npstat::SequentialPolyFilterND Class Reference

#include <SequentialPolyFilterND.hh>

Inheritance diagram for npstat::SequentialPolyFilterND:

Public Member Functions
	SequentialPolyFilterND (const LocalPolyFilter1D **filters, unsigned nFilters, bool takeOwnership)
	SequentialPolyFilterND (const std::vector< AllSymbetaParams1D > &params)
	SequentialPolyFilterND (const SequentialPolyFilterND &)
SequentialPolyFilterND &	operator= (const SequentialPolyFilterND &)
bool	operator== (const SequentialPolyFilterND &r) const
bool	operator!= (const SequentialPolyFilterND &r) const
unsigned	dim () const
bool	takesOwnership () const
ArrayShape	dataShape () const
const LocalPolyFilter1D *	filter (unsigned dimNumber) const
ArrayND< double >	getFilter (const unsigned *index, unsigned lenIndex) const
Matrix< double >	getFilterMatrix () const
template<class Triplet >
CPP11_auto_ptr< std::vector< Triplet > >	sparseFilterTriplets () const
template<class Array >
bool	isCompatible (const Array &in) const
double	selfContribution (const unsigned *index, unsigned lenIndex) const
double	linearSelfContribution (unsigned long index) const
template<class ArrIn , class ArrOut >
void	filter (const ArrIn &in, ArrOut *out) const
template<class ArrIn , class ArrOut >
void	convolve (const ArrIn &in, ArrOut *out) const
virtual gs::ClassId	classId () const
virtual bool	write (std::ostream &os) const

Static Public Member Functions
static const char *	classname ()
static unsigned	version ()
static SequentialPolyFilterND *	read (const gs::ClassId &id, std::istream &in)

Detailed Description

This class performs local polynomial filtering in multiple dimensions by sequential application of one-dimensional filters

Constructor & Destructor Documentation

SequentialPolyFilterND()

npstat::SequentialPolyFilterND::SequentialPolyFilterND	(	const LocalPolyFilter1D **	filters,
		unsigned	nFilters,
		bool	takeOwnership
	)

Main constructor. The arguments are as follows:

filters – Array of pointers to LocalPolyFilter1D objects which will be used to perform filtering in each dimension. The filters should be compatible with the expected span of the data arrays.

nFilters – Number of elements in the "filters" array. Should be equal to the dimensionality of the data arrays which will be processed by the "filter" function.

takeOwnership – If this argument is set to "true", we will call the "delete" operator on each element of the "filters" array the destructor.

Member Function Documentation

convolve()

template<class ArrIn , class ArrOut >

void npstat::SequentialPolyFilterND::convolve	(	const ArrIn &	in,
		ArrOut *	out
	)		const

A diffent filtering method in which the shapes of the kernels are determined by the positions of the "sources" (i.e., sample points) instead of the positions at which the density (or response) is estimated. Note that elements of "out" array themselves are used as result accumulators.

dataShape()

ArrayShape npstat::SequentialPolyFilterND::dataShape ( ) const

virtual

Required shape of the data array

Implements npstat::AbsPolyFilterND.

dim()

unsigned npstat::SequentialPolyFilterND::dim ( ) const

inlinevirtual

Inspect object properties

Implements npstat::AbsPolyFilterND.

filter() [1/2]

template<class ArrIn , class ArrOut >

void npstat::SequentialPolyFilterND::filter	(	const ArrIn &	in,
		ArrOut *	out
	)		const

This method performs the filtering

filter() [2/2]

const LocalPolyFilter1D* npstat::SequentialPolyFilterND::filter

(

unsigned

dimNumber

)

const

Get the filter for the given dimension

getFilter()

ArrayND<double> npstat::SequentialPolyFilterND::getFilter	(	const unsigned *	index,
		unsigned	lenIndex
	)		const

Get the effective multivariate filter coefficients for the given grid point. Note that this operation is quite slow (it involves building a multivariate array using outer products of 1-d arrays) and should not be used when fast performance is essential.

getFilterMatrix()

Matrix<double> npstat::SequentialPolyFilterND::getFilterMatrix

(

)

const

Get the complete effective filter matrix

isCompatible()

template<class Array >

bool npstat::SequentialPolyFilterND::isCompatible

(

const Array &

in

)

const

Check compatibility of an array with the filter

linearSelfContribution()

double npstat::SequentialPolyFilterND::linearSelfContribution ( unsigned long  index ) const

virtual

Contribution of a single point into the density estimate using the linear index of the point

Implements npstat::AbsPolyFilterND.

selfContribution()

double npstat::SequentialPolyFilterND::selfContribution ( const unsigned *  index, unsigned  lenIndex  ) const

virtual

Contribution of a single point into the density estimate at that point (not normalized). This is needed for various leaving-one-out cross validation procedures.

Implements npstat::AbsPolyFilterND.

sparseFilterTriplets()

template<class Triplet >

CPP11_auto_ptr<std::vector<Triplet> > npstat::SequentialPolyFilterND::sparseFilterTriplets

(

)

const

Get the info needed to construct the sparse filter matrix

---

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

• stat/SequentialPolyFilterND.hh