def TEST_MINIMAX():
from basicplayer import minimax
tup_tree = ("A", None,
("B", None,
("C", None,
("D", 2),
("E", 2)),
("F", None,
("G", 0),
("H", 4))
),
("I", None,
("J", None,
("K", 6),
("L", 8)),
("M", None,
("N", 4),
("O", 6))
)
)
tree = make_tree(tup_tree)
print "%s:\n%s" % ("TREE_1", tree_as_string(tree))
v = minimax(tree, 10,
tree_eval_minimax,
tree_get_next_move,
is_leaf)
print "NEXT STEP: %s" % (v)
def TEST_4(expected):
from basicplayer import minimax
tup_tree = ("A", None,
("B", None,
("C", None,
("D", 2),
("E", 2)),
("F", None,
("G", 0),
("H", 4))
),
("I", None,
("J", None,
("K", 6),
("L", 8)),
("M", None,
("N", 4),
("O", 6))
)
)
tree = make_tree(tup_tree)
print "%s:\n%s" %("TREE_1", tree_as_string(tree))
v,x = minimax(tree, 10,
tree_eval,
tree_get_next_move,
is_leaf)
print "BEST MOVE: %s" %(v)
print "EXPECTED: %s" %(expected)
def TEST_MINIMAX():
from basicplayer import minimax
tup_tree = ("A", None, ("B", None, ("C", None, ("D", 2), ("E", 2)),
("F", None, ("G", 0), ("H", 4))),
("I", None, ("J", None, ("K", 6), ("L", 8)),
("M", None, ("N", 4), ("O", 6))))
tree = make_tree(tup_tree)
print "%s:\n%s" % ("TREE_1", tree_as_string(tree))
v = minimax(tree, 10, tree_eval_minimax, tree_get_next_move, is_leaf)
print "NEXT STEP: %s" % (v)
def TEST_4(expected):
from basicplayer import minimax
tup_tree = ("A", None, ("B", None, ("C", None, ("D", 2), ("E", 2)),
("F", None, ("G", 0), ("H", 4))),
("I", None, ("J", None, ("K", 6), ("L", 8)),
("M", None, ("N", 4), ("O", 6))))
tree = make_tree(tup_tree)
print "%s:\n%s" % ("TREE_1", tree_as_string(tree))
v, x = minimax(tree, 10, tree_eval, tree_get_next_move, is_leaf)
print "BEST MOVE: %s" % (v)
print "EXPECTED: %s" % (expected)
def TEST_3(expected):
from basicplayer import minimax
tup_tree = ("A", None,
("B", None,
("E", None,
("K", 8),
("L", 2)),
("F", 6)
),
("C", None,
("G", None,
("M", None,
("S", 4),
("T", 5)),
("N", 3)),
("H", None,
("O", 9),
("P", None,
("U", 10),
("V", 8))
),
),
("D", None,
("I", 1),
("J", None,
("Q", None,
("W", 7),
("X", 12)),
("K", None,
("Y", 11),
("Z", 15)
),
)
)
)
tree = make_tree(tup_tree)
print "%s:\n%s" %("TREE_3",
tree_as_string(tree))
v,x = minimax(tree, 10,
tree_eval,
tree_get_next_move,
is_leaf)
print "BEST-MOVE: %s" %(v)
print "EXPECTED: %s" %(expected)
def TEST_3(expected):
from basicplayer import minimax
tup_tree = ("A", None,
("B", None, ("E", None, ("K", 8), ("L", 2)), ("F", 6)), (
"C",
None,
("G", None, ("M", None, ("S", 4), ("T", 5)), ("N", 3)),
("H", None, ("O", 9), ("P", None, ("U", 10), ("V", 8))),
), ("D", None, ("I", 1), (
"J",
None,
("Q", None, ("W", 7), ("X", 12)),
("K", None, ("Y", 11), ("Z", 15)),
)))
tree = make_tree(tup_tree)
print "%s:\n%s" % ("TREE_3", tree_as_string(tree))
v, x = minimax(tree, 10, tree_eval, tree_get_next_move, is_leaf)
print "BEST-MOVE: %s" % (v)
print "EXPECTED: %s" % (expected)
else:
score = board.longest_chain(board.get_current_player_id()) * 10
# Prefer having your pieces in the center of the board.
for row in range(6):
for col in range(7):
if board.get_cell(row, col) == board.get_current_player_id():
score -= abs(3 - col)
elif board.get_cell(row, col) == board.get_other_player_id():
score += abs(3 - col)
return score
# Create a "player" function that uses the focused_evaluate function
# You can test this player by choosing 'quick' in the main program.
quick_to_win_player = lambda board: minimax(
board, depth=4, eval_fn=focused_evaluate)
# TODO Write an alpha-beta-search procedure that acts like the minimax-search
# procedure, but uses alpha-beta pruning to avoid searching bad ideas
# that can't improve the result. The tester will check your pruning by
# counting the number of static evaluations you make.
# You can use minimax() in basicplayer.py as an example.
# NOTE: You should use get_next_moves_fn when generating
# next board configurations, and is_terminal_fn when
# checking game termination.
# The default functions for get_next_moves_fn and is_terminal_fn set here will work for connect_four.
def alphabeta_find_board_value(alpha,
beta,
maxTurn,
def test_search_2(self):
actual_score = minimax(self.BARELY_WINNING_BOARD, 2, focused_evaluate)
expected_score = 3
self.assertEqual(actual_score, expected_score)
