Derived from 'PPC_KO_base' class for computing PPCKO algorithm with cross-validation on the number of retained PPCs. More...

#include <PPC_KO_CV_k.hpp>

Inheritance diagram for PPC_KO_CV_k< solver, k_imp, valid_err_ret, cv_strat, cv_err_eval >:

Public Member Functions
template<typename STOR_OBJ>
	PPC_KO_CV_k (STOR_OBJ &&X, std::vector< int > &k_s, double alpha, double toll, int min_size_ts, int max_size_ts, int number_threads)
	Constructor for cv version on number of retained PPCs.

void	solving ()
	Method to perform PPCKO if cv is performed on the number of retained PPCs (k will be always imposed)

Public Member Functions inherited from PPC_KO_base< PPC_KO_CV_k< solver, k_imp, valid_err_ret, cv_strat, cv_err_eval >, solver, k_imp, valid_err_ret, cv_strat, cv_err_eval >
	PPC_KO_base (STOR_OBJ &&X, int number_threads)
	Constructor: centers data, evaluate mean function, sample covariance, sample cross-covariance and its square.

std::size_t	m () const
	Getter for the number of evaluation of the curve/surface.

std::size_t	n () const
	Getter for the number of time instants.

KO_Traits::StoringMatrix	X () const
	Getter for fts data matrix (centered)

KO_Traits::StoringArray	means () const
	Getter for the mean function.

KO_Traits::StoringMatrix	Cov () const
	Getter for the covariance operator estimate.

double	trace_cov () const
	Getter for the covariance operator estimate's trace.

KO_Traits::StoringMatrix &	CovReg ()
	Setter for the regularized sample covariance operator.

KO_Traits::StoringMatrix	rho () const
	Getter for the autoregressive operator estimate.

KO_Traits::StoringMatrix	a () const
	Getter for the predictive loadings (PPCs directions)

KO_Traits::StoringMatrix	b () const
	Getter for the predictive factors factor (PPCs weights)

std::vector< double >	explanatory_power () const
	Getter for the cumulative explanatory power of the PPCs.

double	alpha () const
	Getter for the regularization parameter.

double &	alpha ()
	Setter for the regularization parameter.

int	k () const
	Getter for the number of retained PPCs.

int &	k ()
	Setter for the number of retained PPCs.

double	threshold_ppc () const
	Getter for the requested explanatory power of the retained PPCs.

double &	threshold_ppc ()
	Setter for the requested explanatory power of the retained PPCs.

valid_err_variant	ValidErr () const
	Getter for the validation errors.

valid_err_variant &	ValidErr ()
	Setter for the validation errors.

int	number_threads () const
	Getter for the number of threads for OMP.

std::tuple< int, KO_Traits::StoringVector, KO_Traits::StoringMatrix >	PPC_retained ()
	Retaining the the PPCs: pairs eigenvalue-eigenvector and their number.

void	KO_algo ()
	Performing PPCKO algorithm once regularization parameter is selected and k or it is fixed or to be retained through explanatory power. Computes PPCs, direction and weight, their number and their cumulative explanatory power, and the estimate of the autoregressive operator.

KO_Traits::StoringArray	prediction () const
	Performs one-step ahead prediction of the fts. The mean function is added.

std::vector< double >	scores () const
	Computes the scores of the PPCs, defined as scalar product between the direction and the fts at the last instant.

std::vector< std::array< double, 2 > >	sd_scores_dir_wei () const
	Computes the standard deviation of the scores of directions and weights.

void	solve ()
	Method to solve PPCKO according to which cross-validation is performed, if any.

Detailed Description

template<SOLVER solver, K_IMP k_imp, VALID_ERR_RET valid_err_ret, CV_STRAT cv_strat, CV_ERR_EVAL cv_err_eval>
class PPC_KO_CV_k< solver, k_imp, valid_err_ret, cv_strat, cv_err_eval >

Derived from 'PPC_KO_base' class for computing PPCKO algorithm with cross-validation on the number of retained PPCs.

Template Parameters

solver	if algorithm solved inverting the regularized covariance or avoiding it through gep (not possible if retaining the number of PPCs with explanatory power criterion)
k_imp	if k is imposed or has to be found through explanatory power criterion
valid_err_ret	if validation error are stored
cv_strat	strategy for splitting training/validation sets
err_eval	how to evaluate the loss between prediction on validation set and validation set

Constructor & Destructor Documentation

◆ PPC_KO_CV_k()

template<SOLVER solver, K_IMP k_imp, VALID_ERR_RET valid_err_ret, CV_STRAT cv_strat, CV_ERR_EVAL cv_err_eval>

template<typename STOR_OBJ>

PPC_KO_CV_k< solver, k_imp, valid_err_ret, cv_strat, cv_err_eval >::PPC_KO_CV_k	(	STOR_OBJ &&	X,
		std::vector< int > &	k_s,
		double	alpha,
		double	toll,
		int	min_size_ts,
		int	max_size_ts,
		int	number_threads )

inline

Constructor for cv version on number of retained PPCs.

Parameters

X	fts
k_s	number of retained PPCs input space
alpha	regularization parameter
toll	the cv on the number of retained PPCs continues only if between two parameters, that are checked in increasing order, the absolute difference between two validation errors is bigger than tolerance*trace(covariance). If not, stops and look for k only between the tested ones
min_size_ts	smallest training set size (number of time instants)
max_size_ts	biggest training set size (number of time instants)
number_threads	number of threads for OMP

Universal constructor: move semantic used to optimazing handling big size objects

Note: eventual usage of 'pragma' directive for OMP

Member Function Documentation

◆ solving()

template<SOLVER solver, K_IMP k_imp, VALID_ERR_RET valid_err_ret, CV_STRAT cv_strat, CV_ERR_EVAL cv_err_eval>

void PPC_KO_CV_k< solver, k_imp, valid_err_ret, cv_strat, cv_err_eval >::solving ( )

inline

Method to perform PPCKO if cv is performed on the number of retained PPCs (k will be always imposed)

Selects the number of PPCs through cv, and then call the .KO_algo() method of the base class

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

src/PPC_KO_CV_k.hpp

Public Member Functions

Detailed Description

Constructor & Destructor Documentation

◆ PPC_KO_CV_k()

Member Function Documentation

◆ solving()