Refactored by Sourcery

consts.py

@@ -50,10 +50,7 @@
 DISK_FOUR_COUNTS = np.zeros([HEIGHT, WIDTH], int)
 for row in range(HEIGHT):
   for column in range(WIDTH):
-    disk_fours = []
-    for four in FOURS:
-      if four[row, column]:
-        disk_fours.append(four)
+    disk_fours = [four for four in FOURS if four[row, column]]
     DISK_FOURS[row, column] = disk_fours
     DISK_FOUR_COUNTS[row, column] = len(disk_fours)
 
@@ -88,4 +85,4 @@
   print(ROW_EDGE_DISTANCE.reshape([HEIGHT, WIDTH]))
   print(COLUMN_EDGE_DISTANCE.reshape([HEIGHT, WIDTH]))
   print(ODDS.reshape([HEIGHT, WIDTH]))
-  print(np.array(map(bin, DISK_HASHES.flatten())).reshape(DISK_HASHES.shape))
+  print(np.array(map(bin, DISK_HASHES.flatten())).reshape(DISK_HASHES.shape))

network.py

@@ -106,10 +106,8 @@ def variables(self):
                              self.scope + '/')
 
   def assign(self, other):
-    copy_ops = []
-    for self_var, other_var in zip(self.variables, other.variables):
-      copy_ops.append(tf.assign(other_var, self_var))
-    return copy_ops
+    return [tf.assign(other_var, self_var) for self_var, other_var in zip(self.variables, other.variables)]
+
 
 
 class PolicyNetwork(BaseNetwork):

policy_training.py

@@ -141,9 +141,7 @@ def train_games(self, opponent, games):
 
   def process_results(self, opponent, games, step, summary):
     win_rate = np.mean([game.policy_player_score for game in games])
-    average_moves = sum([len(game.moves)
-                         for game in games]) / self.config.batch_size
-
+    average_moves = sum(len(game.moves) for game in games) / self.config.batch_size
     opponent_summary = tf.Summary()
     opponent_summary.value.add(
         tag=self.training_scope.name + '/' + opponent.name + '/win_rate',
@@ -263,10 +261,7 @@ def move(self, move, policy_player_turn=False):
     self.positions.append(self.position)
     if self.position.gameover():
       self.result = self.position.result
-      if self.result:
-        self.policy_player_score = float(policy_player_turn)
-      else:
-        self.policy_player_score = 0.5
+      self.policy_player_score = float(policy_player_turn) if self.result else 0.5
 
 
 def main(_):