gov.sandia.cognition.learning.algorithm.clustering.initializer

Class NeighborhoodGaussianClusterInitializer

java.lang.Object

gov.sandia.cognition.util.AbstractCloneableSerializable

gov.sandia.cognition.util.AbstractRandomized

gov.sandia.cognition.learning.algorithm.clustering.initializer.NeighborhoodGaussianClusterInitializer

All Implemented Interfaces:

FixedClusterInitializer<GaussianCluster,Vector>, CloneableSerializable, Randomized, java.io.Serializable, java.lang.Cloneable

@CodeReview(reviewer="Kevin R. Dixon",
            date="2008-07-23",
            changesNeeded=false,
            comments={"Now extends AbstractRandomized.","Got rid of C-style comments from inside methods.","Cleaned up javadoc a little bit with code annotations.","Otherwise, looks fine."})
public class NeighborhoodGaussianClusterInitializer
extends AbstractRandomized
implements FixedClusterInitializer<GaussianCluster,Vector>

Creates GaussianClusters near existing, but not on top of, data points. Also GaussianClusters are not created near each other.

Since:

1.0

Author:

Kevin R. Dixon

See Also:

Serialized Form

Field Summary

Fields

Modifier and Type	Field and Description
static double	DEFAULT_COVARIANCE Default covariance to put on the diagonal entries, 1.0.
static double	DEFAULT_RANGE Default range of the "neighborhood", 1.0.

Fields inherited from class gov.sandia.cognition.util.AbstractRandomized

random

Constructor Summary

Constructors

Constructor and Description
NeighborhoodGaussianClusterInitializer() Default constructor.
NeighborhoodGaussianClusterInitializer(double defaultCovariance, double randomRange, java.util.Random random) Creates a new instance of NeighborhoodGaussianClusterInitializer
NeighborhoodGaussianClusterInitializer(java.util.Random random) Creates a new instance of NeighborhoodGaussianClusterInitializer

Method Summary

Modifier and Type	Method and Description
double	getDefaultCovariance() Getter for defaultCovariance
double	getRandomRange() Getter for randomRange
java.util.ArrayList<GaussianCluster>	initializeClusters(int numClusters, java.util.Collection<? extends Vector> elements) Initializes a given number of clusters from the given elements.
void	setDefaultCovariance(double defaultCovariance) Setter for defaultCovariance
void	setRandomRange(double randomRange) Setter for randomRange

Methods inherited from class gov.sandia.cognition.util.AbstractRandomized

clone, getRandom, setRandom

Methods inherited from class java.lang.Object

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

Field Detail

DEFAULT_RANGE

public static final double DEFAULT_RANGE

Default range of the "neighborhood", 1.0.

See Also:

Constant Field Values

DEFAULT_COVARIANCE

public static final double DEFAULT_COVARIANCE

Default covariance to put on the diagonal entries, 1.0.

See Also:

Constant Field Values

Constructor Detail

NeighborhoodGaussianClusterInitializer

public NeighborhoodGaussianClusterInitializer()

Default constructor.

NeighborhoodGaussianClusterInitializer

public NeighborhoodGaussianClusterInitializer(java.util.Random random)

Creates a new instance of NeighborhoodGaussianClusterInitializer

Parameters:

random - random-number generator for the system

NeighborhoodGaussianClusterInitializer

public NeighborhoodGaussianClusterInitializer(double defaultCovariance,
                                              double randomRange,
                                              java.util.Random random)

Creates a new instance of NeighborhoodGaussianClusterInitializer

Parameters:

defaultCovariance - default diagonal covariance scaling factor

randomRange - range of the neighborhood from which to place the cluster

random - random-number generator for the system

Method Detail

getRandomRange

public double getRandomRange()

Getter for randomRange

Returns:

range of the neighborhood from which to place the cluster

setRandomRange

public void setRandomRange(double randomRange)

Setter for randomRange

Parameters:

randomRange - range of the neighborhood from which to place the cluster

getDefaultCovariance

public double getDefaultCovariance()

Getter for defaultCovariance

Returns:

default diagonal covariance scaling factor

setDefaultCovariance

public void setDefaultCovariance(double defaultCovariance)

Setter for defaultCovariance

Parameters:

defaultCovariance - default diagonal covariance scaling factor

initializeClusters

public java.util.ArrayList<GaussianCluster> initializeClusters(int numClusters,
                                                               java.util.Collection<? extends Vector> elements)

Description copied from interface: FixedClusterInitializer

Initializes a given number of clusters from the given elements.

Specified by:

initializeClusters in interface FixedClusterInitializer<GaussianCluster,Vector>

Parameters:

numClusters - The number of clusters to create.

elements - The elements to create the clusters from.

Returns:

The initial clusters to use.