LBFGSpp::LBFGSBSolver< Scalar, LineSearch > Class Template Reference

#include <LBFGSB.h>

Public Member Functions
	LBFGSBSolver (const LBFGSBParam< Scalar > &param)
template<typename Foo >
int	minimize (Foo &f, Vector &x, Scalar &fx, const Vector &lb, const Vector &ub)
const Vector &	final_grad () const
Scalar	final_grad_norm () const

Detailed Description

template<typename Scalar, template< class > class LineSearch = LineSearchMoreThuente>
class LBFGSpp::LBFGSBSolver< Scalar, LineSearch >

L-BFGS-B solver for box-constrained numerical optimization

Definition at line 23 of file LBFGSB.h.

Constructor & Destructor Documentation

LBFGSBSolver()

template<typename Scalar , template< class > class LineSearch = LineSearchMoreThuente>

LBFGSpp::LBFGSBSolver< Scalar, LineSearch >::LBFGSBSolver ( const LBFGSBParam< Scalar > &  param )

inline

Constructor for the L-BFGS-B solver.

Parameters

param

An object of LBFGSParam to store parameters for the algorithm

Definition at line 95 of file LBFGSB.h.

Member Function Documentation

minimize()

template<typename Scalar , template< class > class LineSearch = LineSearchMoreThuente>

template<typename Foo >

int LBFGSpp::LBFGSBSolver< Scalar, LineSearch >::minimize ( Foo &  f, Vector &  x, Scalar &  fx, const Vector &  lb, const Vector &  ub  )

inline

Minimizing a multivariate function subject to box constraints, using the L-BFGS-B algorithm. Exceptions will be thrown if error occurs.

Parameters

f	A function object such that f(x, grad) returns the objective function value at x, and overwrites grad with the gradient.
x	In: An initial guess of the optimal point. Out: The best point found.
fx	Out: The objective function value at x.
lb	Lower bounds for x.
ub	Upper bounds for x.

Returns

Number of iterations used.

Definition at line 117 of file LBFGSB.h.

final_grad()

template<typename Scalar , template< class > class LineSearch = LineSearchMoreThuente>

const Vector & LBFGSpp::LBFGSBSolver< Scalar, LineSearch >::final_grad ( ) const

inline

Returning the gradient vector on the last iterate. Typically used to debug and test convergence. Should only be called after the minimize() function.

Returns

A const reference to the gradient vector.

Definition at line 271 of file LBFGSB.h.

final_grad_norm()

template<typename Scalar , template< class > class LineSearch = LineSearchMoreThuente>

Scalar LBFGSpp::LBFGSBSolver< Scalar, LineSearch >::final_grad_norm ( ) const

inline

Returning the infinity norm of the final projected gradient. The projected gradient is defined as \(P(x-g,l,u)-x\), where \(P(v,l,u)\) stands for the projection of a vector \(v\) onto the box specified by the lower bound vector \(l\) and upper bound vector \(u\).

Definition at line 279 of file LBFGSB.h.

---

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

• LBFGSB.h