RVL::FunctionalProductDomain< Scalar > Class Template Reference

A specialization of Functional which also has partial derivatives. More...

#include <functional.hh>

Inheritance diagram for RVL::FunctionalProductDomain< Scalar >:


Public Member Functions
	FunctionalProductDomain ()
	FunctionalProductDomain (const FunctionalProductDomain< Scalar > &)
virtual	~FunctionalProductDomain ()
const Space< Scalar > &	getDomain () const
	getDomain() returns a reference to the domain of the underlying function.
virtual const ProductSpace< Scalar > &	getProductDomain () const =0
	ProductDomain() returns a reference to the domain of the underlying function, as an instance of HCL_ProductSpace.
Protected Member Functions
virtual void	applyPartialGradient (int i, const Vector< Scalar > &x, Vector< Scalar > &g) const =0
	applyPartialGradient() computes the component of the gradient corresponding to the ith component of the independent variable.
void	export_applyPartialGradient (FunctionalProductDomain< Scalar > const &f, int i, const Vector< Scalar > &x, Vector< Scalar > &g) const
virtual void	applyGradient (const Vector< Scalar > &x, Vector< Scalar > &g) const
	applyGradient() computes the gradient of the function.
virtual void	applyPartialHessian (int i, int j, const Vector< Scalar > &x, const Vector< Scalar > &dxj, Vector< Scalar > &dxi) const =0
	applyPartialHessian() computes the image of the (i,j) block of the Hessian on dxi, giving dxj.
void	export_applyPartialHessian (FunctionalProductDomain< Scalar > const &f, int i, int j, const Vector< Scalar > &x, const Vector< Scalar > &dxj, Vector< Scalar > &dxi) const
virtual void	applyHessian (const Vector< Scalar > &x, const Vector< Scalar > &yin, Vector< Scalar > &yout) const
	applyHessian() computes the image of the Hessian of the function.
virtual FunctionalProductDomain< Scalar > *	clonePD () const =0
Functional< Scalar > *	clone () const
Friends
class	FunctionalEvaluation< Scalar >
class	FunctionalProductDomainEvaluation< Scalar >

Detailed Description

template<class Scalar>
class RVL::FunctionalProductDomain< Scalar >

A specialization of Functional which also has partial derivatives.

Definition at line 244 of file functional.hh.

Constructor & Destructor Documentation

template<class Scalar>

RVL::FunctionalProductDomain< Scalar >::FunctionalProductDomain ( )

Definition at line 351 of file functional.hh.

template<class Scalar>

RVL::FunctionalProductDomain< Scalar >::FunctionalProductDomain ( const FunctionalProductDomain< Scalar > & )

Definition at line 353 of file functional.hh.

template<class Scalar>

virtual RVL::FunctionalProductDomain< Scalar >::~FunctionalProductDomain ( ) [virtual]

Definition at line 355 of file functional.hh.

Member Function Documentation

template<class Scalar>

virtual void RVL::FunctionalProductDomain< Scalar >::applyPartialGradient	(	int	i,
		const Vector< Scalar > &	x,
		Vector< Scalar > &	g
	)			const `[protected, pure virtual]`

applyPartialGradient() computes the component of the gradient corresponding to the ith component of the independent variable.

This is a virtual function and must be implemented in any derived class.

Referenced by RVL::FunctionalProductDomain< Scalar >::applyGradient(), and RVL::FunctionalProductDomain< Scalar >::export_applyPartialGradient().

template<class Scalar>

void RVL::FunctionalProductDomain< Scalar >::export_applyPartialGradient	(	FunctionalProductDomain< Scalar > const &	f,
		int	i,
		const Vector< Scalar > &	x,
		Vector< Scalar > &	g
	)			const `[protected]`

Definition at line 259 of file functional.hh.

References RVL::FunctionalProductDomain< Scalar >::applyPartialGradient().

template<class Scalar>

virtual void RVL::FunctionalProductDomain< Scalar >::applyGradient	(	const Vector< Scalar > &	x,
		Vector< Scalar > &	g
	)			const `[protected, virtual]`

applyGradient() computes the gradient of the function.

This method is implemented in this (the base) class, using PartialGradient(), and can be (but need not be) re-implemented in a derived class.

Implements RVL::Functional< Scalar >.

Definition at line 275 of file functional.hh.

References RVL::FunctionalProductDomain< Scalar >::applyPartialGradient(), RVL::Components< Scalar >::getSize(), and RVL::Vector< Scalar >::linComb().

template<class Scalar>

virtual void RVL::FunctionalProductDomain< Scalar >::applyPartialHessian	(	int	i,
		int	j,
		const Vector< Scalar > &	x,
		const Vector< Scalar > &	dxj,
		Vector< Scalar > &	dxi
	)			const `[protected, pure virtual]`

applyPartialHessian() computes the image of the (i,j) block of the Hessian on dxi, giving dxj.

Note the implicit use of duality: the action of the (i,j) second partial derivative on (dxj,dxi) is the inner product of dxi with applyPartialHessian(i,j,x,dxj,*).

Note also that the adjoint of the (i,j) block is the (j,i) block.

Referenced by RVL::FunctionalProductDomain< Scalar >::applyHessian(), RVL::FunctionalEvaluation< Scalar >::applyPartialHessian(), and RVL::FunctionalProductDomain< Scalar >::export_applyPartialHessian().

template<class Scalar>

void RVL::FunctionalProductDomain< Scalar >::export_applyPartialHessian	(	FunctionalProductDomain< Scalar > const &	f,
		int	i,
		int	j,
		const Vector< Scalar > &	x,
		const Vector< Scalar > &	dxj,
		Vector< Scalar > &	dxi
	)			const `[protected]`

Definition at line 306 of file functional.hh.

References RVL::FunctionalProductDomain< Scalar >::applyPartialHessian().

template<class Scalar>

virtual void RVL::FunctionalProductDomain< Scalar >::applyHessian	(	const Vector< Scalar > &	x,
		const Vector< Scalar > &	yin,
		Vector< Scalar > &	yout
	)			const `[protected, virtual]`