Subject: | |
From: | |
Reply To: | |
Date: | Tue, 21 Oct 2008 12:13:52 -0400 |
Content-Type: | text/plain |
Parts/Attachments: |
|
|
Hi all,
I have a single subpopulation (later I'll work with more than one),
using SimpleEvaluator. My Problem is a SimpleProblemForm. After all of
the individuals in a subpop have been evaluated, I need to adjust
their fitnesses by a factor that depends on the original fitnesses and
sizes of all the individuals in the subpop [I'm using the covar(L,F)/
var(L) idea from Poli & McPhee 2008].
When I wasn't using Master/Slave, I could just put the calculation in
my Problem's finishEvaluating method.
However, with Master/Slave, my Problem's finishEvaluating method isn't
called. The documentation (in ec/eval/README) implies that it should
be called:
"When SimpleEvaluator calls finishEvaluating, the
remaining individuals are sent out as a (possibly short) job, and then
the MasterProblem blocks waiting until all the individuals have come
back."
Where should I stick the fitness adjustment code?
Thanks,
--Rob
|
|
|