de.lmu.ifi.dbs.elki.math.linearalgebra.pca.weightfunctions
Class ErfcStddevWeight

java.lang.Object
  de.lmu.ifi.dbs.elki.math.linearalgebra.pca.weightfunctions.ErfcStddevWeight

All Implemented Interfaces:

WeightFunction

public final class ErfcStddevWeight
extends Object
implements WeightFunction

Gaussian Error Function Weight function, scaled using stddev. This probably is the most statistically sound weight. erfc(1 / sqrt(2) * distance / stddev)

Author:

Erich Schubert

Field Summary

private static double

onebysqrt2 Precomputed value 1 / sqrt(2)

Constructor Summary

ErfcStddevWeight()

Method Summary

double

getWeight(double distance, double max, double stddev) Return Erfc weight, scaled by standard deviation. max is ignored.

Methods inherited from class java.lang.Object

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

Field Detail

onebysqrt2

private static final double onebysqrt2

Precomputed value 1 / sqrt(2)

Constructor Detail

ErfcStddevWeight

public ErfcStddevWeight()

Method Detail

getWeight

public double getWeight(double distance,
                        double max,
                        double stddev)

Return Erfc weight, scaled by standard deviation. max is ignored.

Specified by:

getWeight in interface WeightFunction

Parameters:

distance - distance of the query point

max - maximum distance of all included points

stddev - standard deviation (i.e. quadratic mean / RMS) of the included points

Returns:

weight for the query point