runoptprob: General optimization test runner

In NISTO: Nonlinear least squares and function minimization examples from NIST

Description

A tool to run various pre-defined problems with many different optimization or nonlinear least squares methods.

Usage

runoptprob <- function(pfilename, probclass=NULL, minmeth=NULL, submeth=NULL, nstart=0, 
                 runopts=list(), control=list(), ...) 

Arguments

pfilename	a character string giving the filename of a particular problem located in the inst/extdata/ directory of the package.
probclass	If NULL (default) then attempt all available classes, else one of "uncopt", "boundopt", "formula", or "sumsquares".
minmeth	One of a selected set of minimization tool classes. The list includes??.
submeth	A particular member of the class of methods in minmeth.
nstart	The index of the pre-defined starting set of parameters in the specified problem file, or 0 for all predefined starting vectors, ?? or (?? could have user defined if negative??)
runopts	Any arguments for running a particular minimization method needed.
control	Any controls that should be supplied to the minimization method.
...	Any other (named) arguments

Details

To be filled in!??

Value

A list with components:

par

The best set of parameters found.

pfilename

The name of the original problem file (root thereof??).

nstart

The index of the starting parameter set.

minmeth

The class of the minimization method.

submeth

The particular member of the class of minimizers.

NISTO documentation built on Nov. 24, 2016, 1:19 a.m.