How can I use my own function to optimize using fminunc?

Hi Can anybody help me solving my problem? I have a function in Matlab with two variables x1 and x2 and I wanna optimize my function and find the optimal value of x1 and x2 how can I use fminunc? this is may function :

function f=optim(x)
%%%%%x1 hlow for com 1, x2 for interval
%tau=5;
%H2=0.00125;
H1=0.000125;
CI=5800;
Cp=30000;       
Cr=20000;
for j=2:10 % number of intervals 
    t1(j)=j*x2;
    t2(j)=(j-1)*x2;
     Inspcost1(j)=MyF(x1,t1(j));
     Inspcost2(j)=MyF(x1,t2(j));
     Inspcostall(j)=j.*abs((Inspcost2(j)-Inspcost1(j)));
end
Inspcost=CI*sum(Inspcostall(j))
f=Inspecost;

(MyF is another function) I don't know how to use matlab to optimize this?

Answer 1

[Star Strider]

In the optimization routines, the parameters you want to optimise have to be in a single vector. In this instance, simply change them to subscripted veriables, ‘x(1)’ and ‘x(2)’:

t1(j)=j*x(2);
t2(j)=(j-1)*x(2);
Inspcost1(j)=MyF(x(1),t1(j));
Inspcost2(j)=MyF(x(1),t2(j));

[NooshinY]

I still cannot run the fminunc

fh=optim(x)

    x0=[1.1;3];    % Starting guess     
    options = optimoptions('fminunc','Algorithm','quasi-newton');
[x, fval, exitflag, output] = fminunc(fh,x0,options);

I run this code and Matlab give me errors

[Walter Roberson]

 fh=optim(x)

should be

fh = @optim;

[Star Strider]

@Walter — Thank you.