Wrapper for the CERN ROOT facade of the Minuit2 minimizer. More...

Inheritance diagram for Minuit2Minimizer:

Public Member Functions
	Minuit2Minimizer (const std::string &algorithmName="Migrad")

void	setStrategy (int value)
	Sets minimization strategy (0-low, 1-medium, 2-high minimization quality). More...

int	strategy () const

void	setErrorDefinition (double value)
	Sets error definition factor for parameter error calculation. More...

double	errorDefinition () const

void	setTolerance (double value)
	Sets tolerance on the function value at the minimum. More...

double	tolerance () const

void	setPrecision (double value)
	Sets relative floating point arithmetic precision. More...

double	precision () const

void	setPrintLevel (int value)
	Sets minimizer internal print level. More...

int	printLevel () const

void	setMaxFunctionCalls (int value)
	Sets maximum number of objective function calls.

int	maxFunctionCalls () const

std::string	statusToString () const override
	Returns string representation of current minimizer status.

std::map< std::string, std::string >	statusMap () const override
	Returns map of string representing different minimizer statuses.

bool	requiresResiduals () override
	Returns true if minimizer computations are residual-based, false otherwise.

Public Member Functions inherited from RootMinimizerAdapter
Fit::MinimizerResult	minimize_scalar (fcn_scalar_t fcn, Fit::Parameters parameters) override
	run minimization

Fit::MinimizerResult	minimize_residual (fcn_residual_t fcn, Fit::Parameters parameters) override

std::string	minimizerName () const override final
	Returns name of the minimizer.

std::string	algorithmName () const override final
	Returns name of the minimization algorithm.

void	setParameters (const Fit::Parameters &parameters)

double	minValue () const override final
	Returns minimum function value.

MinimizerOptions &	options ()

const MinimizerOptions &	options () const

bool	providesError () const
	Returns true if minimizer provides error and error matrix.

void	setOptions (const std::string &optionString) override final
	Sets option string to the minimizer.

Public Member Functions inherited from IMinimizer
	IMinimizer (const IMinimizer &other)=delete

IMinimizer &	operator= (const IMinimizer &other)=delete

virtual void	clear ()
	clear resources (parameters) for consecutives minimizations

Protected Member Functions
void	propagateOptions () override
	Propagate options down to ROOT's Minuit2Minimizer.

const root_minimizer_t *	rootMinimizer () const override

Protected Member Functions inherited from RootMinimizerAdapter
	RootMinimizerAdapter (const MinimizerInfo &minimizerInfo)

Fit::MinimizerResult	minimize (Fit::Parameters parameters)

void	propagateResults (Fit::Parameters &parameters)
	Propagates results of minimization to fit parameter set.

virtual void	setParameter (unsigned int index, const Fit::Parameter &par)

size_t	fitDimension () const
	Returns number of fit parameters defined (i.e. dimension of the function to be minimized).

std::vector< double >	parValuesAtMinimum () const
	Returns value of the variables at minimum.

std::vector< double >	parErrorsAtMinimum () const
	Returns errors of the variables at minimum.

root_minimizer_t *	rootMinimizer ()

template<class T >
OptionContainer::option_t	addOption (const std::string &optionName, T value, const std::string &description="")

template<class T >
void	setOptionValue (const std::string &optionName, T value)

template<class T >
T	optionValue (const std::string &optionName) const

Additional Inherited Members
Public Types inherited from RootMinimizerAdapter
typedef ROOT::Math::Minimizer	root_minimizer_t

Detailed Description

Wrapper for the CERN ROOT facade of the Minuit2 minimizer.

See Minuit2 user manual https://root.cern.ch/root/htmldoc/guides/minuit2/Minuit2.pdf.

Definition at line 33 of file Minuit2Minimizer.h.

Member Function Documentation

◆ setStrategy()

void Minuit2Minimizer::setStrategy ( int value )

Sets minimization strategy (0-low, 1-medium, 2-high minimization quality).

At low quality number of function calls will be economized.

Definition at line 61 of file Minuit2Minimizer.cpp.

◆ setErrorDefinition()

void Minuit2Minimizer::setErrorDefinition ( double value )

Sets error definition factor for parameter error calculation.

If objective function (OF) is the usual chisquare function and if the user wants the usual one-standard-deviation errors, then the error definition should be 1.0. If OF is a negative-log-likelihood function, then 0.5. If OF is a chisquare, but the user wants two-standard-deviation errors, 4.0.

Definition at line 71 of file Minuit2Minimizer.cpp.

◆ setTolerance()

void Minuit2Minimizer::setTolerance ( double value )

Sets tolerance on the function value at the minimum.

Minimization will stop when the estimated vertical distance to the minimum (EDM) is less than 0.001*tolerance*ErrorDef. Here ErrorDef=1.0 for chi squared fit and ErrorDef=0.5 for negative log likelihood fit.

Definition at line 81 of file Minuit2Minimizer.cpp.

◆ setPrecision()

void Minuit2Minimizer::setPrecision ( double value )

Sets relative floating point arithmetic precision.

Should be adjusted when the user knows that objectiove function value is not calculated to the nominal machine accuracy. Typical values are between 10^-5 and 10^-14.

Definition at line 91 of file Minuit2Minimizer.cpp.

◆ setPrintLevel()

void Minuit2Minimizer::setPrintLevel ( int value )

Sets minimizer internal print level.

Default value is 0 (silent).

Definition at line 101 of file Minuit2Minimizer.cpp.

The documentation for this class was generated from the following files:

/home/www/ba/Fit/RootAdapter/Minuit2Minimizer.h
/home/www/ba/Fit/RootAdapter/Minuit2Minimizer.cpp

Public Member Functions

Protected Member Functions

Additional Inherited Members