vec_math
Class AbstractDataModel.LocalM

java.lang.Object
  vec_math.AbstractDataModel.LocalM

All Implemented Interfaces:

Multidimensional

Enclosing class:

AbstractDataModel

static class AbstractDataModel.LocalM
extends Object
implements Multidimensional

A localM-estimate model for the data. Depending on the choice of the Function rho, which is the negative logarithm of the measurement error distribution, we can either construct a normal-distributed M-estimate model ρ(z)=-1/2*z², or more robust estimates with a double-exponential or even Lorentzian distribution.

Field Summary

(package private) double[]	measures
(package private) Function	rho
(package private) DataModel	root
(package private) double[]	sigma

Constructor Summary

private

AbstractDataModel.LocalM(Function f, DataModel fit, double[] yvec, double[] sigvec)

Method Summary

int	dimension() The dimension equals the number of parameters in the model.
double	evaluate(VectorG a) We evaluate

Methods inherited from class java.lang.Object

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

Field Detail

rho

Function rho

root

DataModel root

measures

double[] measures

sigma

double[] sigma

Constructor Detail

AbstractDataModel.LocalM

private AbstractDataModel.LocalM(Function f,
                                 DataModel fit,
                                 double[] yvec,
                                 double[] sigvec)

Method Detail

dimension

public int dimension()

The dimension equals the number of parameters in the model.

Specified by:

dimension in interface Multidimensional

evaluate

public double evaluate(VectorG a)

We evaluate

           Σ_i(measure_i-model(x_i))^2/sigma_i^2
           

Specified by:

evaluate in interface Multidimensional