CostFunction Class Reference

Cost function abstract class for optimization problem. More...

#include <ql/math/optimization/costfunction.hpp>

Inheritance diagram for CostFunction:

Public Member Functions
virtual Real	value (const Array &x) const
	method to overload to compute the cost function value in x
virtual Array	values (const Array &x) const =0
	method to overload to compute the cost function values in x
virtual void	gradient (Array &grad, const Array &x) const
	method to overload to compute grad_f, the first derivative of
virtual Real	valueAndGradient (Array &grad, const Array &x) const
	method to overload to compute grad_f, the first derivative of
virtual void	jacobian (Matrix &jac, const Array &x) const
	method to overload to compute J_f, the jacobian of
virtual Array	valuesAndJacobian (Matrix &jac, const Array &x) const
	method to overload to compute J_f, the jacobian of
virtual Real	finiteDifferenceEpsilon () const
	Default epsilon for finite difference method :

Detailed Description

Cost function abstract class for optimization problem.

Examples

GlobalOptimizer.cpp.

Member Function Documentation

value()

virtual Real value ( const Array & x ) const

virtual

method to overload to compute the cost function value in x

Reimplemented in LeastSquareFunction, PenaltyFunction< Curve >, and ProjectedCostFunction.

values()

virtual Array values ( const Array & x ) const

pure virtual

method to overload to compute the cost function values in x

Implemented in LeastSquareFunction, PenaltyFunction< Curve >, and ProjectedCostFunction.

gradient()

virtual void gradient ( Array & grad, const Array & x ) const

virtual

method to overload to compute grad_f, the first derivative of

Reimplemented in LeastSquareFunction.

valueAndGradient()

virtual Real valueAndGradient ( Array & grad, const Array & x ) const

virtual

method to overload to compute grad_f, the first derivative of

Reimplemented in LeastSquareFunction.