Using maximize.1 fixed it.

The resulting Pareto front is significantly different than it was when I
use the reciprocal fitness to hack minimization.  I'll have to think about
that.

Good that you updated the manual.  I looked there first (in an old
version), but hadn't seen it.

Thanks a bunch!
Siggy

PS: Yes, I'm using the SVN, but a somewhat outdated version.


On Tue, Sep 24, 2013 at 11:27 PM, Sean Luke <[log in to unmask]> wrote:

> First things first: are you using the SVN version?  BTW, I just updated
> the manual to make clear how maximize/minimize works.
>
> Next...
>
> On Sep 24, 2013, at 10:27 PM, Eric 'Siggy' Scott wrote:
>
> > I have tried this:
> >
> > multi.fitness.0.maximize=true
> > multi.fitness.1.maximize=false
> > multi.fitness.2.maximize=false
> >
> > And this:
> >
> > pop.subpop.0.species.fitness.0.maximize=true
> > pop.subpop.0.species.fitness.1.maximize=false
> > pop.subpop.0.species.fitness.2.maximize=false
>
> It's maximize.1, not 1.maximize.  Any chance this is the issue?
>
> If we've got further concerns with per-objective maximization and
> minimization, time to call in Khaled, who was instrumental in updating ECJ
> to handle it.
>
> Sean
>
>
>


-- 

Ph.D student in Computer Science
George Mason University
http://mason.gmu.edu/~escott8/