In the case of the Fumili algorithm the Error matrix (or the Hessian matrix containing the (approximate) second derivatives) is calculated using a linearization of the model function negleting second derivatives.

(In some sense the Name Updator is a little bit misleading as the Error matrix is not calculated by iteratively updating, like in Davidon's or other similar variable metric methods, but by recalculating each time).

Author: Andras Zsenei and Lorenzo Moneta, Creation date: 28 Sep 2004

See also: MINUIT Tutorial on function minimization, section 5; DavidonErrorUpdator

Definition at line 47 of file FumiliErrorUpdator.h.

Public Member Functions
	FumiliErrorUpdator ()

	~FumiliErrorUpdator ()

virtual MinimumError	Update (const MinimumState &fMinimumState, const MinimumParameters &fMinimumParameters, const GradientCalculator &fGradientCalculator, double lambda) const
	Member function that calculates the Error matrix (or the Hessian matrix containing the (approximate) second derivatives) using a linearization of the model function negleting second derivatives. More...

virtual MinimumError	Update (const MinimumState &, const MinimumParameters &, const FunctionGradient &) const
	Member function which is only present due to the design already in place of the software. More...

Public Member Functions inherited from ROOT::Minuit2::MinimumErrorUpdator
virtual	~MinimumErrorUpdator ()

#include <Minuit2/FumiliErrorUpdator.h>

Inheritance diagram for ROOT::Minuit2::FumiliErrorUpdator:

[legend]

Constructor & Destructor Documentation

◆ FumiliErrorUpdator()

ROOT::Minuit2::FumiliErrorUpdator::FumiliErrorUpdator ( )

inline

Definition at line 51 of file FumiliErrorUpdator.h.

◆ ~FumiliErrorUpdator()

ROOT::Minuit2::FumiliErrorUpdator::~FumiliErrorUpdator ( )

inline

Definition at line 53 of file FumiliErrorUpdator.h.

Member Function Documentation

◆ Update() [1/2]

MinimumError ROOT::Minuit2::FumiliErrorUpdator::Update	(	const MinimumState &	fMinimumState,
		const MinimumParameters &	fMinimumParameters,
		const GradientCalculator &	fGradientCalculator,
		double	lambda
	)		const

virtual

Member function that calculates the Error matrix (or the Hessian matrix containing the (approximate) second derivatives) using a linearization of the model function negleting second derivatives.

Parameters

fMinimumState	used to calculate the change in the covariance matrix between the two iterations
fMinimumParameters	the parameters at the present iteration
fGradientCalculator	the Gradient calculator used to retrieved the Parameter transformation
fFumiliFCNBase	the function calculating the figure of merit.

Definition at line 50 of file FumiliErrorUpdator.cxx.

◆ Update() [2/2]

MinimumError ROOT::Minuit2::FumiliErrorUpdator::Update	(	const MinimumState &	s0,
		const MinimumParameters &	p1,
		const FunctionGradient &	g1
	)		const

virtual

Member function which is only present due to the design already in place of the software.

As all classes calculating the Error matrix are supposed inherit from the MinimumErrorUpdator they must inherit this method. In some methods calculating the aforementioned matrix some of these parameters are not needed and other parameters are necessary... Hopefully, a more elegant solution will be found in the future.

Implements ROOT::Minuit2::MinimumErrorUpdator.

Definition at line 37 of file FumiliErrorUpdator.cxx.

Libraries for ROOT::Minuit2::FumiliErrorUpdator: