autopep8 -r . -i --max-line-length 1000 --ignore E309

AlphaGo/ai.py

@@ -87,7 +87,7 @@ def get_moves(self, states):
 class MCTSPlayer(object):
     def __init__(self, value_function, policy_function, rollout_function, lmbda=.5, c_puct=5, rollout_limit=500, playout_depth=40, n_playout=100):
         self.mcts = mcts.MCTS(value_function, policy_function, rollout_function, lmbda, c_puct,
-                    rollout_limit, playout_depth, n_playout)
+                              rollout_limit, playout_depth, n_playout)
 
     def get_move(self, state):
         sensible_moves = [move for move in state.get_legal_moves(include_eyes=False)]

AlphaGo/go.py

@@ -427,7 +427,7 @@ def do_move(self, action, color=None):
         # Check for end of game
         if len(self.history) > 1:
             if self.history[-1] is PASS_MOVE and self.history[-2] is PASS_MOVE \
-                and self.current_player == WHITE:
+                    and self.current_player == WHITE:
                 self.is_end_of_game = True
         return self.is_end_of_game
 

AlphaGo/mcts.py

@@ -11,6 +11,7 @@ class TreeNode(object):
     """A node in the MCTS tree. Each node keeps track of its own value Q, prior probability P, and
     its visit-count-adjusted prior score u.
     """
+
     def __init__(self, parent, prior_p):
         self._parent = parent
         self._children = {}  # a map from action to TreeNode

AlphaGo/models/nn_util.py

@@ -118,6 +118,7 @@ class Bias(Layer):
     Largely copied from the keras docs:
     http://keras.io/layers/writing-your-own-keras-layers/#writing-your-own-keras-layers
     """
+
     def __init__(self, **kwargs):
         super(Bias, self).__init__(**kwargs)
 

AlphaGo/preprocessing/preprocessing.py

@@ -2,7 +2,7 @@
 import AlphaGo.go as go
 
 ##
-## individual feature functions (state --> tensor) begin here
+# individual feature functions (state --> tensor) begin here
 ##
 
 

AlphaGo/training/reinforcement_policy_trainer.py

@@ -24,6 +24,7 @@ class BatchedReinforcementLearningSGD(Optimizer):
         lr: float >= 0. Learning rate.
         ng: int > 0. Number of games played in parallel. Each one has its own cumulative gradient.
     '''
+
     def __init__(self, lr=0.01, ng=20, **kwargs):
         super(BatchedReinforcementLearningSGD, self).__init__(**kwargs)
         self.__dict__.update(locals())

interface/Play.py

@@ -4,6 +4,7 @@
 
 class play_match(object):
     """Interface to handle play between two players."""
+
     def __init__(self, player1, player2, save_dir=None, size=19):
         # super(ClassName, self).__init__()
         self.player1 = player1
@@ -19,7 +20,7 @@ def _play(self, player):
         # self.state.write_to_disk()
         if len(self.state.history) > 1:
             if self.state.history[-1] is None and self.state.history[-2] is None \
-                and self.state.current_player == -1:
+                    and self.state.current_player == -1:
                 end_of_game = True
             else:
                 end_of_game = False