#include <AbsCopulaSmootherBase.hh>

Inheritance diagram for npstat::AbsCopulaSmootherBase:

Public Member Functions
	AbsCopulaSmootherBase (const unsigned *nBinsInEachDim, unsigned dim, double tolerance, unsigned maxNormCycles)

unsigned	dim () const

ArrayShape	copulaShape () const

void	setArchive (gs::AbsArchive ar, const char category=0)

template<class Point >
const HistoND< double > &	smooth (unsigned long uniqueId, std::vector< OrderedPointND< Point > > &in, double *bandwidthUsed=0)

template<class Point >
const HistoND< double > &	weightedSmooth (unsigned long uniqueId, const std::vector< std::pair< const Point , double > > &in, const unsigned dimsToUse, unsigned nDimsToUse, double *bandwidthUsed=0)

Detailed Description

Interface definition for classes which build discrete copulas out of data samples (typically, by LOrPE, KDE, or some other nonparametric method). The derived classes are supposed to override the "smoothHistogram" member function.

Constructor & Destructor Documentation

◆ AbsCopulaSmootherBase()

npstat::AbsCopulaSmootherBase::AbsCopulaSmootherBase	(	const unsigned *	nBinsInEachDim,
		unsigned	dim,
		double	tolerance,
		unsigned	maxNormCycles
	)

Parameters "tolerance" and "maxNormaCycles" define how many copula normalization cycles to perform after smoothing. These cycles make sure that the resulting distribution has uniform marginal densities. The cycles are needed because, in general, marginals are not preserved by smoothing procedures.

Member Function Documentation

◆ copulaShape()

ArrayShape npstat::AbsCopulaSmootherBase::copulaShape ( ) const

inline

Return the shape of the copula array

◆ setArchive()

void npstat::AbsCopulaSmootherBase::setArchive	(	gs::AbsArchive *	ar,
		const char *	category = `0`
	)

Set the archive for storing the histograms

◆ smooth()

template<class Point >

const HistoND<double>& npstat::AbsCopulaSmootherBase::smooth	(	unsigned long	uniqueId,
		std::vector< OrderedPointND< Point > > &	in,
		double *	bandwidthUsed = `0`
	)

Smoothing function for unweighted samples

◆ weightedSmooth()