Caching of Fitness Values for Elites/Kept Parents

[kingjin94]

Hi there,

Thank you for your great toolbox!

I have a question regarding caching of fitness values in pygad. It seems that any elite or kept parents fitness function will always be just retrieved. Is there a way to force recalculation? In my case, the fitness is non-deterministic, and I have local search capabilities, i.e., Lamarkian evolution, that would be especially useful for the promising elites or selected parents to further improve their fitness if they somehow could be evaluated again.

[ahmedfgad]

This is covered in this section of the documentation: https://pygad.readthedocs.io/en/latest/pygad_more.html#solve-non-deterministic-problems

import pygad
...
ga_instance = pygad.GA(...,
                       keep_elitism=0,
                       keep_parents=0,
                       save_solutions=False,
                       save_best_solutions=False,
                       ...)