Update game logic and error handling in spiel.py

games/spiel.py

@@ -1,8 +1,10 @@
 import datetime
 import pathlib
+from typing import Tuple
 
 import numpy
 import torch
+from pyspiel import Game, SpielError
 
 from .abstract_game import AbstractGame
 
@@ -19,8 +21,23 @@
         "You need to install open_spiel by running pip install open_spiel. For a full documentation, see: https://github.com/deepmind/open_spiel/blob/master/docs/install.md"
     )
 
+
+def get_observation_tensor_shape(game: Game) -> Tuple:
+    # Dimensions of the game observation, must be 3D (channel, height, width).
+    # We reshape it for a 1D array to (1, 1, length of array) or return a default shape if it's not implemented in spiel game.
+    try:
+        shape = game.observation_tensor_shape()
+    except SpielError:
+        print('ObservationTensorShape unimplemented. Returning default tensor shape (1,1,1)')
+        return 1, 1, 1
+
+    for _ in range(3 - len(shape)):
+        shape.insert(0, 1)
+    return tuple(shape)
+
+
 # The game you want to run. See https://github.com/deepmind/open_spiel/blob/master/docs/games.md for a list of games
-game = pyspiel.load_game("tic_tac_toe")
+game = pyspiel.load_game("backgammon")
 
 
 class MuZeroConfig:
@@ -34,9 +51,8 @@ def __init__(self):
         self.max_num_gpus = None  # Fix the maximum number of GPUs to use. It's usually faster to use a single GPU (set it to 1) if it has enough memory. None will use every GPUs available
 
 
-
         ### Game
-        self.observation_shape =  tuple(self.game.observation_tensor_shape()) # Dimensions of the game observation, must be 3D (channel, height, width). For a 1D array, please reshape it to (1, 1, length of array)
+        self.observation_shape = get_observation_tensor_shape(self.game)
         self.action_space = list(range(self.game.policy_tensor_shape()[0]))  # Fixed list of all possible actions. You should only edit the length
         self.players = list(range(self.game.num_players()))  # List of players. You should only edit the length
         self.stacked_observations = 0  # Number of previous observations and previous actions to add to the current observation
@@ -270,14 +286,17 @@ def get_observation(self):
         else:
             current_player = 0
         return numpy.array(self.board.observation_tensor(current_player)).reshape(
-            self.game.observation_tensor_shape()
+            get_observation_tensor_shape(self.game)
         )
 
     def legal_actions(self):
         return self.board.legal_actions()
 
     def have_winner(self):
-        rewards = self.board.rewards()
+        try:
+            rewards = self.board.rewards()
+        except SpielError:
+            return False
 
         if self.player == 1: