Edited docstrings and added some comments

gastop/genalg.py

@@ -161,25 +161,27 @@ def run(self, num_generations=None, progress_fitness=None, progress_truss=None,
         if progress_truss is None:
             progress_truss = self.config['monitor_params']['progress_truss']
 
-        # initialize progress monitor object
+        # Initialize progress monitor object
         progress = ProgMon(progress_fitness, progress_truss, num_generations, domain=self.random_params['domain'],
                            loads=self.config['evaluator_params']['boundary_conditions']['loads'],
                            fixtures=self.config['evaluator_params']['boundary_conditions']['fixtures'])
 
+        # Set chunksize used in multithreading based on size of population
         if self.ga_params['pop_size'] < 1e4:
             chunksize = int(np.amax((self.ga_params['pop_size']/100, 1)))
         else:
             chunksize = int(np.sqrt(self.ga_params['pop_size']))
-        # Loop over all generations:
 
+        # Loop over all generations, updating progress bar with tqdm:
         for current_gen in tqdm(range(num_generations), desc='Overall', position=0):
             self.ga_params['current_generation'] = current_gen
-            # no parallelization
+            # No parallelization
             if num_threads == 1:
                 for current_truss in tqdm(self.population, desc='Evaluating', position=1):
                     self.evaluator(current_truss)
                 for current_truss in tqdm(self.population, desc='Scoring', position=1):
                     self.fitness_function(current_truss)
+            # With multithreading
             else:
                 with Pool(num_threads) as pool:
                     self.population = list(tqdm(pool.imap(
@@ -189,12 +191,16 @@ def run(self, num_generations=None, progress_fitness=None, progress_truss=None,
                         self.fitness_function, self.population, chunksize),
                         total=self.ga_params['pop_size'], desc='Scoring', position=1))
 
+            # Sort population by fitness score (lowest score = most fit)
             self.population.sort(key=lambda x: x.fitness_score)
 
+            # Update progress monitor plots
             progress.progress_monitor(current_gen, self.population)
 
+            # Create next generation
             self.update_population()
 
+            # Periodically save config and population to JSON files
             if self.ga_params['save_frequency'] != 0 and (current_gen % self.ga_params['save_frequency']) == 0:
                 self.save_state(
                     dest_config=self.ga_params['config_save_name'], dest_pop=self.ga_params['pop_save_name'])

gastop/selector.py

@@ -76,14 +76,13 @@ def inverse_square_rank_probability(num_parents, population):
     def tournament(num_parents, population, tourn_size, tourn_prob):
         ''' Selects parents according to tournament method.
 
-        (ASSUMES POPULATION IS SORTED BY FITNESS) Randomly selects truss
-        indices from population in groups called tournaments according to
-        "tourn_size." Each tournament is then sorted by index (lower means
-        more fit) in ascending order and a single index from each tournament
-        is selected. The selection from each tournament is chosen
-        probabilistically, assigning the first, most fit, index with probability
-        p = tourn_prob, and then subsequent indices by p*(1-p)^n. The winners
-        of each tournament are then returned as the parents array.
+        Randomly selects truss indices from population in groups called
+        tournaments according to "tourn_size." Each tournament is then sorted
+        by index (lower means more fit) in ascending order and a single index
+        from each tournament is selected. The selection from each tournament is
+        chosen probabilistically, assigning the first, most fit, index with
+        probability p = tourn_prob, and then subsequent indices by p*(1-p)^n.
+        The winners of each tournament are then returned as the parents array.
 
         Args:
             num_parents (int): The number of parents to select.

gastop/utilities.py

@@ -123,7 +123,8 @@ def init_file_parser(init_file_path):  # Cristian
 
     Creates ConfigObj object, which reads input parameters as a nested
     dictionary of strings. The string are then converted to their correct types
-    using the ConfigObj walk method and a transform function.
+    using the ConfigObj walk method and a transform function. Defaults are then
+    set with if statements.
 
     Args:
         init_file_path (string): Path to the init file, relative to
@@ -178,7 +179,10 @@ def transform(section, key):
     # transform function.
     config.walk(transform)
 
+    # Parse 'properties' CSV file
     properties_dict = beam_file_parser(config['general']['properties_path'])
+
+    # ---------------------------Set Defaults---------------------------------
     user_spec_nodes = config['general']['user_spec_nodes']
     num_user_nodes = user_spec_nodes.shape[0]
     num_rand_nodes = config['general']['num_rand_nodes']
@@ -188,9 +192,6 @@ def transform(section, key):
     loads = config['general']['loads']
     fixtures = config['general']['fixtures']
 
-    # progress_fitness = config['monitor_params']['progress_fitness']  # sfr
-    # progress_truss = config['monitor_params']['progress_truss']  # sfr
-
     if loads.ndim < 3:
         loads = np.reshape(loads, (loads.shape + (1,)))
     if fixtures.ndim < 3:
@@ -299,7 +300,7 @@ def save_progress_history(progress_history, path_progress_history='progress_hist
     Returns:
         Nothing
     '''
-    # Save pop_progress data
+    # Save progress_history data
     with open(path_progress_history, 'w') as f:
         progress_history_dumped = json.dumps(
             copy.deepcopy(progress_history), cls=encoders.PopulationEncoder)
@@ -316,7 +317,7 @@ def load_progress_history(path_progress_history='progress_history.json'):
         progress_history (dict): History of each generation, including generation
         number, fittest truss, etc.
     '''
-    # Load pop_progress data
+    # Load progress_history data
     with open(path_progress_history, 'r') as f:
         progress_history_loaded = json.load(f)
     progress_history = json.loads(

tests/test_selector.py

@@ -15,8 +15,11 @@
 
 
 class TestSelector(unittest.TestCase):  # Cristian
+    '''Tests for Selector() method_params
+    '''
     def testInvSqrRankProp(self):
-        # Tests the inverse_square_rank_probability method of Selector()
+        '''Tests the inverse_square_rank_probability method of Selector()
+        '''
         nodes = np.array([[1, 2, 3], [2, 3, 4]])
         edges = np.array([[0, 1]])
         properties = np.array([[0, 3]])
@@ -59,7 +62,8 @@ def testInvSqrRankProp(self):
         self.assertEqual(num_parents, len(parents))
 
     def testTournament(self):
-        # Tests the tournament method of Selector()
+        '''Tests the tournament method of Selector()
+        '''
         nodes = np.array([[1, 2, 3], [2, 3, 4]])
         edges = np.array([[0, 1]])
         properties = np.array([[0, 3]])

tests/test_utilities.py

@@ -20,9 +20,7 @@ def testInitFileParser(self):
 
         init_file_path = 'gastop-config/boolean_parse_test_init.txt'
         config = utilities.init_file_parser(init_file_path)
-        # for key in config:
-        #     print(key)
-        #     print(config[key])
+
         self.assertTrue(isinstance(config['ga_params']['num_elite'], int))
         self.assertTrue(
             isinstance(config['ga_params']['percent_crossover'], float))
@@ -40,7 +38,7 @@ def testInitFileParser(self):
         self.assertEqual(config['ga_params']['save_frequency'], 0)
 
     def testInitFileParser2(self):
-        """tests edge cases for invalid file path"""
+        """Tests edge cases for invalid file path"""
 
         init_file_path = 'gastop-config/foo'
         self.assertRaises(IOError, utilities.init_file_parser, init_file_path)