def test_do_action(self):
# X - O
# - - X
# - - O
game = TicTacToe().make_moves([1, 3, 6, 9])
mdp = FixedGameMDP(game.copy(), AlphaBeta(), 1)
env = Environment(mdp)
env.do_action(7)
expected = TicTacToe().make_moves([1, 3, 6, 9, 7, 4])
self.assertEqual(env.cur_state(), expected)
def test_reward_when_game_is_over_and_is_draw(self):
cur_state = TicTacToe().make_moves([1, 3, 2, 4, 6, 5, 7, 8])
# cur_state:
# X X O
# O O X
# X O -
action = 9
next_state = cur_state.copy().make_move(action)
# next_state:
# X X O
# O O X
# X O X
reward = self.mdp.reward(cur_state, action, next_state)
self.assertEqual(reward, 0.0)
def test_transitions(self):
# X - O
# - - X
# - - O
game = TicTacToe().make_moves([1, 3, 6, 9])
# Create an mdp where AlphaBeta is the
# second player to move
mdp = FixedGameMDP(game, AlphaBeta(), 1)
# Put X in 7 slot, and the mdp should
# make the transition assuming O moving 4
transitions = mdp.transitions(game, 7)
self.assertEqual(len(transitions), 1)
next_game, prob = transitions[0]
self.assertEqual(next_game, game.copy().make_moves([7, 4]))
self.assertEqual(prob, 1.0)
def test_reward_when_agent_moves_second_and_losses(self):
ab = AlphaBeta()
opp_idx = 0
mdp = FixedGameMDP(self.game, ab, opp_idx)
cur_state = TicTacToe().make_moves([1, 4, 2, 5])
# cur_state:
# X X -
# O O -
# - - -
action = 3
next_state = cur_state.copy().make_move(action)
# next_state:
# X X X
# O O -
# - - -
reward = mdp.reward(cur_state, action, next_state)
self.assertEqual(reward, -1.0)
def test_reward_when_agent_moves_second_and_wins(self):
ab = AlphaBeta()
opp_idx = 0
mdp = FixedGameMDP(self.game, ab, opp_idx)
# Opponent moves first, Agent second
cur_state = TicTacToe().make_moves([1, 4, 2, 5, 7])
# cur_state:
# X X -
# O O -
# X - -
action = 6
next_state = cur_state.copy().make_move(action)
# next_state:
# X X -
# O O O
# X - -
reward = mdp.reward(cur_state, action, next_state)
self.assertEqual(reward, 1.0)
class TestTicTacToe(unittest.TestCase):
def setUp(self):
self.game = TicTacToe()
def test_name(self):
self.assertEqual(TicTacToe.name, 'Tic-Tac-Toe')
def test_with_board(self):
game = TicTacToe([['X', ' ', 'X'], ['O', ' ', 'X'], [' ', 'O', ' ']])
self.assertEqual(game.legal_moves(), [2, 5, 7, 9])
self.assertEqual(game.cur_player(), 1)
def test_copy(self):
self.game.make_moves([1, 2])
game_copy = self.game.copy()
self.assertEqual(game_copy, self.game)
self.assertIsNot(game_copy, self.game)
self.assertIsInstance(game_copy, TicTacToe)
self.assertIsInstance(game_copy, TicTacToe)
self.assertEqual(game_copy.legal_moves(), self.game.legal_moves())
self.assertIsNot(game_copy.legal_moves(), self.game.legal_moves())
def test_hash(self):
game1 = TicTacToe().make_moves([1, 7])
table = {game1: 'game1'}
self.assertEqual(len(table), 1)
self.assertEqual(table[game1], 'game1')
game2 = TicTacToe().make_moves([1, 7])
table[game2] = 'game2'
self.assertEqual(table[game1], 'game2')
self.assertEqual(table[game2], 'game2')
game3 = TicTacToe().make_moves([1, 7, 9])
table[game3] = 'game3'
self.assertEqual(len(table), 2)
self.assertEqual(table[game3], 'game3')
def test_not_equal(self):
other = TicTacToe().make_moves([1])
self.assertNotEqual(self.game, other)
def test_cur_player_start(self):
self.assertEqual(self.game.cur_player(), 0)
def test_cur_player_after_one_move(self):
self.game.make_move(1)
self.assertEqual(self.game.cur_player(), 1)
def test_cur_player_after_two_moves(self):
self.game.make_moves([1, 2])
self.assertEqual(self.game.cur_player(), 0)
def test_equal(self):
moves = [1, 2]
self.game.make_moves(moves)
other = TicTacToe().make_moves(moves)
self.assertEqual(self.game, other)
def test_is_not_over_at_start(self):
self.assertFalse(self.game.is_over())
def test_is_over_at_end_of_game(self):
self.game.make_moves([1, 4, 2, 5, 3])
self.assertTrue(self.game.is_over())
def test_legal_moves_start(self):
actual = self.game.legal_moves()
expected = [1, 2, 3, 4, 5, 6, 7, 8, 9]
self.assertEqual(actual, expected)
def test_legal_moves_after_one_move(self):
self.game.make_move(1)
actual = self.game.legal_moves()
expected = [2, 3, 4, 5, 6, 7, 8, 9]
self.assertEqual(actual, expected)
def test_legal_moves_after_two_moves(self):
self.game.make_moves([3, 7])
actual = self.game.legal_moves()
expected = [1, 2, 4, 5, 6, 8, 9]
self.assertEqual(actual, expected)
def test_legal_moves_are_empty_when_is_over(self):
game = TicTacToe([['X', 'X', 'X'], ['O', 'O', ' '], [' ', ' ', ' ']])
self.assertTrue(len(game.legal_moves()) == 0)
def test_make_move_returns_self(self):
self.assertIs(self.game.make_move(1), self.game)
def test_make_moves(self):
self.game.make_moves([1, 2, 3])
actual = self.game.legal_moves()
expected = [4, 5, 6, 7, 8, 9]
self.assertEqual(actual, expected)
def test_make_moves_returns_self(self):
self.assertIs(self.game.make_moves([1, 2, 3]), self.game)
def test_make_moves_returns_self(self):
actual = self.game.make_moves([1, 2, 3])
expected = self.game
self.assertEqual(actual, expected)
def test_outcomes_win_first_player(self):
game = TicTacToe([['X', 'X', 'X'], ['O', 'O', ' '], [' ', ' ', ' ']])
self.assertEqual(game.outcomes(), ['W', 'L'])
def test_outcomes_win_second_player(self):
game = TicTacToe([['X', 'X', ' '], ['O', 'O', 'O'], [' ', ' ', 'X']])
self.assertEqual(game.outcomes(), ['L', 'W'])
def test_outcomes_draw(self):
game = TicTacToe([['X', 'X', 'O'], ['O', 'O', 'X'], ['X', 'O', 'X']])
self.assertEqual(game.outcomes(), ['D', 'D'])