Hi Jeannie,
The way it is implemented, the steady-state replacement strategy proceeds by processing n individuals, where n is equal to the population size, to simulate a generation of the generational model. It first selects the m best individuals of the population (where n is equal to the value in ec.elite.keepsize) and set them apart in the population -- no variation operation is applied on them. Then, for the remaining (n-m) individuals it proceeds as usual, producing new individuals by crossover, mutation, and possibly reproduction (copy individual as is) if the breeder tree implements this.
Be sure to implement your steady-state algorithm using the SteadyStateOp as replacement strategy.
The crossovers and mutation operators work otherwize the same way than with in the generational model.
Best,
Christian
> --
>
> ---
> You received this message because you are subscribed to the Google Groups "openbeagle-users" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to
openbeagle-use...@googlegroups.com.
> For more options, visit
https://groups.google.com/groups/opt_out.
--
Christian Gagné
http://vision.gel.ulaval.ca/~cgagne