Why does my function abc give error with optimtool

I have a function file abc. I want to run it with optimtool ga. I do settings in the Optimtool as below:

Solver: ga-Genetic Algorithm

Fitness function: @abc

Number of variable 6

Bounds Lower: [0 0 0 0 0 0] Upper:[180 180 180 180 180 180]

-----------------------------

After running it, it gives the following error:

Optimization running.

Error running optimization.

Incorrect dimensions for matrix multiplication. Check that the number of columns in the first matrix matches the number of rows in the second matrix. To perform elementwise multiplication, use '.*'.

Answer 1

[Torsten]

The product r*k in the line

A = exp(-1j*r*k);% Steering matrix

is not defined.

k is (3x3), r is (Nx3) where N = 10.

Setting N = 3 would make the product to be defined.

[Torsten]

See above.

b must be of the same size as [az,el].

[Torsten]

And if you knew this, why did you ask what caused the error in your call of abc ?

[Torsten]

I asked that when the code is correct, then why does it give error with the optimtool?

Because the optimtool works with a (6x1) vector (as I see from the definition of your lower and upper bounds) whereas you need it reshaped as a (3x2) matrix for abc.

[Torsten]

So what do you suggest?

I already wrote what you have to do.

Before calling abc, you have to reshape the vector b supplied by ga as

b = reshape(b,3,2)

[Sadiq Akbar]

Thank you very much dear Torsten for your help. But now there is one issue. When I keep the bounds as:

Lower=[0;0;0;0;0;0] Upper=[360;360;360;180;180;180]

It gives wrong results. But when I change them to:

Lower=[0;0;0;0;0;0] Upper=[180;180;180;180;180;180]

then it gives good results. Although the actual range is the 1st case. So what do I do now?

[Torsten]

I have no experience with ga.

If your problem is such that you can use "fmincon" or some other deterministic optimizer instead of ga, you should do so.