I am trying to minimize the function ARE in file abc.m. But I m getting the error "your objective function must return a scalar value", I know that I minimizing this function with a variable which is not constant. It works when I take any single val

========================================= mytest.m file ========================================= close all; clear all; clc;

a0 = [.0001 .0001 .0001 .0001 .0001 ]

f =@(a)abc(a);

[a fval] = simulannealbnd(f,a0);

disp(a); ===============================================

second file =============================================== abc.m ===============================================

function ARE = abc(a) ;

w = linspace(0,pi,1000)

ARE = ((1/(j*w)) - ((a(1)*exp(2*j*w) + a(2)*exp(j*w) + a(3)) / (exp(2*j*w)+a(4)*exp(j*w) + a(5)))/(1/(j*w)));

Answer 1

[Torsten]

What are you trying to do ?

Find different a(1),...,a(5) that minimize ARE for each single value of w ?

Best wishes

Torsten.

[Nitin Rawal]

It minimizes with each single value of w But I want the function to be minimized with range of values at a time by function simulannealbnd.

That is objective function should be non scalar.

[Torsten]

What exactly do you mean by

"I want the function to be minimized with range of values at a time"

?

If you want to minimize f for a range of values for w, use a for-loop over the elements of the w-array and therein, call "simulannealbnd" for each element of w separately.

Best wishes

Torsten.

[Nitin Rawal]

I already tried what are you saying. It comes up with an error message:

??? Error using ==> samakedata at 28 Your objective function must return a scalar value.

Error in ==> simulannealcommon at 113 solverData = samakedata(solverData,problem,options);

Error in ==> simulanneal at 44 [x,fval,exitflag,output,solverData,problem,options] = ...

Error in ==> simulannealbnd at 129 [x, fval, exitflag, output] = simulanneal(FUN, x0, [], [], [], [], lb, ub, options);

Error in ==> simul at 6 [x,fval]= simulannealbnd(fh,x0);

This is because simulannealbnd's objective function can work properly with constant parameters But here I m using w =linspace(0,pi,10) stat# which evaluates ARE for each value of w but simulannealbnd's objective function can return only scalar value not a vector.

So please help me that how should I code objective function that my objective function minimized: Following is code of objective function

function ARE = abc(a) ; w = linspace(0,pi,1000) ARE =abs(((1./(j*w))-((a(1)*exp(2*j*w)+a(2)*exp(j*w)+a(3))./(exp(2*j*w)+a(4)*exp(j*w)+a(5))))./(1/(j*w)));

If it works like lsqcurvefit non linear function of matlab it is done with my work.

So please also have a look in lsqcurvefit function.

Thanks

Nitin Rawal

Answer 2

[Torsten]

a0 = [.0001 .0001 .0001 .0001 .0001 ];
w = linspace(0,pi,1000);
for i=1:length(w)
w_scal=w(i);
f =@(a)abc(a,w_scal);
[a_vec fval] = simulannealbnd(f,a0);
a_mat(:,i)=a_vec(:);
end

function ARE = abc(a,w)

ARE = ((1/(j*w)) - ((a(1)*exp(2*j*w) + a(2)*exp(j*w) + a(3)) / (exp(2*j*w)+a(4)*exp(j*w) + a(5)))/(1/(j*w)));

I wonder what j is - the imaginary unit ?

Best wishes

Torsten.