42 MINIMIZATION OF FUNCTIONS

The minimization of general functions of one or more variables can be carried out using the command:

MINIMIZE, func, x$_1$[, x$_2$, x$_3$, ...]

where func represents a function of up to 50 variables x$_1$, x$_2$, .... Two different optimization methods can be selected as described below which do or do not use numerical derivative information.

The optimization method, as well as finer control over func, can be chosen using the METHOD directive

METHOD, key [, key1=value, key2=value, ...]

where key defines the optimization method. Valid options for key are:

BFGS

Broyden-Fletcher-Goldfarb-Shanno conjugate gradient method, which uses numerical gradients (default)

SIMPLEX

Downhill simplex method, which uses only function evaluations

Options to these methods, key1, key2, ..., are:

VARSCALE=vscale

Optimization in space of scaled variables.
vscale=0 no scaling (not recommended)
vscale=1 optimization in the space of ln($x$)
vscale=2 optimization in space of initial value scaling, e.g., $x_1/x_{1i}$ (default)

THRESH=thresh

Required accuracy of either the optimized gradient (BFGS) or function value (SIMPLEX). The default is $1 \cdot 10^{-4}$.

VSTEP=epsd

Step size for numerical gradients (BFGS) or initial SIMPLEX vertices

PROC=procname

Specifies the procedure to be executed in each optimization step. This defines a complete function evaluation (if needed, numerical gradients will be evaluated using this procedure as well)

STARTCMD=command

Specifies a start command. In each optimization step all input beginning with command to the current MINIMIZE is processed.

Miscellaneous directives

MAXIT=maxit

maximum number of optimization cycles. The default is 30 for BFGS and 100 for SIMPLEX.