Genetic Algorithm, argument should be real?

Hi all.. I want to ask about my objective function :

function f = function1(x)
p=(atand((x(4))/(x(5)-100)));
g=(((x(5)-100)^2+x(4)^2)^0.5);
s=((x(1)^2+g^2-x(2)^2)/(2*x(1)*g));
t=((x(2)^2+g^2-x(1)^2)/(2*x(2)*g));
alpha=((acosd(s))+p);
betha=((acosd(t))-p);
f=((x(2)*sind(alpha+betha))/(2*x(3)*cosd(betha)));

after execution with genetic algorithm solver the result are :

??? Error using ==> sind at 16 Argument should be real.

Error in ==> fungsi1 at 8 f=((x(2)*sind(alpha+betha))/(2*x(3)*cosd(betha))); Error in ==> validate>@(x)fitness(x,FitnessFcnArgs{:}) at 135 fitness = @(x) fitness(x,FitnessFcnArgs{:});

Error in ==> fcnvectorizer at 23 y = fun(pop);

Error in ==> makeState at 48 Score = fcnvectorizer(state.Population(initScoreProvided+1:end,:),FitnessFcn,1,options.SerialUserFcn);

Error in ==> gaunc at 41 state = makeState(GenomeLength,FitnessFcn,Iterate,output.problemtype,options);

Error in ==> ga at 279 [x,fval,exitFlag,output,population,scores] = gaunc(FitnessFcn,nvars, ...

sind not real argument? why? because value of p not real?

[EDITED, Jan, code formatted]

[Jan]

Please learn how to format your code. It is easy and has been explained repeatedly. Thanks.

Answer 1

[Jan]

If p is not real, alpha and betha are not real also in general. Then even alpha+betha is not real also in general. And then sind() cannot operate on this value, as the error message clearly explains.

You can use the debugger to observe the values of the local variables to find the reasons for the imaginary values.

[Andy Rahmawan]

then i use absolute on value of p, then GA still didn't work,, what happened?

??? Error using ==> sind at 16 Argument should be real.

function f = function1(x) p=(abs(atand((x(4))/(x(5)-100)))); g=(((x(5)-100)^2+x(4)^2)^0.5); s=((x(1)^2+g^2-x(2)^2)/(2*x(1)*g)); t=((x(2)^2+g^2-x(1)^2)/(2*x(2)*g)); alpha=((acosd(s))+p); betha=((acosd(t))-p); f=((x(2)*sind(alpha+betha))/(2*x(3)*cosd(betha)));

[Jan]

Please, Andy, format your code properly. A bad readability is not useful, when you want to get help.

Again I suggest to use the debugger to find the reason for the imaginary values. We do not have your input data, such that we cannot do the debugging for you. E.g.:

dbstop if error

Now Matlab stops when the error occurres and you can check the values of the local variables in the command window or workspace browser.

[Andy Rahmawan]

%code starts from here
function f = function1(x)
p=(abs(atand((x(4))/(x(5)-100))));
g=(((x(5)-100)^2+x(4)^2)^0.5);
s=((x(1)^2+g^2-x(2)^2)/(2*x(1)*g));
t=((x(2)^2+g^2-x(1)^2)/(2*x(2)*g));
alpha=((acosd(s))+p);
betha=((acosd(t))-p);
f=((x(2)*sind(alpha+betha))/(2*x(3)*cosd(betha)));

this code u mean? sorry i'm noob here..

then i use absolute on value of p, then GA still didn't work,, what happened?