Use of the optimisation toolbox,

I have the code to display 2 functions in a graphical form I need to find the minimum point with several constraints, my code to display the equations is

x = 0.1:.01:2.1; 
y = 0:.05:10;
[X,Y] = meshgrid(x,y);
Z = meshgrid(0.125:.025:5.125);
C = (1.10471.*((Z).^2) .* ((X).^2))+(0.0481.*(Y.*Z)*14.*X);
surfc(X,Y,Z,C);
axis([0 2.1 0 10 0 5.125]);
hold on;

R = 6000./sqrt(2.*Z.*X);
A = 6000*(14+0.5*X)*sqrt(0.25*(X).^2 +(Z+Y).^2);
B= 2*0.707*Z.*X.*(((X).^2)/(12+0.25*(Z+Y)^2));
S =A./B;
T=sqrt((((R).^2)+((S).^2)+X.*R.*S)./sqrt(0.25*((X).^2)+(Z+Y).^2));
surfc(X,Y,Z,T);

I'm trying to now use the function fminunc to minimalise the term for C whereby T is less than 13600 so far I thought it best to do firstly minimise the innital equation before inputing the futher constraints so my code is as follows,

function f = myfun(x)
f = (1.10471*(x(1)^2) * (x(1)^2))+(0.0481*(x(2)*x(3))*14*x(1));
x0 = [1,1];
[x,fval] = fminunc(@myfun,x0);

however when run this returns the statement

??? Input argument "x" is undefined.

Error in ==> myfun at 2
f = (1.10471*(x(1)^2) * (x(1)^2))+(0.0481*(x(2)*x(3))*14*x(1));

>>

Any clues as to where I'm going wrong?

Many thanks

Answer 1

[Walter Roberson]

Split your code like

function [x, fval] = minimize_myfun
  x0 = [1,1];
  [x,fval] = fminunc(@myfun,x0);

function f = myfun(x)
  f = (1.10471*(x(1)^2) * (x(1)^2))+(0.0481*(x(2)*x(3))*14*x(1));

It is important that you do not have the call to fminunc within the myfun function.

Answer 2

[james]

ok that makes sense thanks for that, I'm now getting an error whereby it thinks there are too many input arguments? is that because my code would call upon x(3)? Also would I just put the constraints into the bottom of the code?

Sorry for my lack of coding knowledge

[Walter Roberson]

Your x0 does need to be as long as the maximum subscript of x that you reference.