de.lmu.ifi.dbs.elki.math.linearalgebra.fitting
Class GaussianFittingFunction

java.lang.Object
  de.lmu.ifi.dbs.elki.math.linearalgebra.fitting.GaussianFittingFunction

All Implemented Interfaces:

FittingFunction

public class GaussianFittingFunction
extends Object
implements FittingFunction

Gaussian function for parameter fitting Based mostly on fgauss in "Numerical Recpies in C".
However we've removed some small optimizations at the benefit of having easier to use parameters. Instead of position, amplitude and width used in the book, we use the traditional Gaussian parameters mean, standard deviation and a linear scaling factor (which is mostly useful when combining multiple distributions) The cost are some additional computations such as a square root. This could of course have been handled by an appropriate wrapper instead. They are also arranged differently: the book uses

 amplitude, position, width
 

whereas we use

 mean, stddev, scaling
 

The function also can use a mixture of gaussians, just use an appropriate number of parameters (which obviously needs to be a multiple of 3)

Author:

Erich Schubert

Field Summary

protected static double

Sqrt2PI Precomputed constant value of Sqrt(2*PI)

Constructor Summary

GaussianFittingFunction()

Method Summary

FittingFunctionResult

eval(double x, double[] params) compute the mixture of Gaussians at the given position

Methods inherited from class java.lang.Object

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

Field Detail

Sqrt2PI

protected static final double Sqrt2PI

Precomputed constant value of Sqrt(2*PI)

Constructor Detail

GaussianFittingFunction

public GaussianFittingFunction()

Method Detail

eval

public FittingFunctionResult eval(double x,
                                  double[] params)

compute the mixture of Gaussians at the given position

Specified by:

eval in interface FittingFunction

Parameters:

x - Current coordinate

params - Function parameters parameters

Returns:

Array consisting of y value and parameter gradients