roboptim::finiteDifferenceGradientPolicies::Simple< T > Class Template Reference

Fast finite difference gradient computation. More...

#include <roboptim/core/finite-difference-gradient.hh>

List of all members.

Public Member Functions
	ROBOPTIM_DIFFERENTIABLE_FUNCTION_FWD_TYPEDEFS_ (GenericDifferentiableFunction< T >)
void	computeGradient (const GenericFunction< T > &adaptee, value_type epsilon, gradient_t &gradient, const argument_t &argument, size_type idFunction, argument_t &xEps) const throw ()
template<>
void	computeGradient (const GenericFunction< EigenMatrixSparse > &adaptee, value_type epsilon, gradient_t &gradient, const argument_t &argument, size_type idFunction, argument_t &xEps) const throw()

---

Detailed Description

template<typename T>
class roboptim::finiteDifferenceGradientPolicies::Simple< T >

Fast finite difference gradient computation.

Finite difference is computed using forward difference.

Examples:

finite-difference-gradient.cc.

---

Member Function Documentation

template<typename T >

void roboptim::finiteDifferenceGradientPolicies::Simple< T >::computeGradient	(	const GenericFunction< T > &	adaptee,
		value_type	epsilon,
		gradient_t &	gradient,
		const argument_t &	argument,
		size_type	idFunction,
		argument_t &	xEps
	)		const throw ()

References roboptim::GenericFunction< T >::inputSize(), and roboptim::GenericFunction< T >::outputSize().

template<>

void roboptim::finiteDifferenceGradientPolicies::Simple< EigenMatrixSparse >::computeGradient	(	const GenericFunction< EigenMatrixSparse > &	adaptee,
		value_type	epsilon,
		gradient_t &	gradient,
		const argument_t &	argument,
		size_type	idFunction,
		argument_t &	xEps
	)		const throw() [inline]

References roboptim::GenericFunction< T >::inputSize(), and roboptim::GenericFunction< T >::outputSize().

template<typename T >

roboptim::finiteDifferenceGradientPolicies::Simple< T >::ROBOPTIM_DIFFERENTIABLE_FUNCTION_FWD_TYPEDEFS_

(

GenericDifferentiableFunction< T >

)