewNNSearch
          weka.core.neighboursearch.TransformNNSearch
              weka.core.neighboursearch.PLSNNSearch

All Implemented Interfaces:

Serializable, weka.core.AdditionalMeasureProducer, weka.core.OptionHandler, weka.core.RevisionHandler

public class PLSNNSearch
extends TransformNNSearch

Class implementing the brute force search algorithm for nearest neighbour search, filtered using PLS.

Valid options are:

 -S
  Skip identical instances (distances equal to zero).
 

Version:

$Revision: 4584 $

Author:

Dale

See Also:

Serialized Form

Nested Class Summary

Nested classes/interfaces inherited from class weka.core.neighboursearch.NewNNSearch

NewNNSearch.InstanceNode

Nested classes/interfaces inherited from class weka.core.neighboursearch.NearestNeighbourSearch

weka.core.neighboursearch.NearestNeighbourSearch.MyHeap, weka.core.neighboursearch.NearestNeighbourSearch.MyHeapElement, weka.core.neighboursearch.NearestNeighbourSearch.NeighborList, weka.core.neighboursearch.NearestNeighbourSearch.NeighborNode

Field Summary

protected weka.filters.supervised.attribute.PLSFilter	m_plsfilter The neighbourhood of instances to find neighbours in.
protected int	m_Preprocessing the type of preprocessing
protected int	mNumComponents

Fields inherited from class weka.core.neighboursearch.TransformNNSearch

m_myInstances

Fields inherited from class weka.core.neighboursearch.NewNNSearch

m_Distances, m_neighbours, m_SkipIdentical

Fields inherited from class weka.core.neighboursearch.NearestNeighbourSearch

m_DistanceFunction, m_Instances, m_kNN, m_MeasurePerformance, m_Stats

Constructor Summary

PLSNNSearch()
Constructor.

PLSNNSearch(weka.core.Instances insts)
Constructor that uses the supplied set of instances.

Method Summary

void	addInstanceInfo(weka.core.Instance ins) Adds the given instance info.
protected weka.filters.supervised.attribute.PLSFilter	buildFilter(int numcomponents)
int	getNumComponents()
String[]	getOptions() Gets the current settings.
weka.core.SelectedTag	getPreprocessing() Gets the type of preprocessing to use
weka.core.Instances	kNearestNeighbours(weka.core.Instance target, int kNN) Returns k nearest instances in the current neighbourhood to the supplied instance.
Enumeration	listOptions() Returns an enumeration describing the available options.
void	setInstances(weka.core.Instances insts) Sets the instances comprising the current neighbourhood.
void	setNumComponents(int nc)
void	setOptions(String[] options) Parses a given list of options.
void	setPreprocessing(weka.core.SelectedTag value) Sets the type of preprocessing to use
weka.core.Instance	transformInstance(weka.core.Instance in)
protected weka.core.Instances	transformInstances(weka.core.Instances in)
void	update(weka.core.Instance ins) Updates the LinearNNSearch to cater for the new added instance.

Methods inherited from class weka.core.neighboursearch.TransformNNSearch

getTransformedInstances

Methods inherited from class weka.core.neighboursearch.NewNNSearch

getDistances, getRevision, getSkipIdentical, globalInfo, nearestNeighbour, setSkipIdentical, skipIdenticalTipText

Methods inherited from class weka.core.neighboursearch.NearestNeighbourSearch

combSort11, distanceFunctionTipText, enumerateMeasures, getDistanceFunction, getInstances, getMeasure, getMeasurePerformance, getPerformanceStats, measurePerformanceTipText, partition, quickSort, setDistanceFunction, setMeasurePerformance

Methods inherited from class java.lang.Object

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

Field Detail

m_plsfilter

protected weka.filters.supervised.attribute.PLSFilter m_plsfilter

The neighbourhood of instances to find neighbours in.

mNumComponents

protected int mNumComponents

m_Preprocessing

protected int m_Preprocessing

the type of preprocessing

Constructor Detail

PLSNNSearch

public PLSNNSearch()

Constructor. Needs setInstances(Instances) to be called before the class is usable.

PLSNNSearch

public PLSNNSearch(weka.core.Instances insts)

Constructor that uses the supplied set of instances.

Parameters:

insts - the instances to use

Method Detail

setNumComponents

public void setNumComponents(int nc)

getNumComponents

public int getNumComponents()

transformInstances

protected weka.core.Instances transformInstances(weka.core.Instances in)
                                          throws Exception

Throws:

Exception

transformInstance

public weka.core.Instance transformInstance(weka.core.Instance in)
                                     throws Exception

Specified by:

transformInstance in class TransformNNSearch

Throws:

Exception

buildFilter

protected weka.filters.supervised.attribute.PLSFilter buildFilter(int numcomponents)

kNearestNeighbours

public weka.core.Instances kNearestNeighbours(weka.core.Instance target,
                                              int kNN)
                                       throws Exception

Returns k nearest instances in the current neighbourhood to the supplied instance.

Overrides:

kNearestNeighbours in class NewNNSearch

Parameters:

target - The instance to find the k nearest neighbours for.

kNN - The number of nearest neighbours to find.

Returns:

the k nearest neighbors

Throws:

Exception - if the neighbours could not be found.

setInstances

public void setInstances(weka.core.Instances insts)
                  throws Exception

Sets the instances comprising the current neighbourhood.

Overrides:

setInstances in class NewNNSearch

Parameters:

insts - The set of instances on which the nearest neighbour search is carried out. Usually this set is the training set.

Throws:

Exception - if setting of instances fails

update

public void update(weka.core.Instance ins)
            throws Exception

Updates the LinearNNSearch to cater for the new added instance. This implementation only updates the ranges of the DistanceFunction class, since our set of instances is passed by reference and should already have the newly added instance.

Overrides:

update in class NewNNSearch

Parameters:

ins - The instance to add. Usually this is the instance that is added to our neighbourhood i.e. the training instances.

Throws:

Exception - if the given instances are null

addInstanceInfo

public void addInstanceInfo(weka.core.Instance ins)

Adds the given instance info. This implementation updates the range datastructures of the DistanceFunction class.

Overrides:

addInstanceInfo in class NewNNSearch

Parameters:

ins - The instance to add the information of. Usually this is the test instance supplied to update the range of attributes in the distance function.

listOptions

public Enumeration listOptions()

Description copied from class: NewNNSearch

Returns an enumeration describing the available options.

Specified by:

listOptions in interface weka.core.OptionHandler

Overrides:

listOptions in class NewNNSearch

Returns:

an enumeration of all the available options.

setPreprocessing

public void setPreprocessing(weka.core.SelectedTag value)

Sets the type of preprocessing to use

Parameters:

value - the preprocessing type

setOptions

public void setOptions(String[] options)
                throws Exception

Description copied from class: NewNNSearch

Parses a given list of options.

Valid options are:

 -S
  Skip identical instances (distances equal to zero).
 

Specified by:

setOptions in interface weka.core.OptionHandler

Overrides:

setOptions in class NewNNSearch

Parameters:

options - the list of options as an array of strings

Throws:

Exception - if an option is not supported

getPreprocessing

public weka.core.SelectedTag getPreprocessing()

Gets the type of preprocessing to use

Returns:

the current preprocessing type.

getOptions

public String[] getOptions()

Description copied from class: NewNNSearch

Gets the current settings.

Specified by:

getOptions in interface weka.core.OptionHandler

Overrides:

getOptions in class NewNNSearch

Returns:

an array of strings suitable for passing to setOptions()