Default implementation of a profileable negative log-likelihood. More...

#include <BaseLikelihood.h>

Inheritance diagram for BaseLikelihood:

[legend]

Collaboration diagram for BaseLikelihood:

[legend]

Public Member Functions
	BaseLikelihood (const ModelFn &model, std::shared_ptr< LikelihoodContext > ctx, size_t p_dim)
	Constructs a likelihood from a model and statistical context.

double	nll (const std::vector< double > &theta) const override
	Evaluates the negative log-likelihood for a full parameter vector.

std::vector< fit_app::ParameterDefinition >	get_param_defs () const override
	Returns all parameter definitions in likelihood-vector order.

std::size_t	dim () const override
	Returns the total likelihood dimension.

void	enable_debug_trace (std::size_t max_evals=25)
	Enables debug tracing for likelihood evaluations.

void	disable_debug_trace ()
	Disables debug tracing of likelihood evaluations.

std::size_t	p_dimension () const override
	Returns the dimension of the fitted-parameter block.

std::size_t	eta_dimension () const override
	Returns the dimension of the nuisance-parameter block.

std::vector< double >	central_p () const override
	Returns the central values of the fitted parameters.

std::vector< double >	central_eta () const override
	Returns the central values of the nuisance parameters.

std::vector< double >	predict (const std::vector< double > &p, const std::vector< double > &eta) const override

std::vector< double >	residuals (const std::vector< double > &p, const std::vector< double > &eta) const override

double	nll_from_split (const std::vector< double > &p, const std::vector< double > &eta) const override

RealMatrix	observable_curvature (const std::vector< double > &r) const override

RealMatrix	nuisance_curvature (const std::vector< double > &eta) const override

Public Member Functions inherited from ILikelihood
virtual	~ILikelihood ()=default
	Virtual destructor for safe polymorphic deletion.

Protected Attributes
std::shared_ptr< LikelihoodContext >	ctx
	Shared statistical context used by the likelihood.

ModelFn	model
	Model function evaluated by the likelihood.

std::size_t	p_dim
	Dimension of the fitted-parameter block.

bool	debug_trace_enabled_ = false
	Whether evaluation debug tracing is enabled.

std::size_t	debug_trace_max_evals_ = 0
	Maximum number of traced evaluations.

std::size_t	debug_eval_count_ = 0
	Number of evaluations already traced.

bool	debug_have_ref_theta_ = false
	Whether the debug reference parameter vector is set.

bool	debug_have_ref_res_ = false
	Whether the debug reference residual vector is set.

std::vector< double >	debug_ref_theta_
	First traced parameter vector used as debug reference.

std::vector< double >	debug_ref_res_
	First traced residual vector used as debug reference.

Detailed Description

Default implementation of a profileable negative log-likelihood.

BaseLikelihood evaluates

$-\log L(p, \eta) = -\left[\log f_{obs}(m(p,\eta) - y_{obs}) + \log f_{nuis}(\eta)\right],$

where $m(p,\eta)$ is the model prediction and $y_{obs}$ are the nominal experimental observable values stored in LikelihoodContext.

The class also exposes diagnostic helpers for profiling workflows, including central values, split-parameter evaluation, residual computation and local curvature matrices.

Note: If model evaluation or density evaluation produces a non-finite value, the implementation returns a large finite penalty value instead of propagating NaNs through the optimizer.

Definition at line 75 of file BaseLikelihood.h.

Constructor & Destructor Documentation

◆ BaseLikelihood()

BaseLikelihood::BaseLikelihood	(	const ModelFn &	model,
		std::shared_ptr< LikelihoodContext >	ctx,
		size_t	p_dim
	)

Constructs a likelihood from a model and statistical context.

Parameters

model	Model function used to compute observable predictions.
ctx	Shared likelihood context containing distributions and parameter definitions.
p_dim	Dimension of the fitted-parameter block .

Definition at line 3 of file BaseLikelihood.cpp.

Member Function Documentation

◆ central_eta()

std::vector< double > BaseLikelihood::central_eta ( ) const

overridevirtual

Returns the central values of the nuisance parameters.

Returns: Vector containing the nominal values of $\eta$ .

Implements IProfileableLikelihood.

Definition at line 137 of file BaseLikelihood.cpp.

◆ central_p()

std::vector< double > BaseLikelihood::central_p ( ) const

overridevirtual

Returns the central values of the fitted parameters.

Returns: Vector containing the nominal values of .

Implements IProfileableLikelihood.

Definition at line 126 of file BaseLikelihood.cpp.

◆ dim()

std::size_t BaseLikelihood::dim ( ) const

overridevirtual

Returns the total likelihood dimension.

Returns: Sum of fitted-parameter and nuisance-parameter dimensions.

Implements ILikelihood.

Reimplemented in ChiSquaredLikelihood.

Definition at line 114 of file BaseLikelihood.cpp.

◆ disable_debug_trace()

void BaseLikelihood::disable_debug_trace ( )

inline

Disables debug tracing of likelihood evaluations.

Definition at line 144 of file BaseLikelihood.h.

◆ enable_debug_trace()

void BaseLikelihood::enable_debug_trace ( std::size_t max_evals = 25 )

inline

Enables debug tracing for likelihood evaluations.

When enabled, the first max_evals evaluations print diagnostic information such as the negative log-likelihood value, observable and nuisance log-density terms, and maximum shifts relative to the first traced point.

Parameters

max_evals Maximum number of evaluations to print.

Definition at line 131 of file BaseLikelihood.h.

◆ eta_dimension()

std::size_t BaseLikelihood::eta_dimension ( ) const

overridevirtual

Returns the dimension of the nuisance-parameter block.

Returns: Number of scalar parameters in $\eta$ .

Implements IProfileableLikelihood.

Definition at line 122 of file BaseLikelihood.cpp.

◆ get_param_defs()

std::vector< fit_app::ParameterDefinition > BaseLikelihood::get_param_defs ( ) const

inlineoverridevirtual

Returns all parameter definitions in likelihood-vector order.

The returned list contains fitted-parameter definitions first, followed by nuisance-parameter definitions.

Returns: Ordered list of fitted and nuisance parameter definitions.

Implements ILikelihood.

Definition at line 108 of file BaseLikelihood.h.

◆ nll()

double BaseLikelihood::nll ( const std::vector< double > & theta ) const

overridevirtual

Evaluates the negative log-likelihood for a full parameter vector.

The input vector is split as $\theta = (p, \eta)$ using the fitted parameter dimension provided at construction.

Parameters

theta Full parameter vector ordered as fitted parameters followed by nuisance parameters.

Returns: Negative log-likelihood value, or a large finite penalty if the evaluation is numerically invalid.

Implements ILikelihood.

Reimplemented in ChiSquaredLikelihood.

Definition at line 68 of file BaseLikelihood.cpp.

◆ nll_from_split()

double BaseLikelihood::nll_from_split	(	const std::vector< double > &	p,
		const std::vector< double > &	eta
	)		const

overridevirtual

Implements IProfileableLikelihood.

Definition at line 172 of file BaseLikelihood.cpp.

◆ nuisance_curvature()

RealMatrix BaseLikelihood::nuisance_curvature ( const std::vector< double > & eta ) const

overridevirtual

Implements IProfileableLikelihood.

Reimplemented in ChiSquaredLikelihood.

Definition at line 206 of file BaseLikelihood.cpp.

◆ observable_curvature()

RealMatrix BaseLikelihood::observable_curvature ( const std::vector< double > & r ) const

overridevirtual

Implements IProfileableLikelihood.

Reimplemented in ChiSquaredLikelihood.

Definition at line 184 of file BaseLikelihood.cpp.

◆ p_dimension()

std::size_t BaseLikelihood::p_dimension ( ) const

overridevirtual

Returns the dimension of the fitted-parameter block.

Returns: Number of scalar parameters in .

Implements IProfileableLikelihood.

Definition at line 118 of file BaseLikelihood.cpp.

◆ predict()

std::vector< double > BaseLikelihood::predict	(	const std::vector< double > &	p,
		const std::vector< double > &	eta
	)		const

overridevirtual

Implements IProfileableLikelihood.

Definition at line 148 of file BaseLikelihood.cpp.

◆ residuals()

std::vector< double > BaseLikelihood::residuals	(	const std::vector< double > &	p,
		const std::vector< double > &	eta
	)		const

overridevirtual

Implements IProfileableLikelihood.

Definition at line 155 of file BaseLikelihood.cpp.

Member Data Documentation

◆ ctx

std::shared_ptr<LikelihoodContext> BaseLikelihood::ctx

protected

Shared statistical context used by the likelihood.

Definition at line 189 of file BaseLikelihood.h.

◆ debug_eval_count_

std::size_t BaseLikelihood::debug_eval_count_ = 0

mutableprotected

Number of evaluations already traced.

Definition at line 195 of file BaseLikelihood.h.

◆ debug_have_ref_res_

bool BaseLikelihood::debug_have_ref_res_ = false

mutableprotected

Whether the debug reference residual vector is set.

Definition at line 198 of file BaseLikelihood.h.

◆ debug_have_ref_theta_

bool BaseLikelihood::debug_have_ref_theta_ = false

mutableprotected

Whether the debug reference parameter vector is set.

Definition at line 197 of file BaseLikelihood.h.

◆ debug_ref_res_

std::vector<double> BaseLikelihood::debug_ref_res_

mutableprotected

First traced residual vector used as debug reference.

Definition at line 200 of file BaseLikelihood.h.

◆ debug_ref_theta_

std::vector<double> BaseLikelihood::debug_ref_theta_

mutableprotected

First traced parameter vector used as debug reference.

Definition at line 199 of file BaseLikelihood.h.

◆ debug_trace_enabled_

bool BaseLikelihood::debug_trace_enabled_ = false

mutableprotected

Whether evaluation debug tracing is enabled.

Definition at line 193 of file BaseLikelihood.h.

◆ debug_trace_max_evals_

std::size_t BaseLikelihood::debug_trace_max_evals_ = 0

mutableprotected

Maximum number of traced evaluations.

Definition at line 194 of file BaseLikelihood.h.

◆ model

ModelFn BaseLikelihood::model

protected

Model function evaluated by the likelihood.

Definition at line 190 of file BaseLikelihood.h.

◆ p_dim

std::size_t BaseLikelihood::p_dim

protected

Dimension of the fitted-parameter block.

Definition at line 191 of file BaseLikelihood.h.

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

/home/runner/work/Hyperiso/Hyperiso/Hyperiso/Hyperiso/core/src/Statistic/domain/Likelihood/BaseLikelihood.h
/home/runner/work/Hyperiso/Hyperiso/Hyperiso/Hyperiso/core/src/Statistic/domain/Likelihood/BaseLikelihood.cpp

Public Member Functions

Protected Attributes

Detailed Description

Constructor & Destructor Documentation

◆ BaseLikelihood()

Member Function Documentation

◆ central_eta()

◆ central_p()

◆ dim()

◆ disable_debug_trace()

◆ enable_debug_trace()

◆ eta_dimension()

◆ get_param_defs()

◆ nll()

◆ nll_from_split()

◆ nuisance_curvature()

◆ observable_curvature()

◆ p_dimension()

◆ predict()

◆ residuals()

Member Data Documentation

◆ ctx

◆ debug_eval_count_

◆ debug_have_ref_res_

◆ debug_have_ref_theta_

◆ debug_ref_res_

◆ debug_ref_theta_

◆ debug_trace_enabled_

◆ debug_trace_max_evals_

◆ model

◆ p_dim