Minpack functions

Wrappers or bindings for Minpack functions are provided. More...

Classes
struct	mimas::lmdif_t
	Function-object for lmdif. More...
struct	mimas::lmdif1_t
	Function-object for lmdif1. More...
Functions
lmdif_t::Vector	mimas::lmdif (const lmdif_t::Vector &A, void(*fnc)(int *m, int *n, double *x, double *fvec, int *iflag), int nb, int maxfev, double tolerance=1e-7, int mode=1, int factor=1, double epsfcn=0)
	Levenberg-Marquardt algorithm.
lmdif1_t::Vector	mimas::lmdif1 (const lmdif1_t::Vector &A, void(*fnc)(int *m, int *n, double *x, double *fvec, int *iflag), int nb, double tolerance=1e-7)
	Levenberg-Marquardt algorithm.

---

Detailed Description

Wrappers or bindings for Minpack functions are provided.

The vector-classes of boost are used as datatypes.

Read on how to install MINPACK, if it is not included in your distribution.

Author:

Julien Faucher (faucherj@gmail.com)

---

Function Documentation

lmdif_t::Vector mimas::lmdif	(	const lmdif_t::Vector &	A,
		void(*)(int *m, int *n, double *x, double *fvec, int *iflag)	fnc,
		int	nb,
		int	maxfev,
		double	tolerance = 1e-7,
		int	mode = 1,
		int	factor = 1,
		double	epsfcn = 0
	)			[inline]

Levenberg-Marquardt algorithm.

(Full version)

Parameters:

	A	a vector containing a first approximation
	fnc	is the user-supplied subroutine which calculates the values of the functions in a special point
	nb	the number of functions
	maxfev	the number of iterations to do
	tolerance	termination occurs when both the actual and predicted relative reductions in the sum of squares are at most tol.
	mode	is an integer input variable. If mode = 1, the variables will be scaled internally. If mode = 2, the scaling is specified by the input diag.
	factor	is a positive input variable used in determining the initial step bound.
	epsfcn	is an input variable used in determining a suitable step length for the forward-difference approximation.

See also:

lmdif_ for more information on the parameters (especially on fnc)

Definition at line 61 of file nlinalg.h.

lmdif1_t::Vector mimas::lmdif1	(	const lmdif1_t::Vector &	A,
		void(*)(int *m, int *n, double *x, double *fvec, int *iflag)	fnc,
		int	nb,
		double	tolerance = 1e-7
	)			[inline]

Levenberg-Marquardt algorithm.

(Simplified version)

Parameters:

	A	a vector containing a first approximation
	fnc	is the user-supplied subroutine which calculates the values of the functions in a special point
	nb	the number of functions
	tolerance	termination occurs when both the actual and predicted relative reductions in the sum of squares are at most tol.

See also:

lmdif_ for more information on the parameters (especially on fnc)

Definition at line 102 of file nlinalg.h.

---