brute force procedure for a function

Dear collegues,

I have a question about design of flexiable brute force procedure over functions parameters. suppose I have a function

kpi = func(a,b,c)

in brute force function i want to select only for which parameters optimize and in which range e.g.: Optimization over a = 1:5 and c= 100:125 and b = 2

result = f_opt(@func,[1 0 1], {1:5,2,100:125})

the number of parameters in func must be canchagable. func is not vectorized.

What is the best way to design such function?

Answer 1

[Yu Jiang]

If you would like to fix one parameter, e.g., b and optimize the function by finding appropriate a and c in their given range, you may try the following

x = fmincon(@(x) func(x(1),b,x(2)), x0, [], [], [], [], [1 100], [5 125])

where x0 is the initial condition.