How to "always evaluate"?

[Anthony Liekens]

Hi,

I'm using GALib for optimizing royal road functions, but I have a small
question...

The fitness function is using a random sampler - i can't generate a
fitness function that returns a complete result (over the whole pack of
data)
This means that a fitness function can be lower or higher than it's
average value, since the sampled data used for the fitness function is
different each time a genome is evaluated with this fitness function.

Therefore I'm looking for a possibility to evaluate the fitness function
of all genomes in the population, each generation, whether the genes
have changed or not.

If you have an answer to this problem - please mail your answer to
liekens at alife.santafe.edu - it looks like I'm not subscribed in *some*
way to the mailing list...

Thanx,

Anthony Liekens