[SOLVED] Only one chromosome shows good results using Jenetics

Only one chromosome shows good results using Jenetics

I am using Jenetics library to solve a problem with ga. I am extending the official example to use several chromosomes like this:

    List<BitChromosome> arr = new ArrayList<>();
    arr.add(BitChromosome.of(16, 0.5));
    arr.add(BitChromosome.of(16, 0.5));
    arr.add(BitChromosome.of(16, 0.5));
    Factory<Genotype<BitGene>> gtf = Genotype.of(arr);

And change the eval function to have exactly 8 1s and 8 0s:

    private static int eval(Genotype<BitGene> gt) {
    return 10 - Math.abs(gt.getChromosome()
            .as(BitChromosome.class)
            .bitCount()-8);

The other parts were left unchanged:

    // 3.) Create the execution environment.
    Engine<BitGene, Integer> engine = Engine
            .builder(Test1::eval, gtf)
            .build();

    // 4.) Start the execution (evolution) and
    //     collect the result.
    Genotype<BitGene> result = engine.stream()
            .limit(100)
            .collect(EvolutionResult.toBestGenotype());

I was expecting the ga go produce 3 chromosomes which maximize this eval function but I am getting:

[01110010|00010111,01000000|00000100,10011101|01110110]

As you can see, only the first result satisfy the condition. How can I extend this example so all chromosomes will maximize the evaluation function?

Solution

This is exactly what I would expect after looking at the fitness function. You are using only the first chromosome for calculating the fitness. The Genotype.getChromosome() method returns the first chromosome. It is a shortcut for Genotype.getChromosome(0). The other two chromosomes are not considered in your fitness function.