jsat.distributions.multivariate

Class ProductKDE

java.lang.Object

jsat.distributions.multivariate.MultivariateDistributionSkeleton

jsat.distributions.multivariate.MultivariateKDE

jsat.distributions.multivariate.ProductKDE

All Implemented Interfaces:

Serializable, Cloneable, MultivariateDistribution

public class ProductKDE
extends MultivariateKDE

The Product Kernel Density Estimator is a generalization of the KernelDensityEstimator to the multivariate case. This is done by using a kernel and bandwidth for each dimension, such that the bandwidth for each dimension can be determined using the same methods as the univariate KDE. This can simplify the difficulty in bandwidth selection for the multivariate case.

Author:

Edward Raff

See Also:

MetricKDE, Serialized Form

Constructor Summary

Constructors

Constructor and Description
ProductKDE() Creates a new KDE that uses the EpanechnikovKF kernel.
ProductKDE(KernelFunction k) Creates a new KDE that uses the specified kernel

Method Summary

Modifier and Type	Method and Description
ProductKDE	clone()
KernelFunction	getKernelFunction()
List<VecPaired<VecPaired<Vec,Integer>,Double>>	getNearby(Vec x) Returns the list of vectors that have a non zero contribution to the density of the query point x.
List<VecPaired<VecPaired<Vec,Integer>,Double>>	getNearbyRaw(Vec x) Returns the list of vectors that have a non zero contribution to the density of the query point x.
double	pdf(Vec x) Returns the probability of a given vector from this distribution.
List<Vec>	sample(int count, Random rand) Performs sampling on the current distribution.
void	scaleBandwidth(double scale) A caller may want to increase or decrease the bandwidth after training has been completed to get smoother model, or decrease it to observe behavior.
<V extends Vec> boolean	setUsingData(List<V> dataSet) Sets the parameters of the distribution to attempt to fit the given list of vectors.
boolean	setUsingDataList(List<DataPoint> dataPoints) Sets the parameters of the distribution to attempt to fit the given list of data points.

Methods inherited from class jsat.distributions.multivariate.MultivariateDistributionSkeleton

logPdf, logPdf, pdf, setUsingData, setUsingData, setUsingData, setUsingDataList

Methods inherited from class java.lang.Object

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

Constructor Detail

ProductKDE

public ProductKDE()

Creates a new KDE that uses the EpanechnikovKF kernel.

ProductKDE

public ProductKDE(KernelFunction k)

Creates a new KDE that uses the specified kernel

Parameters:

k - the kernel method to use

Method Detail

clone

public ProductKDE clone()

Specified by:

clone in interface MultivariateDistribution

Specified by:

clone in class MultivariateKDE

getNearby

public List<VecPaired<VecPaired<Vec,Integer>,Double>> getNearby(Vec x)

Description copied from class: MultivariateKDE

Returns the list of vectors that have a non zero contribution to the density of the query point x. Each vector is paired with its integer index from the original constructing list vectors, and a double indicating its weight given the kernel function in use.

Specified by:

getNearby in class MultivariateKDE

Parameters:

x - the query point

Returns:

the list of near by vectors and their weights

getNearbyRaw

public List<VecPaired<VecPaired<Vec,Integer>,Double>> getNearbyRaw(Vec x)

Description copied from class: MultivariateKDE

Returns the list of vectors that have a non zero contribution to the density of the query point x. Each vector is paired with its integer index from the original constructing list vectors, and a double indicating its distance from the query point divided by the bandwidth of the point.

Specified by:

getNearbyRaw in class MultivariateKDE

Parameters:

x - the query point

Returns:

the list of near by vectors and their weights

pdf

public double pdf(Vec x)

Description copied from interface: MultivariateDistribution

Returns the probability of a given vector from this distribution. By definition, the probability will always be in the range [0, 1].

Parameters:

x - the vector the get the log probability of

Returns:

the probability

setUsingData

public <V extends Vec> boolean setUsingData(List<V> dataSet)

Description copied from interface: MultivariateDistribution

Sets the parameters of the distribution to attempt to fit the given list of vectors. All vectors are assumed to have the same weight.

Type Parameters:

V - the vector type

Parameters:

dataSet - the list of data points

Returns:

true if the distribution was fit to the data, or false if the distribution could not be fit to the data set.

setUsingDataList

public boolean setUsingDataList(List<DataPoint> dataPoints)

Description copied from interface: MultivariateDistribution

Sets the parameters of the distribution to attempt to fit the given list of data points. The weights of the data points will be used.

Parameters:

dataPoints - the list of data points to use

Returns:

true if the distribution was fit to the data, or false if the distribution could not be fit to the data set.

sample

public List<Vec> sample(int count,
                        Random rand)

Description copied from interface: MultivariateDistribution

Performs sampling on the current distribution.

Parameters:

count - the number of iid samples to draw

rand - the source of randomness

Returns:

a list of sample vectors from this distribution

getKernelFunction

public KernelFunction getKernelFunction()

Specified by:

getKernelFunction in class MultivariateKDE

Returns:

the kernel function used

scaleBandwidth

public void scaleBandwidth(double scale)

Description copied from class: MultivariateKDE

A caller may want to increase or decrease the bandwidth after training has been completed to get smoother model, or decrease it to observe behavior. This method will scaled the bandwidth of each data point by the given factor

Specified by:

scaleBandwidth in class MultivariateKDE

Parameters:

scale - the value to scale the bandwidth used