org.openimaj.ml.clustering.assignment.soft

Class ShortKNNAssigner

java.lang.Object

org.openimaj.ml.clustering.assignment.soft.ShortKNNAssigner

All Implemented Interfaces:

Assigner<short[]>, SoftAssigner<short[],float[]>

public class ShortKNNAssigner
extends Object
implements SoftAssigner<short[],float[]>

A SoftAssigner that picks a fixed number of nearest neighbours. Weights returned are actually the distances to the centroids.

Author:

Jonathon Hare (jsh2@ecs.soton.ac.uk)

Field Summary

Fields

Modifier and Type	Field and Description
protected ShortNearestNeighbours	nn
protected int	numNeighbours

Constructor Summary

Constructors

Constructor and Description
ShortKNNAssigner(CentroidsProvider<short[]> provider, boolean exact, int numNeighbours) Construct the assigner using the given cluster data.
ShortKNNAssigner(CentroidsProvider<short[]> provider, ShortFVComparison comparison, int numNeighbours) Construct the assigner using the given cluster data and distance function.
ShortKNNAssigner(short[][] data, boolean exact, int numNeighbours) Construct the assigner using the given cluster data.
ShortKNNAssigner(short[][] data, ShortFVComparison comparison, int numNeighbours) Construct the assigner using the given cluster data and distance function.

Method Summary

Modifier and Type	Method and Description
int[]	assign(short[] data) Assign a single point to some clusters.
int[][]	assign(short[][] data) Assign data to clusters.
IndependentPair<int[],float[]>	assignWeighted(short[] data) Assign a single point to some clusters.
void	assignWeighted(short[][] data, int[][] assignments, float[][] weights) Assign data to clusters.
int	numDimensions() Get the number of dimensions of the input vectors.
int	size() The number of clusters.

Methods inherited from class java.lang.Object

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

Field Detail

nn

protected ShortNearestNeighbours nn

numNeighbours

protected int numNeighbours

Constructor Detail

ShortKNNAssigner

public ShortKNNAssigner(CentroidsProvider<short[]> provider,
                        boolean exact,
                        int numNeighbours)

Construct the assigner using the given cluster data. The assigner is backed by either a ShortNearestNeighboursExact or ShortNearestNeighboursKDTree, depending on whether the exact parameter is true or false. If the parameter is true, then the resultant ShortNearestNeighboursExact will use Euclidean distance.

Parameters:

provider - the cluster data provider

exact - if true, then use exact mode; false implies approximate mode.

numNeighbours - the number of nearest neighbours to select.

ShortKNNAssigner

public ShortKNNAssigner(short[][] data,
                        boolean exact,
                        int numNeighbours)

Construct the assigner using the given cluster data. The assigner is backed by either a ShortNearestNeighboursExact or ShortNearestNeighboursKDTree, depending on whether the exact parameter is true or false. If the parameter is true, then the resultant ShortNearestNeighboursExact will use Euclidean distance.

Parameters:

data - the cluster data

exact - if true, then use exact mode; false implies approximate mode.

numNeighbours - the number of nearest neighbours to select.

ShortKNNAssigner

public ShortKNNAssigner(CentroidsProvider<short[]> provider,
                        ShortFVComparison comparison,
                        int numNeighbours)

Construct the assigner using the given cluster data and distance function. The assigner will operate in exact mode, using a ShortNearestNeighboursExact.

Parameters:

provider - the cluster data provider

comparison - the distance function

numNeighbours - the number of nearest neighbours to select.

ShortKNNAssigner

public ShortKNNAssigner(short[][] data,
                        ShortFVComparison comparison,
                        int numNeighbours)

Construct the assigner using the given cluster data and distance function. The assigner will operate in exact mode, using a ShortNearestNeighboursExact.

Parameters:

data - the cluster data

comparison - the distance function

numNeighbours - the number of nearest neighbours to select.

Method Detail

assign

public int[][] assign(short[][] data)

Description copied from interface: SoftAssigner

Assign data to clusters.

Specified by:

assign in interface SoftAssigner<short[],float[]>

Parameters:

data - the data.

Returns:

The cluster indices which the data was assigned to.

assign

public int[] assign(short[] data)

Description copied from interface: SoftAssigner

Assign a single point to some clusters.

Specified by:

assign in interface SoftAssigner<short[],float[]>

Parameters:

data - datum to assign.

Returns:

the assigned cluster indices.

assignWeighted

public void assignWeighted(short[][] data,
                           int[][] assignments,
                           float[][] weights)

Description copied from interface: SoftAssigner

Assign data to clusters. The results are returned in the indices and distances arrays. The return arrays must have the same length as the data array.

Specified by:

assignWeighted in interface SoftAssigner<short[],float[]>

Parameters:

data - the data.

assignments - the cluster indices for each data point.

weights - the weights to the for each cluster for each data point.

assignWeighted

public IndependentPair<int[],float[]> assignWeighted(short[] data)

Description copied from interface: SoftAssigner

Assign a single point to some clusters.

Specified by:

assignWeighted in interface SoftAssigner<short[],float[]>

Parameters:

data - point to assign.

Returns:

the assigned cluster indices and weights.

numDimensions

public int numDimensions()

Description copied from interface: Assigner

Get the number of dimensions of the input vectors.

Specified by:

numDimensions in interface Assigner<short[]>

Returns:

the number of dimensions

size

public int size()

Description copied from interface: SoftAssigner

The number of clusters.

Specified by:

size in interface SoftAssigner<short[],float[]>

Returns:

The number of clusters or unique ids that can be generated.