function [ret, popt, info, covar]=levmar(fname, jacname, p0, x, itmax, opts, type) % LEVMAR matlab MEX interface to the levmar non-linear least squares minimization % library available from http://www.ics.forth.gr/~lourakis/levmar/ % % levmar can be used in any of the 4 following ways: % [ret, popt, info, covar]=levmar(fname, jacname, p0, x, itmax, opts, 'unc', ...) % [ret, popt, info, covar]=levmar(fname, jacname, p0, x, itmax, opts, 'bc', lb, ub, ...) % [ret, popt, info, covar]=levmar(fname, jacname, p0, x, itmax, opts, 'lec', A, b, ...) % [ret, popt, info, covar]=levmar(fname, jacname, p0, x, itmax, opts, 'blec', lb, ub, A, b, wghts, ...) % % The dots at the end denote any additional, problem specific data that are passed uniterpreted to % all invocations of fname and jacname, see below for details. % % In the following, the word "vector" is meant to imply either a row or a column vector. % % required input arguments: % - fname: String defining the name of a matlab function implementing the function to be minimized. % fname will be called as fname(p, ...), where p denotes the parameter vector and the dots any % additional data passed as extra arguments during the invocation of levmar (refer to Meyer's % problem in lmdemo.m for an example). % % - p0: vector of doubles holding the initial parameters estimates. % % - x: vector of doubles holding the measurements vector. % % - itmax: maximum number of iterations. % % - opts: vector of doubles specifying the minimization parameters, as follows: % opts(1) scale factor for the initial damping factor % opts(2) stopping threshold for ||J^T e||_inf % opts(3) stopping threshold for ||Dp||_2 % opts(4) stopping threshold for ||e||_2 % opts(5) step used in finite difference approximation to the Jacobian. % If an empty vector (i.e. []) is specified, defaults are used. % % optional input arguments: % - jacname: String defining the name of matlab function implementing the Jacobian of function fname. % jacname will be called as jacname(p, ...) where p is again the parameter vector and the dots % denote any additional data passed as extra arguments to the invocation of levmar. If omitted, % the Jacobian is approximated with finite differences through repeated invocations of fname. % % - type: String defining the minimization type. It should be one of the following: % 'unc' specifies unconstrained minimization. % 'bc' specifies minimization subject to box constraints. % 'lec' specifies minimization subject to linear equation constraints. % 'blec' specifies minimization subject to box and linear equation constraints. % If omitted, a default of 'unc' is assumed. Depending on the minimization type, the MEX % interface will invoke one of dlevmar_XXX, dlevmar_bc_XXX, dlevmar_lec_XXX or dlevmar_blec_XXX % % - lb, ub: vectors of doubles specifying lower and upper bounds for p, respectively % % - A, b: k x m matrix and k vector specifying linear equation constraints for p, i.e. A*p=b % A should have full rank. % % - wghts: vector of doubles specifying the weights for the penalty terms corresponding to % the box constraints, see lmblec_core.c for more details. If omitted and a 'blec' type % minimization is to be carried out, default weights are used. % % % output arguments % - ret: return value of levmar, corresponding to the number of iterations if successful, -1 otherwise. % % - popt: estimated minimizer, i.e. minimized parameters vector. % % - info: optional array of doubles, which upon return provides information regarding the minimization. % See lm_core.c for more details. % % - covar: optional covariance matrix corresponding to the estimated minimizer. % error('levmar.m is used only for providing documentation to levmar; make sure that levmar.c has been compiled using mex');