Loop around a toolbox

Dear community,

I have a question concerning the optimization toolbox.

I have used the optimization toolbox to solve a problem. And that is working fine so far.

Now, I would like to see the solutions for different input variables (let's say for values from 40 to 800) using a loop and store the solutions in a matrix.

The problem is, if I just wrap the editor in a loop (as I would do with usual code), I get an error.

I think MATLAB does not recognize that the "for N = 40:800" at the beginning of the editor and the "end" at the editor's end belong together.

Does somebody know a solution to this problem?

I really appreciate any help you can provide.

[Cris LaPierre]

It would be helpful if you could share your code, as well as the full error message (all the red text).

[Andrew]

Hi, in the code below I have converted the task to editable code.

The error I get is: "Function definition are not supported in this context. Functions can only be created as local or nested functions in code files."

for N = 40:80

v = [1];
m = zeros(1,N); 
for n = 1:N
    m(n)=n;
end

nvar = 5; 
LB = [5 0 -5 -10 -20];
UB = [40 3 5 20 20]; 
q = zeros(1,N); 



% Pass fixed parameters to objfun
objfun3 = @(optimInput)objectiveFcn(optimInput,Daten,m,N,v);

% Set nondefault solver options
options3 = optimoptions("ga","PopulationSize",300,"MaxGenerations",400);

% Solve
[solution,objectiveValue] = ga(objfun3,nvar,[],[],[],[],LB,UB,[],[],options3);

% Clear variables
clearvars objfun3 options3

function f = objectiveFcn(optimInput, Daten, m, N, v)

Function definition are not supported in this context. Functions can only be created as local or nested functions in code files.

a = optimInput(1);
b = optimInput(2);
c = optimInput(3);
d = optimInput(4);
e = optimInput(5);

for n = 1:N
q(n) = (Daten(v,n)-(a*sin(b*m(n)+c) + (d+e*m(n))))^2;
end

f = sum(q);

end

end

[Walter Roberson]

Your function definition is inside your for N loop.

Answer 1

[Matt J]

This runs:

for N = 40:42 %<---shortened

    Daten=rand(N);%<-----added
    
    v = [1];
    m = zeros(1,N); 
    for n = 1:N
        m(n)=n;
    end
    
    nvar = 5; 
    LB = [5 0 -5 -10 -20];
    UB = [40 3 5 20 20]; 
    q = zeros(1,N); 
    
    
    
    % Pass fixed parameters to objfun
    objfun3 = @(optimInput)objectiveFcn(optimInput,Daten,m,N,v);
    
    % Set nondefault solver options
    options3 = optimoptions("ga","PopulationSize",300,"MaxGenerations",400);
    
    % Solve
    [solution,objectiveValue] = ga(objfun3,nvar,[],[],[],[],LB,UB,[],[],options3);
    
    % Clear variables
    clearvars objfun3 options3

end

ga stopped because the average change in the fitness value is less than options.FunctionTolerance.
ga stopped because the average change in the fitness value is less than options.FunctionTolerance.
ga stopped because it exceeded options.MaxGenerations.

function f = objectiveFcn(optimInput, Daten, m, N, v)
    a = optimInput(1);
    b = optimInput(2);
    c = optimInput(3);
    d = optimInput(4);
    e = optimInput(5);
    
    for n = 1:N
    q(n) = (Daten(v,n)-(a*sin(b*m(n)+c) + (d+e*m(n))))^2;
    end
    
    f = sum(q);

end