Namespace List

The namespaces specified in this document are:

Namespace	Assembly
Net.Kniaz.LMA	Net.Kniaz.LMA
Net.Kniaz.LMA.Tests	Net.Kniaz.LMA

Namespace : Net.Kniaz.LMA.Tests

Net.Kniaz.LMA.Tests Type List

Classes

Type	Summary
GaussianFit	Tests Gaussian Fit
GaussianFunction	Implements Gauss Bell Shape function
LineFit	Summary description for LineFit.
LineFunction	Initializes a new instance of the class.
LorenzianFit	Initializes a new instance of the class.
LorenzianFunction	Represents Lorenzian function. Derivative is calculated using a default f(x+dx)/dx method in the base class
SinFit	NUnit test function.
SinFunction	function represents sinusoidal aplitude

Net.Kniaz.LMA.Tests Classes

GaussianFunction Class

Summary

public class GaussianFunction : LMAFunction

Implements Gauss Bell Shape function

Constructor Members

Name	Access	Summary
GaussianFunction()	public	Initializes a new instance of the class.

Method Members

Name	Access	Summary
GenerateData() : Double[][]	public	(from Net.Kniaz.LMA.LMAFunction) Returns array of x,y values, given x and fitting parameters used by all tests to generate test data for exact fits
GetPartialDerivative() : Double	public	Derivative value
GetY() : Double	public	Returns Gaussian values

GaussianFit Class

Summary

public class GaussianFit

Tests Gaussian Fit

Constructor Members

Name	Access	Summary
GaussianFit()	public	Initializes a new instance of the class.

Method Members

Name	Access	Summary
RunGaussian() : Void	public

LineFunction Class

Summary

public class LineFunction : LMAFunction

Constructor Members

Name	Access	Summary
LineFunction()	public	Initializes a new instance of the class.

Method Members

Name	Access	Summary
GenerateData() : Double[][]	public	(from Net.Kniaz.LMA.LMAFunction) Returns array of x,y values, given x and fitting parameters used by all tests to generate test data for exact fits
GetPartialDerivative() : Double	public
GetY() : Double	public

LineFit Class

Summary

public class LineFit

Summary description for LineFit.

Constructor Members

Name	Access	Summary
LineFit()	public	Initializes a new instance of the class.

Method Members

Name	Access	Summary
RunLine() : Void	public

LorenzianFunction Class

Summary

public class LorenzianFunction : LMAFunction

Represents Lorenzian function. Derivative is calculated using a default f(x+dx)/dx method in the base class

Constructor Members

Name	Access	Summary
LorenzianFunction()	public	Initializes a new instance of the class.

Method Members

Name	Access	Summary
GenerateData() : Double[][]	public	(from Net.Kniaz.LMA.LMAFunction) Returns array of x,y values, given x and fitting parameters used by all tests to generate test data for exact fits
GetPartialDerivative() : Double	public	(from Net.Kniaz.LMA.LMAFunction) The method which gives the partial derivates used in the LMA fit. If you can't provide the functional derivative, use a small a-step (e.g., da = 1e-20) and return dy/da at the given x for each fit parameter. This is provided in the method below as a default implementation
GetY() : Double	public	Returns value ofthe Lorenzian

LorenzianFit Class

Summary

public class LorenzianFit

Constructor Members

Name	Access	Summary
LorenzianFit()	public	Initializes a new instance of the class.

Method Members

Name	Access	Summary
RunLorenzian() : Void	public

SinFunction Class

Summary

public class SinFunction : LMAFunction

function represents sinusoidal aplitude

Constructor Members

Name	Access	Summary
SinFunction()	public	Initializes a new instance of the class.

Method Members

Name	Access	Summary
GenerateData() : Double[][]	public	(from Net.Kniaz.LMA.LMAFunction) Returns array of x,y values, given x and fitting parameters used by all tests to generate test data for exact fits
GetPartialDerivative() : Double	public	Returns derivative
GetY() : Double	public	Returns value of the function

SinFit Class

Summary

public class SinFit

NUnit test function.

Constructor Members

Name	Access	Summary
SinFit()	public	Initializes a new instance of the class.

Method Members

Name	Access	Summary
RunSin() : Void	public	Test method

Namespace : Net.Kniaz.LMA

Net.Kniaz.LMA Type List

Classes

Type	Summary
Constants	Summary description for Constants.
LMA	Ctor. In the LMA fit N is the number of data points, M is the number of fit parameters. Call fit() to start the actual fitting.
LMAFunction	Abstract class implementing the LMAFunction interface

Net.Kniaz.LMA Classes

LMAFunction Class

Summary

public abstract class LMAFunction

Abstract class implementing the LMAFunction interface

Constructor Members

Name	Access	Summary
LMAFunction()	protected	Initializes a new instance of the class.

Method Members

Name	Access	Summary
GenerateData() : Double[][]	public	Returns array of x,y values, given x and fitting parameters used by all tests to generate test data for exact fits
GetPartialDerivative() : Double	public	The method which gives the partial derivates used in the LMA fit. If you can't provide the functional derivative, use a small a-step (e.g., da = 1e-20) and return dy/da at the given x for each fit parameter. This is provided in the method below as a default implementation
GetY() : Double	public	Returns the y value of the function for the given x and vector of parameters

Constants Class

Summary

public class Constants

Summary description for Constants.

Constructor Members

Name	Access	Summary
Constants()	public	Initializes a new instance of the class.

Field Members

Name	Access	Summary
lambda : Double	public

Method Members

Name	Access	Summary

LMA Class

Summary

public class LMA

Constructor Members

Name	Access	Summary
LMA()	public	Ctor. In the LMA fit N is the number of data points, M is the number of fit parameters. Call fit() to start the actual fitting.

Property Members

Name	Access	Summary
Chi2 : Double	public
Iterations : Int32	public
Parameters : Double[]	public

Method Members

Name	Access	Summary
CalculateAlphaElement() : Double	protected	Calculates lambda weighted element for the alpha-matrix. NOTE: Does not change the value of alpha-matrix.
CalculateBetaElement() : Double	protected	Calculates element of the beta (gradient) matrix NOTE: Does not change the value of beta-matrix.
CalculateChi2() : Double	protected	Calculates function value for the current fit parameters Does not change the value of chi2
CalculateChi2() : Double	protected	Calculates value of the function for given parameter array
CalculateIncrementedChi2() : Double	protected	Calculates function value for the incremented parameters (da + a). Does not change the value of chi2.
CheckWeights() : Double[]	protected	Checks if the matrix of weights for each point is a matrix of positive elements. Otherwise it initializes a new matrix and sets each value to 1
Fit() : Void	public	The default fit. If used after calling fit(lambda, minDeltaChi2, maxIterations), uses those values. The stop condition is fetched from this.stop(). Override this.stop() if you want to use another stop condition.
Fit() : Void	public	Initializes and starts the fit. The stop condition is fetched from this.stop(). Override this.stop() if you want to use another stop condition.
SolveIncrements() : Void	protected	Solves the increments array (this.da) using alpha and beta. Then updates the this.incrementedParameters array. NOTE: Inverts alpha. Call at least updateAlpha() before calling this.
Stop() : Boolean	public	The stop condition for the fit. Override this if you want to use another stop condition.
UpdateAlpha() : Void	protected	Calculates all elements for this.alpha.
UpdateBeta() : Void	protected	Calculates all elements for this.beta.
UpdateParameters() : Void	protected	Updates parameters from incrementedParameters.