Is it possible to use ga just to generate the population and use another piece of code to run it?

As the question says, I wonder if it is possible to use ga just to generate the population and use another piece of code to run it. Once I run it I will give the results to ga to generate the next population to be run by the second piece of code and so on. I hope my question is clear. My object here is to have "full control" on the optimization. I am quoting full control because I am pretty sure I can do pretty much everything I want with just ga but I would like to skip the learning phase...

Thank you in advance for any advice. Guido

[Geoff Hayes]

Guido - when you say that you want to use ga just to generate the population and use another piece of code to run it, are you saying that you want your other code to do the parent selection, crossover and mutation? So the function GA should just create a random population of members? Or do you still want the genetic algorithm to do the selection, crossover and mutation and have your other code manipulate this new population just to feed it back into the GA for another round of selection, crossover, and mutation?

Answer 1

[Prateekshya]

Hello Guido,

As per my understanding you want to extract the population of each generation. You may follow the below link to know more about it:

https://www.mathworks.com/matlabcentral/answers/568248-extracting-population-generated-in-each-generation-of-genetic-algorithm-ga

However, you may want to implement the population generation code as well since you will be able to customize it and get more control over the settings of the algorithm. Please go through the "Initialization" section of the below link:

https://en.wikipedia.org/wiki/Genetic_algorithm#Methodology

I hope this helps!

Thank you.