ga-fitness

Dear All,

I need to optimize parameters of a function using the ga module. I have a coefficient matrix say A(9:4):

A =

   -0.0996   -0.0323   -0.0640   -0.0365
    0.0859   -0.0636    0.0552    0.0315
    0.1805    0.0125    0.1159    0.0662
    0.0521    0.0513    0.0335    0.0191
   -0.3028    0.0647   -0.1946   -0.1111
   -0.2242   -0.0382   -0.1440   -0.0822
    0.0475   -0.0191    0.0305    0.0174
    0.2170   -0.0010    0.1394    0.0796
    0.0437    0.0257    0.0281    0.0160
and a vector B where i store the experimental data say B(9:1):

B =

   -4.8700
   -3.2100
    5.1000
    5.3400
   -1.6200
   -8.1700
   -0.4700
    4.8300
    3.0800
my chromosome will be a vector x(4:1)

I would like to write a fitness function like:

fitnessfcn = abs((A*x)-B) [x,fval]=ga(fitnessfcn,4)

I have used the below function

FitnessFcn = @(x) abs( ((A(1,1)*x(:,1))+(A(1,2)*x(:,2))+(A(1,3)*x(:,3))+(A(1,4)*x(:,4)))-B(1,1))

which only uses (the first row of A) * (column x)-(first element of B)

How can I write in general form to use all the data. I have tried fitnessfcn = abs((A*x)-B) but it says Error using ==> mtimes Inner matrix dimensions must agree.

Do I have to use for loops and how to write a function with for loops?

I would really welcome any sugestion.

Kind regards, Christos

Answer 1

[Andrei Bobrov]

variant

FitnessFcn = @(x)bsxfun(@minus,abs(bsxfun(@times,A,x)),B)

[chris]

Many thanks for your answer. It is very helpful. I am using this in the example above with:

[x]=ga(FitnessFcn,4)

and i am getting an error message:

Subscripted assignment dimension mismatch.

Is it easy to find the problem?

Once again thanks

Kind regards, christos

Answer 2

[Star Strider]

If the fitness function is supposed to produce a scalar to be minimized, then I suggest something like:

FitnessFcn = @(A,B,x) ( sum( (A*x-B).^2 ) );

Since A is (9x4), x is (4x1) and B is (9x1), A*x will produce a (9x1) vector. A*x-B produces the error vector, and the FitnessFcn here will produce the sum of the squared errors, a scalar >= 0.

Answer 3

[chris]

Many thanks for your answer. It is very helpful. I am using this in the example above with:

[x]=ga(FitnessFcn,4)

and i am getting an error message:

Subscripted assignment dimension mismatch.

Is it easy to find the problem?

Once again thanks

Kind regards, christos

Answer 4

[chris]

Sorry for your time. I am coming back to this old question since after postponing for one year I start again to try make it work. However I haven't solved the problem above. I would be more than happy for any help since my experience with matlab is rather limited.

All the best, chris