source: framspy/evolalg/examples/niching_novelty.py @ 1113

import argparse
import os
import sys
from enum import Enum

import numpy as np

from FramsticksLib import FramsticksLib
from evolalg.base.lambda_step import LambdaStep
from evolalg.dissimilarity.frams_dissimilarity import FramsDissimilarity
from evolalg.dissimilarity.levenshtein import LevenshteinDissimilarity
from evolalg.experiment import Experiment
from evolalg.fitness.fitness_step import FitnessStep
from evolalg.mutation_cross.frams_cross_and_mutate import FramsCrossAndMutate
from evolalg.population.frams_population import FramsPopulation
from evolalg.repair.remove.field import FieldRemove
from evolalg.repair.remove.remove import Remove
from evolalg.selection.tournament import TournamentSelection
from evolalg.statistics.halloffame_stats import HallOfFameStatistics
from evolalg.statistics.statistics_deap import StatisticsDeap
from evolalg.base.union_step import UnionStep
from evolalg.utils.population_save import PopulationSave


def ensureDir(string):
    if os.path.isdir(string):
        return string
    else:
        raise NotADirectoryError(string)


class Dissim(Enum):
    levenshtein = "levenshtein"
    frams = "frams"

    def __str__(self):
        return self.name


class Fitness(Enum):
    raw = "raw"
    niching = "niching"
    novelty = "novelty"

    def __str__(self):
        return self.name


def parseArguments():
    parser = argparse.ArgumentParser(
        description='Run this program with "python -u %s" if you want to disable buffering of its output.' % sys.argv[
            0])
    parser.add_argument('-path', type=ensureDir, required=True, help='Path to Framsticks without trailing slash.')
    parser.add_argument('-opt', required=True,
                        help='optimization criteria : vertpos, velocity, distance, vertvel, lifespan, numjoints, numparts, numneurons, numconnections. Single or multiple criteria.')
    parser.add_argument('-lib', required=False, help="Filename of .so or .dll with framsticks library")
    parser.add_argument('-genformat', required=False, default="1",
                        help='Genetic format for the demo run, for example 4, 9, or B. If not given, f1 is assumed.')
    parser.add_argument('-sim', required=False, default="eval-allcriteria.sim", help="Name of .sim file")
    parser.add_argument('-dissim', required=False, type=Dissim, default=Dissim.frams,
                        help=' Dissimilarity measure DEFAULT = frams', choices=list(Dissim))
    parser.add_argument('-fit', required=False, default=Fitness.raw, type=Fitness,
                        help=' Fitness criteria DEFAULT = raw', choices=list(Fitness))
    parser.add_argument('-popsize', type=int, default=50, help="Size of population, default 50.")
    parser.add_argument('-num_parts', type=int, default=None, help="Maximum number of parts. Default None")
    parser.add_argument('-checkpoint_path', required=False, default=None, help="Path to checkpoint path")
    parser.add_argument('-checkpoint_interval', required=False, type=int, default=100, help="Checkpoint interval")
    return parser.parse_args()


def extract_fitness(ind):
    return ind.fitness_raw


def print_population_count(pop):
    print("Current:", len(pop))
    return pop  # Each step must return a population


class NumPartsGreater(Remove):
    def __init__(self, numparts):
        super(NumPartsGreater, self).__init__()
        self.numparts = numparts

    def remove(self, individual):
        return individual.numparts > self.numparts


def func_niching(ind): setattr(ind, "fitness", ind.fitness_raw * (1 + ind.dissim))


def func_raw(ind): setattr(ind, "fitness", ind.fitness_raw)


def func_novelty(ind): setattr(ind, "fitness", ind.dissim)


def main():
    print("Running experiment with", sys.argv)
    parsed_args = parseArguments()
    frams = FramsticksLib(parsed_args.path, parsed_args.lib,
                          parsed_args.sim)
    # Steps for generating first population
    init_stages = [FramsPopulation(frams, parsed_args.genformat, parsed_args.popsize)]

    # Selection procedure
    selection = TournamentSelection(5, copy=True) # 'fitness' by default, the targeted attribute can be changed, e.g. fit_attr="fitness_raw"

    # Procedure for generating new population. This steps will be run as long there is less than
    # popsize individuals in the new population
    new_generation_stages = [FramsCrossAndMutate(frams, cross_prob=0.2, mutate_prob=0.9)]

    # Steps after new population is created. Executed exacly once per generation.
    generation_modifications = []

    # -------------------------------------------------
    # Fitness

    fitness_raw = FitnessStep(frams, fields={parsed_args.opt: "fitness_raw", "numparts": "numparts"},
                              fields_defaults={parsed_args.opt: None, "numparts": float("inf")},
                              evaluation_count=1)

    fitness_end = FitnessStep(frams, fields={parsed_args.opt: "fitness_raw"},
                              fields_defaults={parsed_args.opt: None},
                              evaluation_count=100)  # evaluate the contents of the last population 100 times (TODO replace this approach and evaluate HOF instead of the last population)
    # Remove
    remove = []
    remove.append(FieldRemove("fitness_raw", None))  # Remove individuals if they have default value for fitness
    if parsed_args.num_parts is not None:
        # This could be also implemented by "LambdaRemove(lambda x: x.numparts > parsed_args.num_parts)
        # But this will not serialize in checkpoint.
        remove.append(NumPartsGreater(parsed_args.num_parts))
    remove_step = UnionStep(remove)

    fitness_remove = UnionStep([fitness_raw, remove_step])

    init_stages.append(fitness_remove)
    new_generation_stages.append(fitness_remove)

    # -------------------------------------------------
    # Novelty or niching
    dissim = None
    if parsed_args.dissim == Dissim.levenshtein:
        dissim = LevenshteinDissimilarity(reduction="mean", output_field="dissim")
    elif parsed_args.dissim == Dissim.frams:
        dissim = FramsDissimilarity(frams, reduction="mean", output_field="dissim")

    if parsed_args.fit == Fitness.raw:
        # Fitness is equal to finess raw
        raw = LambdaStep(func_raw)
        init_stages.append(raw)
        new_generation_stages.append(raw)
        generation_modifications.append(raw)

    if parsed_args.fit == Fitness.niching:
        niching = UnionStep([
            dissim,
            LambdaStep(func_niching)
        ])
        init_stages.append(niching)
        new_generation_stages.append(niching)
        generation_modifications.append(niching)

    if parsed_args.fit == Fitness.novelty:
        novelty = UnionStep([
            dissim,
            LambdaStep(func_novelty)
        ])
        init_stages.append(novelty)
        new_generation_stages.append(novelty)
        generation_modifications.append(novelty)

    # -------------------------------------------------
    # Statistics
    hall_of_fame = HallOfFameStatistics(100, "fitness_raw")  # Wrapper for halloffamae
    statistics_deap = StatisticsDeap([
        ("avg", np.mean),
        ("stddev", np.std),
        ("min", np.min),
        ("max", np.max)
    ], extract_fitness)  # Wrapper for deap statistics

    statistics_union = UnionStep([
        hall_of_fame,
        statistics_deap
    ])  # Union of two statistics steps.

    init_stages.append(statistics_union)
    generation_modifications.append(statistics_union)

    # -------------------------------------------------
    # End stages: this will execute exacly once after all generations.
    end_stages = [
        fitness_end,
        PopulationSave("halloffame.gen", provider=hall_of_fame.haloffame, fields={"genotype": "genotype",
                                                                                            "fitness": "fitness_raw"})]
    # ...but custom fields can be added, e.g. "custom": "recording"

    # -------------------------------------------------
    # Experiment creation

    experiment = Experiment(init_population=init_stages,
                            selection=selection,
                            new_generation_steps=new_generation_stages,
                            generation_modification=generation_modifications,
                            end_steps=end_stages,
                            population_size=parsed_args.popsize,
                            checkpoint_path=parsed_args.checkpoint_path,
                            checkpoint_interval=parsed_args.checkpoint_interval
                            )
    experiment.init()  # init is mandatory
    experiment.run(10)
    # Next call for experiment.run(10) will do nothing. Parameter 10 specifies how many generations should be
    # in one experiment. Previous call generated 10 generations.
    # Example 1:
    # experiment.init()
    # experiment.run(10)
    # experiment.run(12)
    # #This will run for total of 12 generations
    #
    # Example 2
    # experiment.init()
    # experiment.run(10)
    # experiment.init()
    # experiment.run(10)
    # # All work produced by first run will be "destroyed" by second init().

    for ind in hall_of_fame.haloffame:
        print("%g\t%s" % (ind.fitness, ind.genotype))


if __name__ == '__main__':
    main()