Aggregator class used to declare specialized structures with derivatives computations, using CRTP. More...

#include <basket.h>

Inheritance diagram for Ponca::BasketDiff< BasketType, Type, Ext0, Exts >:

Collaboration diagram for Ponca::BasketDiff< BasketType, Type, Ext0, Exts >:

Public Types
using	Base = typename internal::BasketDiffAggregate< BasketType, Type, Ext0, Exts... >::type

using	WeightFunction = typename BasketType::WeightFunction
	Weight function.

using	DataPoint = typename BasketType::DataPoint
	Point type used for computation.

using	Scalar = typename BasketType::Scalar
	Scalar type used for computation, as defined from Basket.

Public Types inherited from Ponca::BasketComputeObject< BasketDiff< BasketType, Type, Ext0, Exts... >, internal::BasketDiffAggregate< BasketType, Type, Ext0, Exts... >::type >
using	Base = internal::BasketDiffAggregate< BasketType, Type, Ext0, Exts... >::type

using	Derived = BasketDiff< BasketType, Type, Ext0, Exts... >
	<

using	Scalar = typename Base::Scalar
	Alias to the Derived type.

Public Member Functions
bool	addNeighbor (const DataPoint &_nei)
	Add a neighbor to perform the fit.

Public Member Functions inherited from Ponca::BasketComputeObject< BasketDiff< BasketType, Type, Ext0, Exts... >, internal::BasketDiffAggregate< BasketType, Type, Ext0, Exts... >::type >
FIT_RESULT	compute (const IteratorBegin &begin, const IteratorEnd &end)
	Convenience function for STL-like iterators Add neighbors stored in a container using STL-like iterators, and call finalize at the end.

FIT_RESULT	computeWithIds (IndexRange ids, const PointContainer &points)
	Convenience function to iterate over a subset of samples in a PointContainer Add neighbors stored in a PointContainer and sampled using indices stored in ids.

FIT_RESULT	computeMLS (const PointContainer &points, const int mlsIter=5, const Scalar epsilon=Eigen::NumTraits< Scalar >::dummy_precision())

FIT_RESULT	computeWithIdsMLS (const IndexRange &ids, const PointContainer &points, const int mlsIter=5, const Scalar epsilon=Eigen::NumTraits< Scalar >::dummy_precision())

Public Member Functions inherited from Ponca::ComputeObject< Derived >
template<typename Container >
FIT_RESULT	compute (const Container &c)
	Convenience function for STL-like container.

template<typename IteratorBegin , typename IteratorEnd >
FIT_RESULT	compute (const IteratorBegin &, const IteratorEnd &)
	Convenience function for STL-like iterators.

template<typename IndexRange , typename PointContainer >
FIT_RESULT	computeWithIds (IndexRange, const PointContainer &)
	Convenience function to iterate over a subset of samples in a PointContainer.

Additional Inherited Members
Protected Member Functions inherited from Ponca::BasketComputeObject< BasketDiff< BasketType, Type, Ext0, Exts... >, internal::BasketDiffAggregate< BasketType, Type, Ext0, Exts... >::type >
FIT_RESULT	computeMLSImpl (Func &&computeFunc, int mlsIter, Scalar epsilon)
	Computes the fit using the MLS iteration process.

Protected Member Functions inherited from Ponca::ComputeObject< Derived >
Derived &	derived ()
	Retrieve the top layer object Returns a reference to the derived class so that we can use its overwritten methods.

Detailed Description

template<typename BasketType, int Type, template< class, class, int, typename > class Ext0, template< class, class, int, typename > class... Exts>
class Ponca::BasketDiff< BasketType, Type, Ext0, Exts >

Aggregator class used to declare specialized structures with derivatives computations, using CRTP.

Base ComputeObject for the Basket classes.

Implements the compute methods for fitting: compute, computeWithIds, ... Checkout Fitting Module: User Manual for more details

The various implementations of Ponca::Concept are mixed through specializations of the BasketDiff and Basket classes:

typedef
BasketDiff <BasketType,           // Existing Basket, to be differentiated
DiffType,                         // Differentiation space: FitScaleDer, FitSpaceDer, or FitScaleDer|FitSpaceDer
ComputationalDerivativesConcept1, // Implementation of ComputationalDerivativesConcept
ComputationalDerivativesConcept2, // Implementation of ComputationalDerivativesConcept
... ,                             //
> myFitDer;                       // Final structure to fit and derive a primitive over weighted samples

Template Parameters

Derived	Derived class that provides the addNeighbor method (either Basket or BasketDiff)
Base	Base class that provides, through the CRTP the init, startNewPass, addNeighbor and finalize methods
BasketType	Existing Basket, to be differentiated
Type	Differentiation space: FitScaleDer, FitSpaceDer, or FitScaleDer\|FitSpaceDer
Ext0	Implements ComputationalDerivativesConcept
Exts	Implements ComputationalDerivativesConcept (optional)

Definition at line 282 of file basket.h.