Experimental tuning

and minor bugfixes
and added more typing documentation to functions

Author: portaloffreedom

experiments/brain-speciation/MorphologyCompatibility.py

@@ -0,0 +1,138 @@
+from __future__ import annotations
+
+from typing import TYPE_CHECKING
+if TYPE_CHECKING:
+    from typing import Optional
+    from pyrevolve.evolution.individual import Individual
+
+
+class MorphologyCompatibility:
+    def __init__(self,
+                 total_threshold: float = 1.0,
+                 branching_modules_count: float = 0.0,
+                 branching: float = 0.0,
+                 extremities: float = 0.0,
+                 limbs: float = 0.0,
+                 extensiveness: float = 0.0,
+                 length_of_limbs: float = 0.0,
+                 coverage: float = 0.0,
+                 joints: float = 0.0,
+                 active_hinges_count: float = 0.0,
+                 proportion: float = 0.0,
+                 width: float = 0.0,
+                 height: float = 0.0,
+                 z_depth: float = 0.0,
+                 absolute_size: float = 0.0,
+                 size: float = 0.0,
+                 sensors: float = 0.0,
+                 symmetry: float = 0.0,
+                 hinge_count: float = 0.0,
+                 brick_count: float = 0.0,
+                 brick_sensor_count: float = 0.0,
+                 touch_sensor_count: float = 0.0,
+                 free_slots: float = 0.0,
+                 height_base_ratio: float = 0.0,
+                 max_permitted_modules: Optional[int] = None,
+                 symmetry_vertical: float = 0.0,
+                 base_density: float = 0.0,
+                 bottom_layer: float = 0.0,
+                 ):
+        # Total threshold
+        self.total_threshold: float = total_threshold
+
+        # Absolute branching
+        self.branching_modules_count: float = branching_modules_count
+        # Relative branching
+        self.branching: float = branching
+        # Absolute number of limbs
+        self.extremities: float = extremities
+        # Relative number of limbs
+        self.limbs: float = limbs
+        # Absolute length of limbs
+        self.extensiveness: float = extensiveness
+        # Relative length of limbs
+        self.length_of_limbs: float = length_of_limbs
+        # Coverage
+        self.coverage: float = coverage
+        # Relative number of effective active joints
+        self.joints: float = joints
+        # Absolute number of effective active joints
+        self.active_hinges_count: float = active_hinges_count
+        # Proportion
+        self.proportion: float = proportion
+        # Width
+        self.width: float = width
+        # Height
+        self.height: float = height
+        # Z depth
+        self.z_depth: float = z_depth
+        # Absolute size
+        self.absolute_size: float = absolute_size
+        # Relative size in respect of the max body size `max_permitted_modules`
+        self.size: float = size
+        # Proportion of sensor vs empty slots
+        self.sensors: float = sensors
+        # Body symmetry in the xy plane
+        self.symmetry: float = symmetry
+        # Number of active joints
+        self.hinge_count: float = hinge_count
+        # Number of bricks
+        self.brick_count: float = brick_count
+        # Number of brick sensors
+        self.brick_sensor_count: float = brick_sensor_count
+        # Number of touch sensors
+        self.touch_sensor_count: float = touch_sensor_count
+        # Number of free slots
+        self.free_slots: float = free_slots
+        # Ratio of the height over the root of the area of the base
+        self.height_base_ratio: float = height_base_ratio
+        # Maximum number of modules allowed (sensors excluded)
+        self.max_permitted_modules: Optional[int] = max_permitted_modules
+        # Vertical symmetry
+        self.symmetry_vertical: float = symmetry_vertical
+        # Base model density
+        self.base_density: float = base_density
+        # Bottom layer of the robot
+        self.bottom_layer: float = bottom_layer
+
+    def compatible_individuals(self,
+                               individual1: Individual,
+                               individual2: Individual) -> bool:
+        morph_measure_1 = individual1.phenotype.measure_body()
+        morph_measure_2 = individual2.phenotype.measure_body()
+        _1 = morph_measure_1
+        _2 = morph_measure_2
+
+        # TODO consider normalization of some of these values, some are already normalized by definition
+
+        total_distance: float = 0.0
+        total_distance += self.branching_modules_count * abs(_2.branching_modules_count - _1.branching_modules_count)
+        total_distance += self.branching * abs(_2.branching - _1.branching)
+        total_distance += self.extremities * abs(_2.extremities - _1.extremities)
+        total_distance += self.limbs * abs(_2.limbs - _1.limbs)
+        total_distance += self.extensiveness * abs(_2.extensiveness - _1.extensiveness)
+        total_distance += self.length_of_limbs * abs(_2.length_of_limbs - _1.length_of_limbs)
+        total_distance += self.coverage * abs(_2.coverage - _1.coverage)
+        total_distance += self.joints * abs(_2.joints - _1.joints)
+        total_distance += self.active_hinges_count * abs(_2.active_hinges_count - _1.active_hinges_count)
+        total_distance += self.proportion * abs(_2.proportion - _1.proportion)
+        total_distance += self.width * abs(_2.width - _1.width)
+        total_distance += self.height * abs(_2.height - _1.height)
+        total_distance += self.z_depth * abs(_2.z_depth - _1.z_depth)
+        total_distance += self.absolute_size * abs(_2.absolute_size - _1.absolute_size)
+        if self.max_permitted_modules is not None:
+            total_distance += self.size * \
+                              abs(_2.absolute_size - _1.absolute_size) / self.max_permitted_modules
+        total_distance += self.sensors * abs(_2.sensors - _1.sensors)
+        total_distance += self.symmetry * abs(_2.symmetry - _1.symmetry)
+        total_distance += self.hinge_count * abs(_2.hinge_count - _1.hinge_count)
+        total_distance += self.brick_count * abs(_2.brick_count - _1.brick_count)
+        total_distance += self.brick_sensor_count * abs(_2.brick_sensor_count - _1.brick_sensor_count)
+        total_distance += self.touch_sensor_count * abs(_2.touch_sensor_count - _1.touch_sensor_count)
+        total_distance += self.free_slots * abs(_2.free_slots - _1.free_slots)
+        total_distance += self.height_base_ratio * abs(_2.height_base_ratio - _1.height_base_ratio)
+        total_distance += self.symmetry_vertical * abs(_2.symmetry_vertical - _1.symmetry_vertical)
+        total_distance += self.base_density * abs(_2.base_density - _1.base_density)
+        total_distance += self.bottom_layer * abs(_2.bottom_layer - _1.bottom_layer)
+
+        return total_distance <= self.total_threshold

experiments/brain-speciation/manager.py

@@ -1,7 +1,9 @@
 #!/usr/bin/env python3
+from __future__ import annotations
+
 from pyrevolve import parser
 from pyrevolve.evolution import fitness
-from pyrevolve.evolution.selection import multiple_selection, tournament_selection
+from pyrevolve.evolution.selection import multiple_selection_with_duplicates, tournament_selection
 from pyrevolve.evolution.speciation.population_speciated import PopulationSpeciated
 from pyrevolve.evolution.speciation.population_speciated_config import PopulationSpeciatedConfig
 from pyrevolve.evolution.speciation.population_speciated_management import steady_state_speciated_population_management
@@ -11,13 +13,17 @@
 from pyrevolve.genotype.lsystem_neat.mutation import LSystemNeatMutationConf as lMutationConfig
 from pyrevolve.genotype.plasticoding.mutation.mutation import MutationConfig as plasticMutationConfig
 from pyrevolve.genotype.lsystem_neat.mutation import standard_mutation as lmutation
-
 from pyrevolve.util.supervisor.analyzer_queue import AnalyzerQueue
 from pyrevolve.util.supervisor.simulator_queue import SimulatorQueue
 from pyrevolve.custom_logging.logger import logger
 from pyrevolve.genotype.plasticoding import PlasticodingConfig
 from pyrevolve.genotype.lsystem_neat.lsystem_neat_genotype import LSystemCPGHyperNEATGenotype, LSystemCPGHyperNEATGenotypeConfig
 from pyrevolve.genotype.neat_brain_genome.neat_brain_genome import NeatBrainGenomeConfig
+from .MorphologyCompatibility import MorphologyCompatibility
+
+from typing import TYPE_CHECKING
+if TYPE_CHECKING:
+    from pyrevolve.evolution.individual import Individual
 
 
 async def run():
@@ -31,7 +37,7 @@ async def run():
     offspring_size = 50
 
     body_conf = PlasticodingConfig(
-        max_structural_modules=20,  # TODO increase
+        max_structural_modules=20,
         allow_vertical_brick=False,
         use_movement_commands=True,
         use_rotation_commands=False,
@@ -67,17 +73,32 @@ async def run():
     crossover_conf = lCrossoverConfig(
         crossover_prob=0.8,
     )
+
+    compatibitity_tester = MorphologyCompatibility(
+        total_threshold=1.0,
+        size=1.0,
+        brick_count=1.0,
+        proportion=1.0,
+        coverage=1.0,
+        joints=1.5,
+        branching=1.0,
+        symmetry=0.0,
+        max_permitted_modules=body_conf.max_structural_modules,
+    )
+
     # experiment params #
 
     # Parse command line / file input arguments
     args = parser.parse_args()
     experiment_management = ExperimentManagement(args)
+    has_offspring = False
     do_recovery = args.recovery_enabled and not experiment_management.experiment_is_new()
 
     logger.info(f'Activated run {args.run} of experiment {args.experiment_name}')
 
     if do_recovery:
-        gen_num, has_offspring, next_robot_id, next_species_id = experiment_management.read_recovery_state(population_size, offspring_size)
+        gen_num, has_offspring, next_robot_id, next_species_id = \
+            experiment_management.read_recovery_state(population_size, offspring_size, species=True)
 
         if gen_num == num_generations-1:
             logger.info('Experiment is already complete.')
@@ -97,9 +118,15 @@ async def run():
     if next_species_id < 0:
         next_species_id = 1
 
-    def are_genomes_compatible_fn(genotype1: LSystemCPGHyperNEATGenotype,
-                                  genotype2: LSystemCPGHyperNEATGenotype) -> bool:
-        return genotype1.is_brain_compatible(genotype2, genotype_conf)
+    def are_individuals_brains_compatible_fn(individual1: Individual,
+                                             individual2: Individual) -> bool:
+        assert isinstance(individual1.genotype, LSystemCPGHyperNEATGenotype)
+        assert isinstance(individual2.genotype, LSystemCPGHyperNEATGenotype)
+        return individual1.genotype.is_brain_compatible(individual2.genotype, genotype_conf)
+
+    def are_individuals_morphologies_compatible_fn(individual1: Individual,
+                                                   individual2: Individual) -> bool:
+        return compatibitity_tester.compatible_individuals(individual1, individual2)
 
     population_conf = PopulationSpeciatedConfig(
         population_size=population_size,
@@ -111,18 +138,19 @@ def are_genomes_compatible_fn(genotype1: LSystemCPGHyperNEATGenotype,
         crossover_operator=lcrossover,
         crossover_conf=crossover_conf,
         selection=lambda individuals: tournament_selection(individuals, 2),
-        parent_selection=lambda individuals: multiple_selection(individuals, 2, tournament_selection),
+        parent_selection=lambda individuals: multiple_selection_with_duplicates(individuals, 2, tournament_selection),
         population_management=steady_state_speciated_population_management,
         population_management_selector=tournament_selection,
         evaluation_time=args.evaluation_time,
         offspring_size=offspring_size,
         experiment_name=args.experiment_name,
         experiment_management=experiment_management,
         # species stuff
-        are_genomes_compatible_fn=are_genomes_compatible_fn,
+        # are_individuals_compatible_fn=are_individuals_brains_compatible_fn,
+        are_individuals_compatible_fn=are_individuals_morphologies_compatible_fn,
         young_age_threshold=5,
         young_age_fitness_boost=2.0,
-        old_age_threshold=20,
+        old_age_threshold=35,
         old_age_fitness_penalty=0.5,
         species_max_stagnation=30,
     )
@@ -142,12 +170,16 @@ def are_genomes_compatible_fn(genotype1: LSystemCPGHyperNEATGenotype,
                                      next_species_id)
 
     if do_recovery:
-        raise NotImplementedError('recovery not implemented')
         # loading a previous state of the experiment
         population.load_snapshot(gen_num)
         if gen_num >= 0:
             logger.info(f'Recovered snapshot {gen_num}, pop with {len(population.genus)} individuals')
 
+        # TODO partial recovery is not implemented, this is a substitute
+        has_offspring = False
+        next_robot_id = 1 + population.config.population_size + gen_num * population.config.offspring_size
+        population.next_robot_id = next_robot_id
+
         if has_offspring:
             raise NotImplementedError('partial recovery not implemented')
             recovered_individuals = population.load_partially_completed_generation(gen_num, population_size, offspring_size, next_robot_id)

pyrevolve/evolution/individual.py

@@ -6,6 +6,7 @@
     from typing import Optional, List
     from pyrevolve.revolve_bot import RevolveBot
     from pyrevolve.genotype import Genotype
+    from pyrevolve.evolution.speciation.species import Species
 
 
 class Individual:
@@ -38,24 +39,41 @@ def id(self) -> int:
             _id = self.genotype.id
         return _id
 
-    def export_genotype(self, folder) -> None:
+    def export_genotype(self, folder: str) -> None:
         self.genotype.export_genotype(os.path.join(folder, f'genotype_{self.phenotype.id}.txt'))
 
-    def export_phenotype(self, folder) -> None:
+    def export_phenotype(self, folder: str) -> None:
         if self.phenotype is None:
             self.develop()
         self.phenotype.save_file(os.path.join(folder, f'phenotype_{self.phenotype.id}.yaml'), conf_type='yaml')
 
-    def export_fitness(self, folder) -> None:
+    def export_phylogenetic_info(self, folder: str) -> None:
+        """
+        Export phylogenetic information
+        (parents and other possibly other information to build a phylogenetic tree)
+        :param folder: folder where to save the information
+        """
+        if self.parents is not None:
+            parents_ids: List[str] = [str(p.id) for p in self.parents]
+            parents_ids_str = ",".join(parents_ids)
+        else:
+            parents_ids_str = 'None'
+
+        filename = os.path.join(folder, f'parents_{self.id}.yaml')
+        with open(filename, 'w') as file:
+            file.write(f'parents:{parents_ids_str}')
+
+    def export_fitness_single_file(self, folder: str) -> None:
         """
         It's saving the fitness into a file. The fitness can be a floating point number or None
         :param folder: folder where to save the fitness
         """
         with open(os.path.join(folder, f'fitness_{self.id}.txt'), 'w') as f:
             f.write(str(self.fitness))
 
-    def export(self, folder) -> None:
+    def export(self, folder: str) -> None:
         self.export_genotype(folder)
+        self.export_phylogenetic_info(folder)
         self.export_phenotype(folder)
         self.export_fitness(folder)
 

pyrevolve/evolution/population/population.py

@@ -3,13 +3,13 @@
 import os
 
 from pyrevolve.evolution.individual import Individual
-from pyrevolve.tol.manage import measures
 from pyrevolve.custom_logging.logger import logger
 from pyrevolve.evolution.population.population_config import PopulationConfig
 
 from typing import TYPE_CHECKING
 if TYPE_CHECKING:
-    from typing import List, Optional, AnyStr
+    from typing import List, Optional
+    from pyrevolve.evolution.speciation.species import Species
     from pyrevolve.tol.manage.measures import BehaviouralMeasurements
     from pyrevolve.util.supervisor.analyzer_queue import AnalyzerQueue, SimulatorQueue
 
@@ -44,10 +44,15 @@ def __init__(self,
         self.simulator_queue = simulator_queue
         self.next_robot_id = next_robot_id
 
-    def _new_individual(self, genotype):
+    def _new_individual(self,
+                        genotype,
+                        parents: Optional[List[Individual]] = None):
         individual = Individual(genotype)
         individual.develop()
         individual.phenotype.update_substrate()
+        if parents is not None:
+            individual.parents = parents
+
         self.config.experiment_management.export_genotype(individual)
         self.config.experiment_management.export_phenotype(individual)
         self.config.experiment_management.export_phenotype_images(individual)

pyrevolve/evolution/selection.py

@@ -32,9 +32,12 @@ def tournament_selection(population: List[Individual], k=2) -> Individual:
 
 def multiple_selection(population: List[Individual],
                        selection_size: int,
-                       selection_function: Callable[[List[Individual]], Individual]) -> List[Individual]:
+                       selection_function: Callable[[List[Individual]], Individual]
+                       ) -> List[Individual]:
     """
-    Perform selection on population of distinct group, can be used in the form parent selection or survival selection
+    Perform selection on population of distinct group, it can be used in the
+    form parent selection or survival selection.
+    It never selects the same individual more than once
     :param population: list of individuals where to select from
     :param selection_size: amount of individuals to select
     :param selection_function:
@@ -49,3 +52,22 @@ def multiple_selection(population: List[Individual],
                 selected_individuals.append(selected_individual)
                 new_individual = True
     return selected_individuals
+
+
+def multiple_selection_with_duplicates(population: List[Individual],
+                                       selection_size: int,
+                                       selection_function: Callable[[List[Individual]], Individual]
+                                       ) -> List[Individual]:
+    """
+    Perform selection on population of distinct group, it can be used in the
+    form parent selection or survival selection.
+    It can select the same individual more than once
+    :param population: list of individuals where to select from
+    :param selection_size: amount of individuals to select
+    :param selection_function:
+    """
+    selected_individuals = []
+    for _ in range(selection_size):
+        selected_individual = selection_function(population)
+        selected_individuals.append(selected_individual)
+    return selected_individuals