How to write signature of custom GA functions? (Selection, Mutation, Crossover)

Question

It is hard for me to find Documentation on how to write signatures of my custom selection, mutation, crossover functions for Genetic Algorithm.

I can't figure out how ga() functions works. (I know this exists: link ), but how these functions communicate, what they are expecting for inputs and for outputs?

Q&A, More about my problem: link

This are my GA options:

options = gaoptimset(...
                      'PopulationSize',    10, ...
                      'Generations',       50, ...
                      'CrossoverFcn',      {'crossoverscattered'}, ...
                      'MutationFcn',       { @Mutation, 0.01 }, ...
                      'SelectionFcn',      { @RouletteWheelSelection }, ...
                      'UseParallel',       true, ...
                      'Display', 'iter' ...
                     );

My GA function:

lb = 1;  % Lower bound on x
ub = 3;  % Upper bound on x
nvars = 1;
x = ga(@GaFitness,nvars,[],[],[],[],lb,ub,[],[],options);

My GaFitness Fcn signature (for now):

function result  = GaFitness()

My Mutation function signature:

 function chromosome  = Mutation( chromosome )

My Selection function signature:

function selected_chromosome = RouletteWheelSelection( population )

If you need more info about my problem or more code, pls let me know.

This is Error I'm getting:

Error using GaFitness

Too many input arguments.

Error in createAnonymousFcn>@(x)fcn(x,FcnArgs{:}) (line 11)

fcn_handle = @(x) fcn(x,FcnArgs{:});

Error in makeState (line 47)

            firstMemberScore = FitnessFcn(state.Population(initScoreProvided+1,:));

Error in galincon (line 17)
state = makeState(GenomeLength,FitnessFcn,Iterate,output.problemtype,options);

Error in ga (line 374)
            [x,fval,exitFlag,output,population,scores] = galincon(FitnessFcn,nvars, ...

Caused by:
    Failure in initial user-supplied fitness function evaluation. GA cannot continue.

Using Matlab 2016b.

Well, I'm trying to figure out what my input needs to be, also my output. I assume it will be a chromosome which is a set of rules for a fuzzy controller. This is my Q&A about my problem: [link](https://stackoverflow.com/questions/56689104/ga-custom-integer-selection-with-roulette-wheel) — , Jun 25 '19 at 19:06
What's the function you want to optimize? `GaFitness()` takes a chromosome as input and returns a value, the higher values one get their chromosome selected for `RouletteWheelSelection` tournament — Adam, Jun 25 '19 at 19:26
@adam.Yes, I want to optimize `GaFitness()`, more precisely Fuzzy control rules inside `GaFitness`. So, I assume input will be a set of rules (chromosome, exp.: [`1,2,3,1,1..]` and based on that input, return a result of xy variable that was impacted of new fuzzy settings. — , Jun 25 '19 at 20:49
Perhaps [this example](https://stackoverflow.com/a/42838593/3372061) could help you. — Dev-iL, Jun 26 '19 at 10:16
@Dev-iL I looked at your answer that you linked and basically if I understood right, inputs are irrelevant, but the output must be as documented. In ˙GaFitness˙ input will be chromosome, and output result I want to minimize. Could you help me with other functions? (Mutation/Selection) Where can I look at what they need for input/output? Some example maybe? — , Jun 27 '19 at 20:38

How to write signature of custom GA functions? (Selection, Mutation, Crossover)

0 Answers0

Linked