Q: DemeGA Problem / zeroed genomes

Oliver Brockmann oliver.brockmann at numeris.de
Fri Jun 8 04:43:48 EDT 2001

I'm using this config:

- IRIX (cc)
- DemeGA with 4 populations (50 individuals each)
- GARealGenome 
- GARealAlleleSet
  (50, 12000, 50,GAAllele::INCLUSIVE,GAAllele::INCLUSIVE)
- nMigration = 5

When the GA evolves I dump the genes from the objective function
to a log file when evaluating and observe the following:

First there are the 200 genomes from the initialized populations
with acceptable values respecting the definitions from the AlleleSet.
However then 146 genomes are evaluated which consist solely of zeroes

After that the GA continues with sensfull genomes again.

I cannot figure out where the zero-genomes come from and how to get
rid of them. Can anybody help?

Thanks for your comments!



More information about the galib mailing list