MLFitter Class Reference

High-level driver for likelihood minimization and 2D contour extraction. More...

#include <Fit.h>

Public Member Functions
	MLFitter (std::shared_ptr< LikelihoodContext > ctx, const ModelFn &model, MLFitOptions options={})
	Constructs a fitter from a likelihood context and model function.
	MLFitter (std::shared_ptr< BaseLikelihood > like, MLFitOptions options={})
	Constructs a fitter from an already configured base likelihood.
FitResult	maximum_likelihood_fit (const std::vector< double > &p0)
	Runs the global maximum-likelihood fit.
Contour	contour (std::size_t x_id, std::size_t y_id, double z, std::array< double, 4 > bounds, ContourOptions options) const
	Computes a 2D confidence contour after the global fit.

Detailed Description

High-level driver for likelihood minimization and 2D contour extraction.

MLFitter owns a profileable likelihood, performs the global maximum- likelihood fit, stores the best-fit state, and subsequently builds confidence contours for selected fit-parameter pairs.

Typical usage:

• construct the fitter from a LikelihoodContext and a model function,
• call maximum_likelihood_fit,
• call contour for parameter pairs of interest.

Note

Contour computation requires a successful prior call to maximum_likelihood_fit.

Definition at line 129 of file Fit.h.

Constructor & Destructor Documentation

MLFitter() [1/2]

MLFitter::MLFitter	(	std::shared_ptr< LikelihoodContext >	ctx,
		const ModelFn &	model,
		MLFitOptions	options = {}
	)

Constructs a fitter from a likelihood context and model function.

A BaseLikelihood is created internally. The number of fit parameters is inferred from ctx->fp_defs.

Parameters

ctx	Likelihood context containing distributions, observations and parameter definitions.
model	Model function mapping fit and nuisance parameters to predicted observables.
options	Fit options controlling the minimization and covariance extraction.

Definition at line 376 of file Fit.cpp.

MLFitter() [2/2]

MLFitter::MLFitter ( std::shared_ptr< BaseLikelihood >  like, MLFitOptions  options = {}  )

explicit

Constructs a fitter from an already configured base likelihood.

Parameters

like	Likelihood object used by the fitter.
options	Fit options controlling the minimization and covariance extraction.

Exceptions

std::invalid_argument

if like is null.

Definition at line 383 of file Fit.cpp.

Member Function Documentation

contour()

Contour MLFitter::contour	(	std::size_t	x_id,
		std::size_t	y_id,
		double	z,
		std::array< double, 4 >	bounds,
		ContourOptions	options
	)		const

Computes a 2D confidence contour after the global fit.

Parameters

x_id	Index of the first fit parameter in the global fit-parameter block.
y_id	Index of the second fit parameter in the global fit-parameter block.
z	Gaussian-equivalent confidence level in standard deviations.
bounds	Contour bounds ordered as {xmin, xmax, ymin, ymax}.
options	Contour and profiling options.

Returns

Extracted contour at the requested level.

Precondition

maximum_likelihood_fit must have been called successfully.

Definition at line 545 of file Fit.cpp.

maximum_likelihood_fit()

FitResult MLFitter::maximum_likelihood_fit

(

const std::vector< double > &

p0

)

Runs the global maximum-likelihood fit.

The fit parameters are initialized from p0, while nuisance parameters keep the values stored in the likelihood parameter definitions. The returned FitResult contains the best-fit values, profiled nuisance estimates, minimum NLL, and covariance-derived uncertainties when available.

Parameters

p0	Initial values for the fit-parameter block.

Returns

Global fit result.

Definition at line 392 of file Fit.cpp.

---

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

• /home/runner/work/Hyperiso/Hyperiso/Hyperiso/Hyperiso/core/src/Statistic/domain/Fit.h
• /home/runner/work/Hyperiso/Hyperiso/Hyperiso/Hyperiso/core/src/Statistic/domain/Fit.cpp