I try to write simple analytical function Gomez and Levy with ga in optimization toolbox but it cannot run

%fitness function Gomez and Levy
 
function y=fitnessfunctiongl(x)
 
x1=x(1);
x2=x(2);
 
y=(4*x1.^2)-2.1*x1.^4+1/3*x1.^6+(x1*x2)-4*x2.^2+4*x2.^4;
 
end

¬¬¬¬¬%constraints
 
function [c ,ceq]=constraintgl(x)
x1=x(1);
x2=x(2);
c=-sin(4*pi*x1)+2*sin(2*pi*x2)^2-1.5;
ceq=[];
end

%main script to pass function and constraints to ga for optimization 
ObjFcn=@fitnessfunctiongl;
nvars=2;
lb=[-1, -1];
ub=[0.75, 1];
 
nonlcon=@constraintgl;
options = optimoptions('ga','ConstraintTolerance',1e-6,'Display','iter','PlotFcn',{@gaplotrange,@gaplotbestf,@gaplotselection,@gaplotmaxconstr}); %plotting the process of finding the solution 
%solution and function value
[x , fval,exitFlag,output,population,scores]= ga(ObjFcn,nvars,[],[],[],[],lb,ub,nonlcon,options);




%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%errors
Error in gadsplot (line 141)
        [state,optimvalues] = callOnePlotFcn(fname,plotNames{i},state,options.OutputPlotFcnOptions,optimvalues,'init',args{i}{:});

Error in gacon (line 55)
state = gadsplot(options,state,'init','Genetic Algorithm');

Error in ga (line 406)
                [x,fval,exitFlag,output,population,scores] = gacon(FitnessFcn,nvars, ...

Error in maingl (line 11)
[x , fval,exitFlag,output,population,scores]= ga(ObjFcn,nvars,[],[],[],[],lb,ub,nonlcon,options);
 

[Walter Roberson]

What error did you encounter?

[sogol bandekian]

Error using makeState (line 61)

Your fitness function must return a scalar value.

Error in gacon (line 44)

state = makeState(GenomeLength,SubFitness,Iterate,subtype,options);

Error in ga (line 406)

[x,fval,exitFlag,output,population,scores] = gacon(FitnessFcn,nvars, ...

Error in maingl (line 11)

[x , fval,exitFlag,output,population,scores]= ga(ObjFcn,nvars,[],[],[],[],lb,ub,nonlcon,options);

Answer 1

[Walter Roberson]

y=(4*x1.^2)-2.1*x1.^4+1/3*x1.^6+(x1*x2)-4*x.^2+4*x2.^4;

Notice that you have 4*x.^2 . But x is a vector so that would return a vector.

[sogol bandekian]

I did it and change the x to x2 but I still have errors

Error in gadsplot (line 141)

[state,optimvalues] = callOnePlotFcn(fname,plotNames{i},state,options.OutputPlotFcnOptions,optimvalues,'init',args{i}{:});

Error in gacon (line 55)

state = gadsplot(options,state,'init','Genetic Algorithm');

Error in ga (line 406)

[x,fval,exitFlag,output,population,scores] = gacon(FitnessFcn,nvars, ...

Error in maingl (line 11)

[x , fval,exitFlag,output,population,scores]= ga(ObjFcn,nvars,[],[],[],[],lb,ub,nonlcon,options);

[Walter Roberson]

I do not get that error.

Note: in the below, setting the PlotFcn to [] is done for the purpose of being able to run the code here in MATLAB Answers, as the plot functions need a call that is not supported for the MATLAB Answers facility. For your own use, comment out the line that removes the plot fcn.

%main script to pass function and constraints to ga for optimization 
ObjFcn=@fitnessfunctiongl;
nvars=2;
lb=[-1, -1];
ub=[0.75, 1];
 
nonlcon=@constraintgl;
options = optimoptions('ga','ConstraintTolerance',1e-6,'Display','iter','PlotFcn',{@gaplotrange,@gaplotbestf,@gaplotselection,@gaplotmaxconstr}); %plotting the process of finding the solution 
options.PlotFcn = [];
%solution and function value
[x , fval,exitFlag,output,population,scores]= ga(ObjFcn,nvars,[],[],[],[],lb,ub,nonlcon,options);

Single objective optimization:
2 Variable(s)
1 Nonlinear inequality constraint(s)

Options:
CreationFcn:       @gacreationuniform
CrossoverFcn:      @crossoverscattered
SelectionFcn:      @selectionstochunif
MutationFcn:       @mutationadaptfeasible

                              Best       Max        Stall
Generation  Func-count        f(x)     Constraint  Generations
    1           2500      -1.03163            0      0
    2           4950      -1.03163            0      0
    3           7400      -1.03163            0      0
    4           9850      -1.03163            0      0
    5          12300      -1.03163            0      1
Optimization terminated: average change in the fitness value less than options.FunctionTolerance
 and constraint violation is less than options.ConstraintTolerance.

%fitness function Gomez and Levy
 
function y=fitnessfunctiongl(x)
 
x1=x(1);
x2=x(2);
 
y=(4*x1.^2)-2.1*x1.^4+1/3*x1.^6+(x1*x2)-4*x2.^2+4*x2.^4;
 
end

 
function [c ,ceq]=constraintgl(x)
%constraints
x1=x(1);
x2=x(2);
c=-sin(4*pi*x1)+2*sin(2*pi*x2)^2-1.5;
ceq=[];
end

[sogol bandekian]

I didnot get this point:"For your own use, comment out the line that removes the plot fcn"

what do you mean by this?

if I put [] for plotfcn how can I get the prot for the process of code?

[Walter Roberson]

When I remove the line that assigns [] to plot function, and try to show you the results here, then see:

%main script to pass function and constraints to ga for optimization 
ObjFcn=@fitnessfunctiongl;
nvars=2;
lb=[-1, -1];
ub=[0.75, 1];
 
nonlcon=@constraintgl;
options = optimoptions('ga','ConstraintTolerance',1e-6,'Display','iter','PlotFcn',{@gaplotrange,@gaplotbestf,@gaplotselection,@gaplotmaxconstr}); %plotting the process of finding the solution 
%solution and function value
[x , fval,exitFlag,output,population,scores]= ga(ObjFcn,nvars,[],[],[],[],lb,ub,nonlcon,options);

Single objective optimization:
2 Variable(s)
1 Nonlinear inequality constraint(s)

Options:
CreationFcn:       @gacreationuniform
CrossoverFcn:      @crossoverscattered
SelectionFcn:      @selectionstochunif
MutationFcn:       @mutationadaptfeasible

Error using uicontrol
This functionality is not available on remote platforms.

Error in gadsplot (line 98)
pObj = uicontrol('string','Pause', ...

Error in gacon (line 64)
state = gadsplot(options,state,'init','Genetic Algorithm');

Error in ga (line 418)
[x,fval,exitFlag,output,population,scores] = gacon(FitnessFcn,nvars, ...

%fitness function Gomez and Levy
 
function y=fitnessfunctiongl(x)
 
x1=x(1);
x2=x(2);
 
y=(4*x1.^2)-2.1*x1.^4+1/3*x1.^6+(x1*x2)-4*x2.^2+4*x2.^4;
 
end

 
function [c ,ceq]=constraintgl(x)
%constraints
x1=x(1);
x2=x(2);
c=-sin(4*pi*x1)+2*sin(2*pi*x2)^2-1.5;
ceq=[];
end

When you run this version of the code on your own desktop, or on MATLAB Online, you will instead get a plot -- but the MATLAB Answers facility that we are using right now does not allow the code to run because the plotfcn tries to create an interactive "Stop" button for the plot and the Answers faciity does not support interactive buttons.

[sogol bandekian]

I did this in MATLAB online but the plot are empty.how I can fix this?

[Walter Roberson]

The plots are not empty. The optimization is configured for a maximum number of generations and that is the x axis used so that it does not get continually redrawn as more generations are processed. The optimization either succeeded or failed in a small number of generations, on the order of 5 generations.

Answer 2

[Bhavana Ravirala]

Hi Sogol,

When I try to reproduce the issue at my end I encountered the error as Function definition not supported in this context. Create functions in a code file. If you are facing the same error refer to the below link to eliminate the error.

Error: Function definition not supported in this context. Create functions in code file. - (mathworks.com)

If not, can you send us the exact error message that you are getting.

Hope this helps!!

[sogol bandekian]

hello thank you for reply.these below errors are what I got

%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

Error using makeState (line 61)

Your fitness function must return a scalar value.

Error in gacon (line 44)

state = makeState(GenomeLength,SubFitness,Iterate,subtype,options);

Error in ga (line 406)

[x,fval,exitFlag,output,population,scores] = gacon(FitnessFcn,nvars, ...

Error in maingl (line 11)

[x , fval,exitFlag,output,population,scores]= ga(ObjFcn,nvars,[],[],[],[],lb,ub,nonlcon,options);