If you're using NSGA2, your custom fitness class must be a subclass of NSGA2MultiObjectiveFitness. Furthermore, NSGA2 has its own way of handling fitness comparisons, so overriding betterThan will deviate from how NSGA2 works: you need to make sure you know what you're doing.
The only reason you'd subclass directly from MultiObjectiveFitness is if you're implementing your own multiobjective optimization algorithm.
> On Dec 9, 2018, at 4:35 PM, Adrián Romero Cáceres <[log in to unmask]> wrote:
> I have implemented a new betterThan operator into a CustomFitness class, which is an extended class of MultiObjectiveFitness. I wrote pop.subpop.0.species.fitness = ec.multiobjective.CustomFitness inside my experiment param file. Once the execution starts, after the first generation fails, returning the next error:
> Exception in thread "main" java.lang.ClassCastException: ec.multiobjective.CustomFitness cannot be cast to ec.multiobjective.nsga2.NSGA2MultiObjectiveFitness
> at ec.multiobjective.nsga2.NSGA2Breeder.assignFrontRanks(NSGA2Breeder.java:185)
> at ec.multiobjective.nsga2.NSGA2Breeder.buildArchive(NSGA2Breeder.java:121)
> at ec.multiobjective.nsga2.NSGA2Breeder.loadElites(NSGA2Breeder.java:82)
> at ec.simple.SimpleBreeder.breedPopulation(SimpleBreeder.java:225)
> at ec.multiobjective.nsga2.NSGA2Breeder.breedPopulation(NSGA2Breeder.java:113)
> at ec.simple.SimpleEvolutionState.evolve(SimpleEvolutionState.java:104)
> at ec.EvolutionState.run(EvolutionState.java:482)
> at ec.Evolve.main(Evolve.java:771)
> I forgot something? Is it the right way to extend the betterThan operator?
> Thanks in advance,