Class to search optimal regularization parameter using GCV. More...

#include <num_collect/regularization/implicit_gcv.h>

Inheritance diagram for num_collect::regularization::implicit_gcv< Solver >:

Collaboration diagram for num_collect::regularization::implicit_gcv< Solver >:

Public Types
using	data_type = typename solver_type::data_type
	Type of data.
using	scalar_type = typename solver_type::scalar_type
	Type of scalars.
using	solver_type = Solver
	Type of solvers.

Public Member Functions
	implicit_gcv (solver_type &solver, const data_type &data, const data_type &initial_solution)
	Constructor.
auto	max_evaluations (index_type value) -> implicit_gcv &
	Set the maximum number of evaluations of GCV.
auto	noise_rate (scalar_type value) -> implicit_gcv &
	Set the rate of noise.
auto	num_samples (index_type value) -> implicit_gcv &
	Set the number of samples for approximation of denominator of GCV.
auto	opt_param () const -> scalar_type
	Get the optimal regularization parameter.
void	search ()
	Search the optimal regularization parameter.
void	solve (data_type &solution) const
	Solver with the optimal regularization parameter.
Public Member Functions inherited from num_collect::logging::logging_mixin
	logging_mixin (log_tag_view tag)
	Constructor.
template<typename Child>
void	configure_child_algorithm_logger_if_exists (Child &child)
	Configure a logger of a child algorithm if exists.
auto	logger () const noexcept -> const num_collect::logging::logger &
	Access to the logger.
auto	logger () noexcept -> num_collect::logging::logger &
	Access to the logger.

Private Attributes
implicit_gcv_calculator< solver_type >	calculator_
	Calculator of GCV.
scalar_type	opt_param_ {}
	Optimal regularization parameter.
opt::gaussian_process_optimizer< opt::any_objective_function< scalar_type(scalar_type)> >	optimizer_ {}
	Optimizer.

Detailed Description

template<concepts::implicit_regularized_solver Solver>
class num_collect::regularization::implicit_gcv< Solver >

Class to search optimal regularization parameter using GCV.

Template Parameters

Solver Type of solvers.

Examples: regularization/image_denoising_fista.cpp, regularization/image_denoising_tgv2_admm.cpp, and regularization/image_denoising_tv_admm.cpp.

Definition at line 227 of file implicit_gcv.h.

Member Typedef Documentation

◆ data_type

template<concepts::implicit_regularized_solver Solver>

using num_collect::regularization::implicit_gcv< Solver >::data_type = typename solver_type::data_type

Type of data.

Definition at line 236 of file implicit_gcv.h.

◆ scalar_type

template<concepts::implicit_regularized_solver Solver>

using num_collect::regularization::implicit_gcv< Solver >::scalar_type = typename solver_type::scalar_type

Type of scalars.

Definition at line 233 of file implicit_gcv.h.

◆ solver_type

template<concepts::implicit_regularized_solver Solver>

using num_collect::regularization::implicit_gcv< Solver >::solver_type = Solver

Type of solvers.

Definition at line 230 of file implicit_gcv.h.

Constructor & Destructor Documentation

◆ implicit_gcv()

template<concepts::implicit_regularized_solver Solver>

num_collect::regularization::implicit_gcv< Solver >::implicit_gcv	(	solver_type &	solver,
		const data_type &	data,
		const data_type &	initial_solution )

inline

Constructor.

Parameters

[in]	solver	Solver.
[in]	data	Data.
[in]	initial_solution	Initial solution.

Definition at line 245 of file implicit_gcv.h.

Member Function Documentation

◆ max_evaluations()

template<concepts::implicit_regularized_solver Solver>

auto num_collect::regularization::implicit_gcv< Solver >::max_evaluations ( index_type value ) -> implicit_gcv&

inline

Set the maximum number of evaluations of GCV.

Parameters

[in] value Value.

Returns: This object.

Definition at line 330 of file implicit_gcv.h.

◆ noise_rate()

template<concepts::implicit_regularized_solver Solver>

auto num_collect::regularization::implicit_gcv< Solver >::noise_rate ( scalar_type value ) -> implicit_gcv&

inline

Set the rate of noise.

Parameters

[in] value Value.

Returns: This object.

Definition at line 308 of file implicit_gcv.h.

◆ num_samples()

template<concepts::implicit_regularized_solver Solver>

auto num_collect::regularization::implicit_gcv< Solver >::num_samples ( index_type value ) -> implicit_gcv&

inline

Set the number of samples for approximation of denominator of GCV.

Parameters

[in] value Value.

Returns: This object.

Definition at line 319 of file implicit_gcv.h.

◆ opt_param()

template<concepts::implicit_regularized_solver Solver>

auto num_collect::regularization::implicit_gcv< Solver >::opt_param ( ) const -> scalar_type

inlinenodiscard

Get the optimal regularization parameter.

Returns: Optimal regularization parameter.

Definition at line 293 of file implicit_gcv.h.

◆ search()

template<concepts::implicit_regularized_solver Solver>

void num_collect::regularization::implicit_gcv< Solver >::search ( )

inline

Search the optimal regularization parameter.

Examples: regularization/image_denoising_fista.cpp, regularization/image_denoising_tgv2_admm.cpp, and regularization/image_denoising_tv_admm.cpp.

Definition at line 257 of file implicit_gcv.h.

◆ solve()

template<concepts::implicit_regularized_solver Solver>

void num_collect::regularization::implicit_gcv< Solver >::solve ( data_type & solution ) const

inline

Solver with the optimal regularization parameter.

Template Parameters

Solution Type of the solution.

Parameters

[out] solution Solution.

Examples: regularization/image_denoising_fista.cpp, regularization/image_denoising_tgv2_admm.cpp, and regularization/image_denoising_tv_admm.cpp.

Definition at line 296 of file implicit_gcv.h.

Member Data Documentation

◆ calculator_

template<concepts::implicit_regularized_solver Solver>

implicit_gcv_calculator<solver_type> num_collect::regularization::implicit_gcv< Solver >::calculator_

private

Calculator of GCV.

Definition at line 337 of file implicit_gcv.h.

◆ opt_param_

template<concepts::implicit_regularized_solver Solver>

scalar_type num_collect::regularization::implicit_gcv< Solver >::opt_param_ {}

private

Optimal regularization parameter.

Definition at line 345 of file implicit_gcv.h.

◆ optimizer_

template<concepts::implicit_regularized_solver Solver>

opt::gaussian_process_optimizer< opt::any_objective_function<scalar_type(scalar_type)> > num_collect::regularization::implicit_gcv< Solver >::optimizer_ {}

private

Optimizer.

Definition at line 342 of file implicit_gcv.h.

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

include/num_collect/regularization/implicit_gcv.h

Public Types

Public Member Functions

Private Attributes

Detailed Description

Member Typedef Documentation

◆ data_type

◆ scalar_type

◆ solver_type

Constructor & Destructor Documentation

◆ implicit_gcv()

Member Function Documentation

◆ max_evaluations()

◆ noise_rate()

◆ num_samples()

◆ opt_param()

◆ search()

◆ solve()

Member Data Documentation

◆ calculator_

◆ opt_param_

◆ optimizer_