Added code for example assignment results

assignment_plots.py

+import pickle
+import numpy as np
+import matplotlib
+import matplotlib.pyplot as plt
+
+histories = pickle.load(open('assignment_simulation_histories.p', 'r'))
+
+fig, ax = plt.subplots()
+for idx, history in enumerate(histories):
+    x = [gen.generation_number for gen in history]
+    y = [gen.max_fitness  for gen in history]
+    ax.plot(x, y, label = str(idx))
+ax.set(xlabel='Generation #', ylabel='max fitness', title='Three Simulations')
+ax.yaxis.grid(True)
+fig.legend()
+plt.savefig('assignment_fitness_vs_gen.png')
+plt.close()
+
+fig, ax = plt.subplots()
+styles = {'surviving_ancestors': 'b-',
+          'surviving_children': 'r--',
+          'surviving_mutants': 'k-.'}
+
+labels = {'surviving_ancestors': 'previous generation',
+          'surviving_children': 'new children',
+          'surviving_mutants': 'new mutants'}
+for attr in ['surviving_ancestors', 'surviving_children', 'surviving_mutants']:
+    
+    style = styles[attr]
+    ys = []
+    for idx, history in enumerate(histories):
+        history = filter(lambda gen: getattr(gen, attr) is not None, history)
+        ys.append([getattr(gen, attr)  for gen in history])
+    	x = [gen.generation_number for gen in history]
+    y = np.mean(ys, axis=0)
+    ax.plot(x, y, style, label = labels[attr])
+ax.set(xlabel='Generation #', ylabel='Number per Group', title='Types of Surviving Manducas')
+ax.yaxis.grid(True)
+fig.legend()
+plt.savefig('assignment_survivors_vs_gen.png')
+plt.close()
+
+fig, ax = plt.subplots()
+styles = {'duplicate_population': 'b-',
+          'duplicate_children': 'r--',
+          'duplicate_mutants': 'k-.'}
+
+labels = {'duplicate_population': 'previous generation',
+          'duplicate_children': 'new children',
+          'duplicate_mutants': 'new mutants'}
+for attr in ['duplicate_population', 'duplicate_children', 'duplicate_mutants']:
+    
+    style = styles[attr]
+    ys = []
+    for idx, history in enumerate(histories):
+        history = filter(lambda gen: getattr(gen, attr) is not None, history)
+        ys.append([getattr(gen, attr)  for gen in history])
+    	x = [gen.generation_number for gen in history]
+    y = np.mean(ys, axis=0)
+    ax.plot(x, y, style, label = labels[attr])
+ax.set(xlabel='Generation #', ylabel='Number per Group', title='Types of Duplicate Manducas')
+ax.yaxis.grid(True)
+fig.legend()
+plt.savefig('assignment_dups_vs_gen.png')
+plt.close()

assignment_sim.py

+""" This is a script that generates the basic functionality of the
+EE194_Manduca_Simulator package with multithreading support"""
+import numpy as np
+import pickle
+import multiprocessing
+import tqdm
+from manduca import SimpleManduca, EvolutionSimulator, EvolutionParameters
+
+POPULATION_SIZE, NUM_GENERATIONS= 100, 200
+evolution_parameters = EvolutionParameters(POPULATION_SIZE, 50, 50, 10)
+num_legs, time_segments, time_step = 5, 10, 10
+
+rng = np.random.RandomState()
+rng.seed(0xBAAA)
+def random_manduca():
+    return SimpleManduca.random_individual(num_legs, time_segments, time_step, rng=rng)
+
+histories = []
+try:
+    pool = multiprocessing.Pool(int(multiprocessing.cpu_count()))
+    for idx,RND_SEED in enumerate(tqdm.tqdm([0xBADC0DE, 0xDEADBEEF, 0xFACEB00C],
+                                  desc='70 year simulations', unit='sims')):
+        simulator = EvolutionSimulator(random_manduca, pool=pool,
+                                       evolution_parameters = evolution_parameters,
+                                       random_number_generator=rng, random_seed=RND_SEED)
+        simulator.run_simulation(NUM_GENERATIONS)
+        histories.append(simulator.history)
+        simulator.update_fitness()
+        simulator.population[0].save('fittest_manduca_{}.npz'.format(idx))
+finally:
+    pool.close()                                                                                                                                                                                                                     
+    pool.join() 
+
+pickle.dump(histories, open('assignment_simulation_histories.p', 'wb'))