def test_4x4_max_number_for_get_state_number(self):
env = Environment(4, 4)
env.board.fill(2)
state_number = env.get_state_number(env.board)
self.assertEqual(3 ** 16 - 1, state_number)
def test_3x3_max_number_for_get_state_number(self):
env = Environment(3, 3)
env.board.fill(2)
state_number = env.get_state_number(env.board)
self.assertEqual(3 ** 9 - 1, state_number)
def test_game_ended_when_not_ended(self):
env = Environment(2, 2)
env.print_array(env.board)
game_ended = env.check_game_ended()
self.assertEqual(False, game_ended)
def test__game_ended_row_check_when_game_not_ended_board_empty(self):
env = Environment(3, 3)
if_game_ended = env._game_ended_row_check()
env.print_array(env.board)
self.assertEqual(False, if_game_ended)
def test__draw_check_when_no_draw(self):
env = Environment(3, 3)
env.set_x(0, 0)
env.set_x(1, 0)
env.set_o(2, 0)
env.print_array(env.board)
env._draw_check()
self.assertEqual(False, env.draw)
def test_get_board_from_state_number_80(self):
# given
env = Environment(2, 2)
state_number = 80
# when - conversion from state number to board and back to number
board_from_state = env.get_board_from_state_number(state_number, True)
state_number_from_board = env.get_state_number(board_from_state)
# then
self.assertEqual(state_number_from_board, state_number)
def test_get_state_number(self):
env = Environment(3, 3)
env.set_x(0, 0) # 1
env.set_x(0, 1) # 3
env.set_x(0, 2) # 9
env.set_o(1, 1) # 3^4 * 2 = 162
state_number = env.get_state_number(env.board)
self.assertEqual(175, state_number)
def test__game_ended_row_check_when_game_not_ended(self):
env = Environment(3, 3)
env.set_o(0, 0)
env.set_o(0, 1)
env.set_x(0, 2)
if_game_ended = env._game_ended_row_check()
env.print_array(env.board)
self.assertEqual(False, if_game_ended)
def test2__game_ended_diagonal_check_on_12x12_board(self):
env = Environment(12, 12)
env.set_o(4, 4)
env.set_o(5, 5)
env.set_o(6, 6)
env.print_array(env.board)
if_game_ended = env._game_ended_diagonal_check()
self.assertEqual(True, if_game_ended)
def test__game_ended_diagonal_check_on_12x12_board(self):
env = Environment(12, 12)
env.set_o(9, 2)
env.set_o(10, 1)
env.set_o(11, 0)
env.print_array(env.board)
if_game_ended = env._game_ended_diagonal_check()
self.assertEqual(True, if_game_ended)
def test__game_ended_diagonal_check_when_not_ended_on_4x4_board(self):
env = Environment(4, 4)
env.set_o(0, 0)
env.set_o(2, 2)
env.set_o(3, 3)
env.print_array(env.board)
if_game_ended = env._game_ended_diagonal_check()
self.assertEqual(False, if_game_ended)
def test2__game_ended_diagonal_check_when_ended_right_to_left_on_4x4_board(self):
env = Environment(4, 4)
env.set_o(1, 2)
env.set_o(2, 1)
env.set_o(3, 0)
env.print_array(env.board)
if_game_ended = env._game_ended_diagonal_check()
self.assertEqual(True, if_game_ended)
def test__game_ended_diagonal_check_when_ended_left_to_right(self):
env = Environment(3, 3)
env.set_x(0, 0)
env.set_x(1, 1)
env.set_x(2, 2)
env.print_array(env.board)
if_game_ended = env._game_ended_diagonal_check()
self.assertEqual(True, if_game_ended)
def test__game_ended_column_check_when_game_not_ended_4x3_board(self):
env = Environment(4, 3)
env.set_o(0, 1)
env.set_o(2, 1)
env.set_o(3, 1)
env.print_array(env.board)
if_game_ended = env._game_ended_column_check()
self.assertEqual(False, if_game_ended)
def test_game_ended_when_x_has_won_(self):
env = Environment(3, 3)
env.set_x(0, 0)
env.set_x(1, 0)
env.set_x(2, 0)
env.print_array(env.board)
game_ended = env.check_game_ended(force_recalculate=True)
self.assertEqual(True, game_ended)
def test__game_ended_row_check_when_game_not_ended_on4x4_board(self):
env = Environment(4, 4)
env.set_o(2, 0)
env.set_o(2, 2)
env.set_o(2, 3)
if_game_ended = env._game_ended_row_check()
env.print_array(env.board)
self.assertEqual(False, if_game_ended)
def test__draw_check_when_draw(self):
env = Environment(3, 3)
env.board.fill(1)
env.set_o(1, 1)
env.print_array(env.board)
env._draw_check()
self.assertEqual(True, env.draw)
def test__game_ended_column_check_when_game_ended(self):
env = Environment(3, 3)
env.set_x(0, 0)
env.set_x(1, 0)
env.set_x(2, 0)
if_game_ended = env._game_ended_column_check()
self.assertEqual(True, if_game_ended)
def test_rows_columns_number(self):
# given
env = Environment(4, 3)
# when
env.set_x(0, 0) # x denoted as 1 in numpy array
env.set_o(0, 1) # o denoted as 2 in numpy array
# then
self.assertEqual(env.rows, 4)
self.assertEqual(env.columns, 3)
self.assertEqual(env.board[0, 0], 1)
self.assertEqual(env.board[0, 1], 2)
if env.winner == p1.player_number and isinstance(p1, Human) or env.winner == p2.player_number and isinstance(p2, Human):
print('Ghrrr! Watch your back human! Next time I will triumph! \n')
if draw:
env.print_array(env.board)
# update the value function
p1.update_state_values(env)
p2.update_state_values(env)
# finally
env.clear_board()
# ----------------------------------------------------
env = Environment(3, 3) # other sizes doesn't really make sense!
p1 = Agent(player_number=1) # player x
p1.eps = 0.25
p1.alpha = 0.1
p2 = Agent(player_number=2) # player o
p2.eps = 0.3
p2.alpha = 0.1
# set initial state values for both players
state_winner_triples = get_state_number_winner_ended_triple(env, verbose_lvl=1)
env.state_number_winner_ended_triple = state_winner_triples
Vx = initialV_x(env, state_winner_triples)
p1.set_state_values(Vx)
def test_get(self):
env = Environment(2, 2)
results = env.get_all_possible_states(env.board, 1, [])
self.assertTrue(len(results) != 0)
