def test_get_move_chooses_random_available_move_if_random_lt_epsilon(
self, random_mock, choice_mock):
random_mock.return_value = 0.099
choice_mock.side_effect = MockRandom(4).choice
assert_get_move_is(self, self.player, self.board, 6, Board.X,
"X--|-O-|---")
choice_mock.assert_called_once_with([1, 2, 3, 5, 6, 7, 8])
def test_get_move_chooses_random_best_available_move_if_random_gte_epsilon_and_multiple_bests(
self, random_mock, choice_mock):
random_mock.return_value = 0.1
choice_mock.side_effect = MockRandom(1).choice
self.player.values[get_board_state_tuple(
"X--|-XO|---")] = 0.501 # position 0
self.player.values[get_board_state_tuple(
"--X|-XO|---")] = 0.501 # position 2
# Symmetries for X--|-XO|---:
# -O-|-X-|X-- (Rotated by 90 degrees)
# ---|OX-|--X (Rotated by 180 degrees)
# --X|-X-|-O- (Rotated by 270 degrees)
# ---|-XO|X-- (Reflected horizontally) - position 6
# --X|OX-|--- (Reflected vertically)
# X--|-X-|-O- (Reflected on left diagonal)
# -O-|-X-|--X (Reflected on right diagonal)
#
# Symmetries for --X|-XO|---:
# XO-|-X-|--- (Rotated by 90 degrees)
# ---|OX-|X-- (Rotated by 180 degrees)
# ---|-X-|-OX (Rotated by 270 degrees)
# ---|-XO|--X (Reflected horizontally) - position 8
# X--|OX-|--- (Reflected vertically)
# ---|-X-|XO- (Reflected on left diagonal)
# -OX|-X-|--- (Reflected on right diagonal)
assert_get_move_is(self, self.player, self.board, 2, Board.X,
"---|-XO|---")
choice_mock.assert_called_once_with([0, 2, 6, 8])
def test_get_move_chooses_best_available_move_if_random_gte_epsilon(
self, random_mock, choice_mock):
random_mock.return_value = 0.1
choice_mock.side_effect = MockRandom(0).choice
self.player.values[get_board_state_tuple("---|-XO|---")] = 0.501
assert_get_move_is(self, self.player, self.board, 5, Board.O,
"---|-X-|---")
choice_mock.assert_called_once_with([5])
def test_get_move_chooses_best_available_move_if_learning_disabled(
self, random_mock, choice_mock):
self.player.disable_learning()
random_mock.return_value = 0.099
choice_mock.side_effect = MockRandom(0).choice
self.player.values[get_board_state_tuple("---|-O-|--X")] = 0.501
assert_get_move_is(self, self.player, self.board, 4, Board.O,
"---|---|--X")
random_mock.assert_not_called()
choice_mock.assert_called_once_with([4])
def test_get_move_chooses_random_best_available_move_if_random_gte_epsilon_and_multiple_bests(
self, random_mock, choice_mock):
random_mock.return_value = 0.1
choice_mock.side_effect = MockRandom(1).choice
self.player.values[get_board_state_tuple("X--|-XO|---")] = 0.501
self.player.values[get_board_state_tuple("--X|-XO|---")] = 0.501
self.player.values[get_board_state_tuple("---|-XO|X--")] = 0.501
assert_get_move_is(self, self.player, self.board, 2, Board.X,
"---|-XO|---")
choice_mock.assert_called_once_with([0, 2, 6])
def test_get_move_chooses_best_available_move_if_random_gte_epsilon(
self, random_mock, choice_mock):
random_mock.return_value = 0.1
choice_mock.side_effect = MockRandom(0).choice
self.player.values[get_board_state_tuple(
"---|-XO|---")] = 0.501 # Original
# Reflected horizontally
# Force other symmetric choices to be of lower value
self.player.values[get_board_state_tuple(
"-O-|-X-|---")] = 0.499 # Rotated by 90 degrees
# Reflected on right diagonal
self.player.values[get_board_state_tuple(
"---|OX-|---")] = 0.499 # Rotated by 180 degrees
# Reflected vertically
self.player.values[get_board_state_tuple(
"---|-X-|-O-")] = 0.499 # Rotated by 270 degrees
# Reflected on left diagonal
assert_get_move_is(self, self.player, self.board, 5, Board.O,
"---|-X-|---")
choice_mock.assert_called_once_with([5])
def test_get_move_returns_random_available_move(self, choice_mock):
choice_mock.side_effect = MockRandom(5).choice
assert_get_move_is(self, self.player, self.board, 7, Board.X,
"---|-XO|---")