MINIMIZE Function

Standard Library function that minimizes a real valued function of n real variables.

Usage

x = MINIMIZE(f, l, u, g, i, y )

Input Parameters

f—A string specifying a user supplied function to be minimized. Input is a (m,n) array of m points in n-space (m variable); output is a (m,p+1) array b, where p is the number of constraints, b(*,0) contains the objective function values at each of the m input points, and b(*,j) contains the corresponding values of the jth constraint. All constraints must be of the form c(x) ≤ 0.

l—An n-element vector of lower bounds for the independent variables

u—An n-element vector of upper bounds for the independent variables

g—An n-element vector giving an initial guess for the solution

i—An integer limit on the number of iterations

Returned Value

x—The n-element solution vector

y—(optional) A (p+1)-element vector containing the objective function value at x followed by the constraint values at x.

Keywords

d—A string specifying a user-supplied gradient function. Input is the n-element vector at which to calculate the gradient(s). Output is a (n,p+1) array that contains the objective function gradient followed by the constraint gradients.

s—An (n,2) array where s(*,0) is the maximum allowable step and s(*,1) is the minimum allowable step. The default is [ [(u-l)/100], [(u-l)/1000] ].

Examples

See wave/lib/user/examples/minimize_ex*.pro.