de.lmu.ifi.dbs.elki.index.preprocessed.localpca
Class KNNQueryFilteredPCAIndex.Factory<V extends NumberVector<V,?>>

java.lang.Object
  extended by de.lmu.ifi.dbs.elki.index.preprocessed.localpca.AbstractFilteredPCAIndex.Factory<V,KNNQueryFilteredPCAIndex<V>>
      extended by de.lmu.ifi.dbs.elki.index.preprocessed.localpca.KNNQueryFilteredPCAIndex.Factory<V>
All Implemented Interfaces:
IndexFactory<V,KNNQueryFilteredPCAIndex<V>>, FilteredLocalPCAIndex.Factory<V,KNNQueryFilteredPCAIndex<V>>, LocalProjectionIndex.Factory<V,KNNQueryFilteredPCAIndex<V>>, InspectionUtilFrequentlyScanned, Parameterizable
Enclosing class:
KNNQueryFilteredPCAIndex<NV extends NumberVector<? extends NV,?>>

public static class KNNQueryFilteredPCAIndex.Factory<V extends NumberVector<V,?>>
extends AbstractFilteredPCAIndex.Factory<V,KNNQueryFilteredPCAIndex<V>>

Factory class


Nested Class Summary
static class KNNQueryFilteredPCAIndex.Factory.Parameterizer<NV extends NumberVector<NV,?>>
          Parameterization class.
 
Field Summary
private  Integer k
          Holds the value of K_ID.
static OptionID K_ID
          Optional parameter to specify the number of nearest neighbors considered in the PCA, must be an integer greater than 0.
 
Fields inherited from class de.lmu.ifi.dbs.elki.index.preprocessed.localpca.AbstractFilteredPCAIndex.Factory
pca, PCA_DISTANCE_ID, pcaDistanceFunction
 
Constructor Summary
KNNQueryFilteredPCAIndex.Factory(DistanceFunction<V,DoubleDistance> pcaDistanceFunction, PCAFilteredRunner<V> pca, Integer k)
          Constructor.
 
Method Summary
 KNNQueryFilteredPCAIndex<V> instantiate(Relation<V> relation)
          Instantiate the index for a given database.
 
Methods inherited from class de.lmu.ifi.dbs.elki.index.preprocessed.localpca.AbstractFilteredPCAIndex.Factory
getInputTypeRestriction
 
Methods inherited from class java.lang.Object
clone, equals, finalize, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait
 

Field Detail

K_ID

public static final OptionID K_ID
Optional parameter to specify the number of nearest neighbors considered in the PCA, must be an integer greater than 0. If this parameter is not set, k is set to three times of the dimensionality of the database objects.

Key: -localpca.k

Default value: three times of the dimensionality of the database objects


k

private Integer k
Holds the value of K_ID.

Constructor Detail

KNNQueryFilteredPCAIndex.Factory

public KNNQueryFilteredPCAIndex.Factory(DistanceFunction<V,DoubleDistance> pcaDistanceFunction,
                                        PCAFilteredRunner<V> pca,
                                        Integer k)
Constructor.

Parameters:
pcaDistanceFunction - distance
pca - PCA class
k - k
Method Detail

instantiate

public KNNQueryFilteredPCAIndex<V> instantiate(Relation<V> relation)
Description copied from interface: FilteredLocalPCAIndex.Factory
Instantiate the index for a given database.

Specified by:
instantiate in interface IndexFactory<V extends NumberVector<V,?>,KNNQueryFilteredPCAIndex<V extends NumberVector<V,?>>>
Specified by:
instantiate in interface FilteredLocalPCAIndex.Factory<V extends NumberVector<V,?>,KNNQueryFilteredPCAIndex<V extends NumberVector<V,?>>>
Specified by:
instantiate in interface LocalProjectionIndex.Factory<V extends NumberVector<V,?>,KNNQueryFilteredPCAIndex<V extends NumberVector<V,?>>>
Specified by:
instantiate in class AbstractFilteredPCAIndex.Factory<V extends NumberVector<V,?>,KNNQueryFilteredPCAIndex<V extends NumberVector<V,?>>>
Parameters:
relation - Relation to use
Returns:
Index

Release 0.4.0 (2011-09-20_1324)