jsat.clustering

Class FLAME

java.lang.Object

jsat.clustering.ClustererBase

jsat.clustering.FLAME

All Implemented Interfaces:

java.io.Serializable, Clusterer, Parameterized

public class FLAME
extends ClustererBase
implements Parameterized

Provides an implementation of the FLAME clustering algorithm. The original FLAME paper does not describe all necessary details for an implementation, so results may differ between implementations.

FLAME is highly sensitive to the number of neighbors chosen. Increasing the neighbors tends to reduce the number of clusters formed.

See: Fu, L.,&Medico, E. (2007). FLAME, a novel fuzzy clustering method for the analysis of DNA microarray data. BMC Bioinformatics, 8(1), 3. Retrieved from here

See Also:

Serialized Form

Constructor Summary

Constructors

Constructor and Description
FLAME(DistanceMetric dm, int k, int maxIterations) Creates a new FLAME clustering object
FLAME(FLAME toCopy) Copy constructor

Method Summary

Modifier and Type	Method and Description
FLAME	clone()
int[]	cluster(DataSet dataSet, boolean parallel, int[] designations) Performs clustering on the given data set.
DistanceMetric	getDistanceMetric() Returns the distance metric to use for the nearest neighbor search
double	getEps() Returns the minimum difference in scores to consider FLAME converged
int	getK() Returns the number of neighbors used
int	getMaxIterations() Returns the maximum number of iterations to perform
double	getStndDevs() Returns the number of standard deviations away from the mean density an outlier must be
void	setDistanceMetric(DistanceMetric dm) Sets the distance metric to use for the nearest neighbor search
void	setEps(double eps) Sets the convergence goal for the minimum difference in score between rounds.
void	setK(int k) Sets the number of neighbors that will be considered in determining Cluster Supporting Points and assignment contributions.
void	setMaxIterations(int maxIterations) Sets the maximum number of iterations to perform.
void	setStndDevs(double stndDevs) Sets the number of standard deviations away from the mean density a candidate outlier must be to be confirmed as an outlier.
void	setVectorCollectionFactory(VectorCollection<VecPaired<Vec,java.lang.Integer>> vc) Sets the vector collection used to accelerate the nearest neighbor search.

Methods inherited from class jsat.clustering.ClustererBase

createClusterListFromAssignmentArray, getDatapointsFromCluster

Methods inherited from class java.lang.Object

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

Methods inherited from interface jsat.parameters.Parameterized

getParameter, getParameters

Methods inherited from interface jsat.clustering.Clusterer

cluster, cluster, cluster, supportsWeightedData

Constructor Detail

FLAME

public FLAME(DistanceMetric dm,
             int k,
             int maxIterations)

Creates a new FLAME clustering object

Parameters:

dm - the distance metric to use

k - the number of neighbors to consider

maxIterations - the maximum number of iterations to perform

FLAME

public FLAME(FLAME toCopy)

Copy constructor

Parameters:

toCopy - the object to copy

Method Detail

setMaxIterations

public void setMaxIterations(int maxIterations)

Sets the maximum number of iterations to perform. FLAME can require far more iterations to converge than necessary to get the same hard clustering result.

Parameters:

maxIterations - the maximum number of iterations to perform

getMaxIterations

public int getMaxIterations()

Returns the maximum number of iterations to perform

Returns:

the maximum number of iterations to perform

setK

public void setK(int k)

Sets the number of neighbors that will be considered in determining Cluster Supporting Points and assignment contributions.

Parameters:

k - the number of neighbors to consider

getK

public int getK()

Returns the number of neighbors used

Returns:

the number of neighbors used

setEps

public void setEps(double eps)

Sets the convergence goal for the minimum difference in score between rounds. Negative values are allowed to force all iterations to occur

Parameters:

eps - the minimum difference in scores for convergence

getEps

public double getEps()

Returns the minimum difference in scores to consider FLAME converged

Returns:

the minimum difference in scores to consider FLAMe converged

setStndDevs

public void setStndDevs(double stndDevs)

Sets the number of standard deviations away from the mean density a candidate outlier must be to be confirmed as an outlier.

Parameters:

stndDevs - the number of standard deviations away from the mean density an outlier must be

getStndDevs

public double getStndDevs()

Returns the number of standard deviations away from the mean density an outlier must be

Returns:

the number of standard deviations away from the mean density an outlier must be

setDistanceMetric

public void setDistanceMetric(DistanceMetric dm)

Sets the distance metric to use for the nearest neighbor search

Parameters:

dm - the distance metric to use

getDistanceMetric

public DistanceMetric getDistanceMetric()

Returns the distance metric to use for the nearest neighbor search

Returns:

the distance metric to use

setVectorCollectionFactory

public void setVectorCollectionFactory(VectorCollection<VecPaired<Vec,java.lang.Integer>> vc)

Sets the vector collection used to accelerate the nearest neighbor search. The nearest neighbor only needs to be done once for each point, so the collection should be faster than the naive method when considering both construction and search time.

Parameters:

vc - the vector collection to use

cluster

public int[] cluster(DataSet dataSet,
                     boolean parallel,
                     int[] designations)

Description copied from interface: Clusterer

Performs clustering on the given data set. Parameters may be estimated by the method, or other heuristics performed.

Specified by:

cluster in interface Clusterer

Parameters:

dataSet - the data set to perform clustering on

parallel - true if multiple threads should be used to perform clustering. false if it should be done in a single threaded manner.

designations - the array which will contain the designated values. The array will be altered and returned by the function. If null is given, a new array will be created and returned.

Returns:

the int[]

clone

public FLAME clone()

Specified by:

clone in interface Clusterer

Specified by:

clone in class ClustererBase