databionics.esom.neighborhood
Class GaussNeighborhood

java.lang.Object
  databionics.esom.neighborhood.Neighborhood
      databionics.esom.neighborhood.CachedNeighborhood
          databionics.esom.neighborhood.GaussNeighborhood

public class GaussNeighborhood
extends CachedNeighborhood

Gaussian neighborhood kernel

Field Summary

protected static org.apache.log4j.Logger	log interface to log4j system
protected double	norm normalization factor
protected int	stddevs number of standard deviations corresponding to radius

Fields inherited from class databionics.esom.neighborhood.CachedNeighborhood

weights

Fields inherited from class databionics.esom.neighborhood.Neighborhood

radius, scaling

Constructor Summary

GaussNeighborhood()
Create a new neighborhood object.

GaussNeighborhood(int stddevs)
Create a new neighborhood object.

GaussNeighborhood(int radius, double scaling)
Create a new neighborhood object.

GaussNeighborhood(int radius, double scaling, int stddevs)
Create a new neighborhood object.

Method Summary

protected double	calculate(int distance) Calculate the weight value for this distance.
void	init() Initialize constants and the cache

Methods inherited from class databionics.esom.neighborhood.CachedNeighborhood

get

Methods inherited from class databionics.esom.neighborhood.Neighborhood

setRadius, setScaling

Methods inherited from class java.lang.Object

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

Field Detail

log

protected static final org.apache.log4j.Logger log

interface to log4j system

stddevs

protected int stddevs

number of standard deviations corresponding to radius

norm

protected double norm

normalization factor

Constructor Detail

GaussNeighborhood

public GaussNeighborhood()

Create a new neighborhood object.

GaussNeighborhood

public GaussNeighborhood(int stddevs)

Create a new neighborhood object.

GaussNeighborhood

public GaussNeighborhood(int radius,
                         double scaling,
                         int stddevs)

Create a new neighborhood object.

Parameters:

radius - radius of kernel

scaling - scaling factor for all weights

stddevs - number of standard deviations corresponding to radius

GaussNeighborhood

public GaussNeighborhood(int radius,
                         double scaling)

Create a new neighborhood object. Radius corresponds to 2 standard deviations.

Parameters:

radius - radius of kernel

scaling - scaling factor for all weights

Method Detail

init

public void init()

Initialize constants and the cache

Overrides:

init in class CachedNeighborhood

Parameters:

offset -

calculate

protected double calculate(int distance)

Calculate the weight value for this distance. Weights are calculated with the Gauss function.

Specified by:

calculate in class Neighborhood

Parameters:

distance - distance from center ([0..radius])

Returns:

weight value for this distance