def test_history_correctly_represents_history_after_one_move():
env = Environment(2, 2, 2)
action = 0
start_state = env.get_state()
reward = env.take_action(action)
assert env.history == [(start_state, action, reward)]
def test_take_action_returns_0_when_no_win():
env = Environment(2, 2, 2)
result = env.take_action(0)
assert result == 0
result = env.take_action(1)
assert result == 0
def test_take_action_invalid_move_returns_neg_1():
env = Environment(2, 2, 2)
env.cur_player = Cell.BLACK
env.take_action(0) # black moves
env.take_action(1) # red moves
env.take_action(0) # black moves
result = env.take_action(0) # Invalid move
assert result == -1
def test_history_has_same_number_elements_as_moves():
env = Environment(2, 2, 2)
env.take_action(0)
assert len(env.history) == 1
env.take_action(0)
assert len(env.history) == 2
def test_reset_new_board():
env = Environment(2, 2, 2)
env.take_action(0)
env.reset()
assert env.b.cells[0, 0] == Cell.EMPTY
assert env.b.cells[0, 1] == Cell.EMPTY
assert env.b.cells[1, 0] == Cell.EMPTY
assert env.b.cells[1, 1] == Cell.EMPTY
def test_take_action_game_over_when_board_full():
env = Environment(2, 2, 2)
env.take_action(0)
env.take_action(0)
env.take_action(1)
env.take_action(1)
# 2x2 board is now full
assert env.is_game_over
def test_get_state_red_replaced_with_M():
'''
The position
R E
R B
should be represented as
Y E
Y M
when playing as black (M=mine, Y=yours)
'''
env = Environment(2, 2, 2)
env.cur_player = Cell.RED
env.take_action(0) # red moves
env.take_action(1) # black moves
env.take_action(0) # red moves
result = env.get_state()
expected = 'YYEM' # what black sees
assert result == expected
def test_history_correctly_represents_history_after_several_moves():
env = Environment(2, 2, 2)
# make some moves
env.take_action(0)
env.take_action(1)
# make one more move
action = 0
most_recent_state = env.get_state()
reward = env.take_action(action)
# check that history correctly appended to
last_move = env.history[-1]
assert last_move == (most_recent_state, action, reward)
def test_take_action_returns_1_when_win():
env = Environment(2, 2, 2)
env.take_action(0) # black moves
env.take_action(1) # red moves
result = env.take_action(0) # black moves and wins
assert result == 1
def test_take_action_curplayer_RED_changes_to_BLACK():
env = Environment(2, 2, 2)
env.cur_player = Cell.RED
env.take_action(0)
assert env.cur_player == Cell.BLACK
def test_num_cells_computed_correctly(board_w, board_h):
env = Environment(board_w, board_h, 2)
assert env.get_num_cells() == board_w * board_h
def test_reset_history_cleared():
env = Environment(2, 2, 2)
env.take_action(0)
env.reset()
assert env.history == []
def test_reset_new_board_num_moves_reset():
env = Environment(2, 2, 2)
env.take_action(0)
env.reset()
assert env.num_moves == 0