| |
| | [No title] (Site not responding. Last check: 2007-10-08) |
 | | function [x, xiter]= steepest(x0,tol,maxiter,act,func,grad,varargin) % % Gradient search (steepest descent) algorithm for min/max f(x): % Golden section is used at each iteration to find step-length. |
 | | end function h = f1d(t,f,x,p,varargin) %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % Evaluate the univariant function % h(t) = f(x + t*p) % for given multi-variant function f % and given vectors x and p. |
 | | h = feval(f, x+t*p, varargin{:}); function [x,f] = golden(xlo, xhi, tol, act, func, varargin) %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%% % % This function implements golden section to search % for an optimum of func(x) in the interval [xlo, xhi]. |
| www.caam.rice.edu /~yzhang/caam378/nlp/steep.m (90 words) |
|