num_collect::rbf::gaussian_process_interpolator< FunctionSignature, RBF, KernelMatrixType, DistanceFunction > Class Template Reference

Class to interpolate using Gaussian process. More...

#include <num_collect/rbf/gaussian_process_interpolator.h>

Inheritance diagram for num_collect::rbf::gaussian_process_interpolator< FunctionSignature, RBF, KernelMatrixType, DistanceFunction >:

Collaboration diagram for num_collect::rbf::gaussian_process_interpolator< FunctionSignature, RBF, KernelMatrixType, DistanceFunction >:

Public Types
using	base_type
	Type of the base class.

Public Member Functions
void	compute (const std::vector< variable_type > &variables, const function_value_vector_type &function_values)
	Compute parameters for interpolation.
auto	evaluate_mean_and_variance_on (const variable_type &variable) const -> std::pair< function_value_type, function_value_type >
	Evaluate mean and variance in the gaussian process for a variable.

Private Attributes
function_value_type	common_coeff_ {}
	Common coefficients for RBF.

Static Private Attributes
static constexpr auto	reg_param = static_cast<kernel_value_type>(0)
	Regularization parameter.

Detailed Description

template<typename FunctionSignature, concepts::rbf RBF = rbfs::gaussian_rbf<impl::get_default_scalar_type<FunctionSignature>>, kernel_matrix_type KernelMatrixType = kernel_matrix_type::dense, concepts::distance_function DistanceFunction = distance_functions::euclidean_distance_function< impl::get_variable_type_t<FunctionSignature>>>
class num_collect::rbf::gaussian_process_interpolator< FunctionSignature, RBF, KernelMatrixType, DistanceFunction >

Class to interpolate using Gaussian process.

Template Parameters

FunctionSignature	Signature of the function to interpolate. (Example: double(double), double(Eigen::Vector3d), ...)
RBF	Type of the RBF.
KernelMatrixType	Type of kernel matrices.
DistanceFunction	Type of the distance function.

Note

This class adds some functions for Gaussian process to num_collect::rbf::rbf_interpolator class.

Definition at line 60 of file gaussian_process_interpolator.h.

Member Typedef Documentation

base_type

template<typename FunctionSignature , concepts::rbf RBF = rbfs::gaussian_rbf<impl::get_default_scalar_type<FunctionSignature>>, kernel_matrix_type KernelMatrixType = kernel_matrix_type::dense, concepts::distance_function DistanceFunction = distance_functions::euclidean_distance_function< impl::get_variable_type_t<FunctionSignature>>>

using num_collect::rbf::gaussian_process_interpolator< FunctionSignature, RBF, KernelMatrixType, DistanceFunction >::base_type

Initial value:

 global_rbf_interpolator<FunctionSignature, RBF,
        KernelMatrixType, DistanceFunction>

Type of the base class.

Definition at line 68 of file gaussian_process_interpolator.h.

Member Function Documentation

compute()

template<typename FunctionSignature , concepts::rbf RBF = rbfs::gaussian_rbf<impl::get_default_scalar_type<FunctionSignature>>, kernel_matrix_type KernelMatrixType = kernel_matrix_type::dense, concepts::distance_function DistanceFunction = distance_functions::euclidean_distance_function< impl::get_variable_type_t<FunctionSignature>>>

void num_collect::rbf::gaussian_process_interpolator< FunctionSignature, RBF, KernelMatrixType, DistanceFunction >::compute ( const std::vector< variable_type > & variables, const function_value_vector_type & function_values )

inline

Compute parameters for interpolation.

Parameters

[in]	variables	Variables.
[in]	function_values	Function values.

Note

Pointer to the variables are saved internally, so do not destruct it.

Definition at line 89 of file gaussian_process_interpolator.h.

evaluate_mean_and_variance_on()

template<typename FunctionSignature , concepts::rbf RBF = rbfs::gaussian_rbf<impl::get_default_scalar_type<FunctionSignature>>, kernel_matrix_type KernelMatrixType = kernel_matrix_type::dense, concepts::distance_function DistanceFunction = distance_functions::euclidean_distance_function< impl::get_variable_type_t<FunctionSignature>>>

auto num_collect::rbf::gaussian_process_interpolator< FunctionSignature, RBF, KernelMatrixType, DistanceFunction >::evaluate_mean_and_variance_on ( const variable_type & variable ) const -> std::pair<function_value_type, function_value_type>

inlinenodiscard

Evaluate mean and variance in the gaussian process for a variable.

Parameters

[in]

variable

Variable on which the function value is interpolated.

Returns

Mean and variance.

Definition at line 102 of file gaussian_process_interpolator.h.

Member Data Documentation

common_coeff_

template<typename FunctionSignature , concepts::rbf RBF = rbfs::gaussian_rbf<impl::get_default_scalar_type<FunctionSignature>>, kernel_matrix_type KernelMatrixType = kernel_matrix_type::dense, concepts::distance_function DistanceFunction = distance_functions::euclidean_distance_function< impl::get_variable_type_t<FunctionSignature>>>

function_value_type num_collect::rbf::gaussian_process_interpolator< FunctionSignature, RBF, KernelMatrixType, DistanceFunction >::common_coeff_ {}

private

Common coefficients for RBF.

Definition at line 134 of file gaussian_process_interpolator.h.

reg_param

template<typename FunctionSignature , concepts::rbf RBF = rbfs::gaussian_rbf<impl::get_default_scalar_type<FunctionSignature>>, kernel_matrix_type KernelMatrixType = kernel_matrix_type::dense, concepts::distance_function DistanceFunction = distance_functions::euclidean_distance_function< impl::get_variable_type_t<FunctionSignature>>>

auto num_collect::rbf::gaussian_process_interpolator< FunctionSignature, RBF, KernelMatrixType, DistanceFunction >::reg_param = static_cast<kernel_value_type>(0)

staticconstexprprivate

Regularization parameter.

Definition at line 129 of file gaussian_process_interpolator.h.

---

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

• include/num_collect/rbf/gaussian_process_interpolator.h