I'm using ga solver with custom population type,
each entities is a binary vector,
I'm writing the crossover and the mutation functions,
How can I avoid the creation of children olready tested?
I'm searching for a solution compatible with the parallelization.
Thanks in advance for the answers,