def testState(self):
# First "real life" example
g = TicTacToe.Game()
self.assertEqual(g.status, -1)
g.play((0,0)); g.play((0,1))
self.assertEqual(g.status, -1)
g.play((1,1)); g.play((1,0))
self.assertEqual(g.status, -1)
g.play((2,2));
self.assertEqual(g.status, 1)
with self.assertRaises(RuntimeError) as cm:
g.play((2,1))
# Some more "to the point" tests
g = TicTacToe.Game()
g._board[0, 0:3] = 1
self.assertEqual(g._compute_status(), 1)
self.assertEqual(g.status, 1)
g = TicTacToe.Game()
g._board[0:3, 2] = 1
self.assertEqual(g._compute_status(), 1)
self.assertEqual(g.status, 1)
g = TicTacToe.Game()
g._board[2, 0] = 2
self.assertEqual(g._compute_status(), -1)
self.assertEqual(g.status, -1)
g._board[1, 1] = 2
self.assertEqual(g._compute_status(), -1)
self.assertEqual(g.status, -1)
g._board[0, 2] = 2
self.assertEqual(g._compute_status(), 2)
self.assertEqual(g.status, 2)
def main():
import TicTacToe as tm
import TicTacToe.bots as tm_bots
bot_minmax = tm_bots.Minimax_Bot("x", "o", "Bob")
bot_random = tm_bots.Random_Bot("o", "x", "Bill")
game = tm.Game([bot_minmax, bot_random])
game.run()
def getWLRs(self, number_of_runs=1000):
env = TicTacToe.Game(self.rewards)
results = [0, 0, 0]
for episode in range(number_of_runs):
# Reset the environment
env.reset()
# run game
steps = env.run(self.qtable)
# get reward from last step
reward = steps[-1][3]
for i in range(3):
if reward == self.rewards[i]:
results[i] += 1
return results
def initLearning(self):
# init Game Environment
self.environment = TicTacToe.Game(self.rewards)
def twoplayer_setup(self):
game1 = TicTacToe.Game()
player1 = TicTacToe.Player('X', game1)
player2 = TicTacToe.Player('O', game1)
return (game1, player1, player2)
def onecompplayer_setup(self):
game1 = TicTacToe.Game()
computer1 = TicTacToe.Computer('X', game1)
player2 = TicTacToe.Player('O', game1)
return (game1, computer1, player2)
import TicTacToe
from utility import loadFromFile
qtable = loadFromFile("qtable")
game = TicTacToe.Game((10, -10, 5))
game.showcase(qtable)