Merged @mstechly's statistics and index reset pull request.

Simplified Population statistics collection, and keep per-species fitness data.
Simplified visualize and statistics functions to take the Population object directly, and factored out common statistical operations on Population statistics into statistics.py.
Updated basic XOR example to show usage of statistics functions.

examples/pole_balancing/single_pole/ctrnn_evolve.py

@@ -34,8 +34,8 @@ def fitness_function(genomes):
 print(winner)
 
 # Plot the evolution of the best/average fitness.
-visualize.plot_stats(pop.most_fit_genomes, pop.fitness_scores, ylog=True, filename="ctrnn_fitness.svg")
+visualize.plot_stats(pop, ylog=True, filename="ctrnn_fitness.svg")
 # Visualizes speciation
-visualize.plot_species(pop.species_log, filename="ctrnn_speciation.svg")
+visualize.plot_species(pop, filename="ctrnn_speciation.svg")
 # Visualize the best network.
 visualize.draw_net(winner, view=True, filename="ctrnn_winner.gv")

examples/pole_balancing/single_pole/nn_evolve.py

@@ -33,8 +33,8 @@ def fitness_function(genomes):
     pickle.dump(winner, f)
 
 # Plot the evolution of the best/average fitness.
-visualize.plot_stats(pop.most_fit_genomes, pop.fitness_scores, ylog=True, filename="nn_fitness.svg")
+visualize.plot_stats(pop, ylog=True, filename="nn_fitness.svg")
 # Visualizes speciation
-visualize.plot_species(pop.species_log, filename="nn_speciation.svg")
+visualize.plot_species(pop, filename="nn_speciation.svg")
 # Visualize the best network.
 visualize.draw_net(winner, view=True, filename="nn_winner.gv")

examples/xor/xor2.py

@@ -1,7 +1,7 @@
 """ 2-input XOR example """
 from __future__ import print_function
 
-from neat import nn, population, visualize
+from neat import nn, population, statistics, visualize
 
 xor_inputs = [[0, 0], [0, 1], [1, 0], [1, 1]]
 xor_outputs = [0, 1, 1, 0]
@@ -35,7 +35,10 @@ def eval_fitness(genomes):
     output = winner_net.serial_activate(inputs)
     print("expected {0:1.5f} got {1:1.5f}".format(expected, output[0]))
 
-# Visualize the winner network and plot statistics.
-visualize.plot_stats(pop.most_fit_genomes, pop.fitness_scores)
-visualize.plot_species(pop.species_log)
+# Visualize the winner network and plot/log statistics.
+visualize.plot_stats(pop)
+visualize.plot_species(pop)
 visualize.draw_net(winner, view=True)
+statistics.save_stats(pop)
+statistics.save_species_count(pop)
+statistics.save_species_fitness(pop)

examples/xor/xor2_parallel.py

@@ -74,8 +74,8 @@ def fitness(genomes):
         print("{0:1.5f} \t {1:1.5f}".format(xor_outputs[i], output[0]))
 
     # Visualize the winner network and plot statistics.
-    visualize.plot_stats(pop.most_fit_genomes, pop.fitness_scores)
-    visualize.plot_species(pop.species_log)
+    visualize.plot_stats(pop)
+    visualize.plot_species(pop)
     visualize.draw_net(winner, view=True)
 
 

examples/xor/xor2_spiking.py

@@ -95,8 +95,8 @@ def run():
     # Visualize the winner network and plot statistics.
     winner = pop.most_fit_genomes[-1]
     visualize.draw_net(winner, view=True, node_names={0: 'A', 1: 'B', 2: 'Out1', 3: 'Out2'})
-    visualize.plot_stats(pop.most_fit_genomes, pop.fitness_scores)
-    visualize.plot_species(pop.species_log)
+    visualize.plot_stats(pop)
+    visualize.plot_species(pop)
 
     # Verify network output against training data.
     print('\nBest network output:')

neat/population.py

@@ -33,8 +33,7 @@ def __init__(self, config, checkpoint_file=None, initial_population=None,
 
         self.population = None
         self.species = []
-        self.species_log = []
-        self.fitness_scores = []
+        self.generation_statistics = []
         self.most_fit_genomes = []
         self.generation = -1
         self.total_evaluations = 0
@@ -50,6 +49,10 @@ def __init__(self, config, checkpoint_file=None, initial_population=None,
         # Partition the population into species based on current configuration.
         self._speciate()
 
+    @staticmethod
+    def clear_indexer(cls):
+        Species.clear_indexer()
+
     def _load_checkpoint(self, checkpoint):
         '''Resumes the simulation from a previous saved point.'''
         with gzip.open(checkpoint) as f:
@@ -130,10 +133,15 @@ def _speciate(self):
 
     def _log_stats(self):
         """ Gather data for visualization/reporting purposes. """
-        species_sizes = dict((s.ID, len(s.members)) for s in self.species)
-        self.species_log.append(species_sizes)
+        # Keep a deep copy of the best genome, so that future modifications to the genome
+        # do not produce an unexpected change in statistics.
         self.most_fit_genomes.append(copy.deepcopy(max(self.population)))
-        self.fitness_scores.append([c.fitness for c in self.population])
+
+        # Store the fitnesses of the members of each currently active species.
+        species_stats = {}
+        for s in self.species:
+            species_stats[s.ID] = [c.fitness for c in s.members]
+        self.generation_statistics.append(species_stats)
 
     def epoch(self, fitness_function, n, report=True, save_best=False, checkpoint_interval=10,
               checkpoint_generation=None):

neat/species.py

@@ -9,6 +9,10 @@ class Species(object):
     """ A collection of genetically similar individuals."""
     indexer = Indexer(1)
 
+    @classmethod
+    def clear_indexer(cls):
+        cls.indexer.clear()
+
     def __init__(self, first_individual, previous_id=None):
         self.representative = first_individual
         self.ID = Species.indexer.next(previous_id)
@@ -75,7 +79,3 @@ def reproduce(self, config):
         self.representative = random.choice(offspring)
 
         return offspring
-
-    def clearIndexer(self):
-        Species.indexer.clear()
-

neat/statistics.py

@@ -1,54 +1,78 @@
 # -*- coding: UTF-8 -*-
 from __future__ import print_function
-import warnings
+
 import csv
 
-try:
-    import numpy as np
-except ImportError:
-    np = None
-    warnings.warn('Could not import optional dependency NumPy.')
+from neat.math_util import mean
+
+
+def get_average_fitness(population):
+    avg_fitness = []
+    for stats in population.generation_statistics:
+        scores = []
+        for fitness in stats.values():
+            scores.extend(fitness)
+        avg_fitness.append(mean(scores))
+
+    return avg_fitness
+
+
+def get_species_sizes(population):
+    all_species = set()
+    for gen_data in population.generation_statistics:
+        all_species = all_species.union(gen_data.keys())
+
+    max_species = max(all_species)
+    species_counts = []
+    for gen_data in population.generation_statistics:
+        species = [len(gen_data.get(sid, [])) for sid in range(1, max_species + 1)]
+        species_counts.append(species)
+
+    return species_counts
 
-def save_stats(best_genomes, avg_scores, ylog=False, view=False, filename='fitness_history.csv'):
-    """ Saves the population's average and best fitness. """
-    csvfile = open(filename, 'wb')
-    statWriter = csv.writer(csvfile, delimiter=' ')
 
-    generation = range(len(best_genomes))
-    fitness = [c.fitness for c in best_genomes]
+def get_species_fitness(population, null_value=''):
+    all_species = set()
+    for gen_data in population.generation_statistics:
+        all_species = all_species.union(gen_data.keys())
 
-    for i in generation:
-        statWriter.writerow([fitness[i], avg_scores[i]])
+    max_species = max(all_species)
+    species_fitness = []
+    for gen_data in population.generation_statistics:
+        member_fitness = [gen_data.get(sid, []) for sid in range(1, max_species + 1)]
+        fitness = []
+        for mf in member_fitness:
+            if mf:
+                fitness.append(mean(mf))
+            else:
+                fitness.append(null_value)
+        species_fitness.append(fitness)
 
-    csvfile.close()
+    return species_fitness
 
 
-def save_species_count(species_log, view=False, filename='speciation.csv'):
-    """ Visualizes speciation throughout evolution. """
-    csvfile = open(filename, 'wb')
-    statWriter = csv.writer(csvfile, delimiter=' ')
+def save_stats(population, delimiter=' ', filename='fitness_history.csv'):
+    """ Saves the population's best and average fitness. """
+    with open(filename, 'w') as f:
+        w = csv.writer(f, delimiter=delimiter)
 
+        best_fitness = [c.fitness for c in population.most_fit_genomes]
+        avg_fitness = get_average_fitness(population)
+        for best, avg in zip(best_fitness, avg_fitness):
+            w.writerow([best, avg])
 
-    num_generations = len(species_log)
-    num_species = max(map(len, species_log))
-    curves = []
-    for gen in species_log:
-        species = [0] * num_species
-        species[:len(gen)] = gen
-        statWriter.writerow(species)
 
-    csvfile.close()
+def save_species_count(population, delimiter=' ', filename='speciation.csv'):
+    """ Log speciation throughout evolution. """
+    with open(filename, 'w') as f:
+        w = csv.writer(f, delimiter=delimiter)
+        for s in get_species_sizes(population):
+            w.writerow(s)
 
-def save_species_fitness(species_fitness_log, view=False, filename='species_fitness.csv'):
-    """ Visualizes speciation throughout evolution. """
-    csvfile = open(filename, 'wb')
-    statWriter = csv.writer(csvfile, delimiter=' ')
-    num_generations = len(species_fitness_log)
-    num_species = max(map(len, species_fitness_log))
-    curves = []
-    for gen in species_fitness_log:
-        species = ["NA"] * num_species
-        species[:len(gen)] = gen
-        statWriter.writerow(species)
 
-    csvfile.close()
+def save_species_fitness(population, delimiter=' ', null_value='NA', filename='species_fitness.csv'):
+    """ Log species' average fitness throughout evolution. """
+    with open(filename, 'w') as f:
+        w = csv.writer(f, delimiter=delimiter)
+        for s in get_species_fitness(population, null_value):
+            w.writerow(s)

neat/visualize.py

@@ -1,7 +1,10 @@
 # -*- coding: UTF-8 -*-
 from __future__ import print_function
+
 import warnings
 
+from neat.statistics import get_average_fitness, get_species_sizes
+
 try:
     import graphviz
 except ImportError:
@@ -20,23 +23,19 @@
     np = None
     warnings.warn('Could not import optional dependency NumPy.')
 
-from neat.math_util import mean
-
 
-def plot_stats(best_genomes, fitness_scores, ylog=False, view=False, filename='avg_fitness.svg'):
+def plot_stats(population, ylog=False, view=False, filename='avg_fitness.svg'):
     """ Plots the population's average and best fitness. """
     if plt is None:
         warnings.warn("This display is not available due to a missing optional dependency (matplotlib)")
         return
 
-    generation = range(len(best_genomes))
-
-    fitness = [c.fitness for c in best_genomes]
-
-    avg_scores = [mean(f) for f in fitness_scores]
+    generation = range(len(population.most_fit_genomes))
+    best_fitness = [c.fitness for c in population.most_fit_genomes]
+    avg_fitness = get_average_fitness(population)
 
-    plt.plot(generation, avg_scores, 'b-', label="average")
-    plt.plot(generation, fitness, 'r-', label="best")
+    plt.plot(generation, avg_fitness, 'b-', label="average")
+    plt.plot(generation, best_fitness, 'r-', label="best")
 
     plt.title("Population's average and best fitness")
     plt.xlabel("Generations")
@@ -95,24 +94,15 @@ def plot_spikes(spikes, view=False, filename=None, title=None):
     plt.close()
 
 
-def plot_species(species_log, view=False, filename='speciation.svg'):
+def plot_species(population, view=False, filename='speciation.svg'):
     """ Visualizes speciation throughout evolution. """
     if plt is None:
         warnings.warn("This display is not available due to a missing optional dependency (matplotlib)")
         return
 
-    num_generations = len(species_log)
-    all_species = set()
-    for gen_data in species_log:
-        for sid, scount in gen_data.items():
-            all_species.add(sid)
-
-    max_species = max(all_species)
-    curves = []
-    for gen_data in species_log:
-        species = [gen_data.get(sid, 0) for sid in range(max_species + 1)]
-        curves.append(np.array(species))
-    curves = np.array(curves).T
+    species_sizes = get_species_sizes(population)
+    num_generations = len(species_sizes)
+    curves = np.array(species_sizes).T
 
     fig, ax = plt.subplots()
     ax.stackplot(range(num_generations), *curves)