cleaning: merged gameUtils.py and game.py

sirmammingtonham · Nov 1, 2018 · 06e633b · 06e633b
1 parent 4714d56
commit 06e633b
Show file tree

Hide file tree

Showing 6 changed files with 683 additions and 69 deletions.
diff --git a/alphabot/Arena.py b/alphabot/Arena.py
@@ -48,15 +48,15 @@ def playGame(self, verbose=False):
                 action = players[curPlayer + 1](current_game)
                 next_state, curPlayer = self.game.getNextState(curPlayer, (action), current_game)
             else:
-                pi = players[curPlayer+1](self.game.getState(current_game))
+                pi = players[curPlayer+1](self.game.getState())
                 pi_reshape = np.reshape(pi, (21, 18))
                 action = np.where(pi_reshape==np.max(pi_reshape))
                 next_state, curPlayer = self.game.getNextState(curPlayer, (action[0][0], action[1][0]), current_game)
         # if verbose:
         #     assert(self.display)
         #     print("Game over: Turn ", str(it), "Result ", str(self.game.getGameEnded(board, 1)))
         #     self.display(board)
-        return self.game.getGameEnded(curPlayer, current_game)
+        return self.game.getGameEnded(current_game)
 
     def playGames(self, num, verbose=False):
         """

diff --git a/alphabot/Coach.py b/alphabot/Coach.py
@@ -51,15 +51,15 @@ def executeEpisode(self):
             pi = self.mcts.getActionProb(state, temp=temp)
             pi_reshape = np.reshape(pi, (21, 18))
             # sym = self.game.getSymmetries(state, pi)
-            s = self.game.getState(self.curPlayer)
+            s = self.game.getState()
             trainExamples.append([s, self.curPlayer, pi, None])
             # for b,p in sym:
             #     trainExamples.append([b, self.curPlayer, p, None])
             action = np.random.choice(len(pi), p=pi)
             a, b = np.unravel_index(np.ravel(action, np.asarray(pi).shape), pi_reshape.shape)
             current_game, self.curPlayer = self.game.getNextState(self.curPlayer, (a[0], b[0]))
 
-            r = self.game.getGameEnded(self.curPlayer)
+            r = self.game.getGameEnded()
 
             if r!=0:
                 return [(x[0],x[2],r*((-1)**(x[1]!=self.curPlayer))) for x in trainExamples]