def best_move(self, board, tleft):
logging.info("get_move {} {}".format(board, tleft))
logging.info("winmat {}".format(board.winmatrix()))
csm = board.canscorematrix()
logging.info("canscoremat {}".format(csm))
for sbi in range(9):
if board.macroboard[sbi] == -1 and csm[self.myid][sbi]:
scoring_loc = csm[self.myid][sbi]
logging.info("we can score on board {} loc {}".format(
sbi, scoring_loc))
scoring_move = board.locs_to_move(sbi, scoring_loc)
logging.info("we can score with move {}".format(scoring_move))
logging.info("sb\n{}".format(pprint(board.get_subboard(sbi))))
logging.info("canscore \n{}".format(
sb_logic.can_score(board.get_subboard(sbi), self.myid)))
return scoring_move
# logging.info("board {}".format(board.get_board()))
lmoves = board.legal_moves()
logging.info("lmoves {}".format(lmoves))
if len(lmoves) < 1:
logging.error("There is no legal moves")
for y in range(9):
for x in range(9):
if board.field[y * 9 + x] == 0:
return (x, y)
rm = randint(0, len(lmoves) - 1)
logging.info("rm {}".format(rm))
return board.locs_to_move(*lmoves[rm])
def set_values(self, board):
values = {(b, l): 0.0 for b in range(9) for l in range(9)}
win_value = 2000.0
lose_board_value = -15.0
score_value = 18.0
bad_score_value = -12.0
free_play_value = -18.0
cats_play_value = 20.0
center_board_value = 0.5
#priotize the center
for i in range(9):
values[(4, i)] += center_board_value
# priortize the board that could cause us to win
masked_macrob = [0 if i < 0 else i for i in board.macroboard]
canwin = sb_logic.can_score(masked_macrob, self.myid)
if canwin is not False:
logging.info("CAN WIN BOARD {}".format(canwin))
for l in range(9):
values[canwin, l] += win_value
# depriortize the board that could cause us to win
canlose = sb_logic.can_score(masked_macrob, self.oppid)
if canlose is not False:
for b in range(9):
values[b, canlose] += lose_board_value
# priortize scoring
csm = board.canscorematrix()
for b in range(9):
if csm[self.myid][b] is not False:
values[b, csm[self.myid][b]] += score_value
# depriortize him scoring
for b in range(9):
if csm[self.oppid][b] is not False:
for b2 in range(9):
values[b2, b] += bad_score_value
# prioritize letting him play on cats
catsmatrix = board.catsmatrix()
for b in range(9):
if catsmatrix[b] is True:
for b2 in range(9):
values[b2, b] += cats_play_value
# don't give him free play
for b in range(9):
if board.macroboard[b] > 0:
for b2 in range(9):
values[b2, b] += free_play_value
legal_values = {
k: v
for k, v in values.iteritems() if k in board.legal_moves()
}
return legal_values
def set_values(self, board):
values = {(b,l): 0.0 for b in range(9) for l in range(9)}
win_value = 2000.0
lose_board_value = -15.0
score_value = 18.0
bad_score_value = -12.0
free_play_value = -18.0
cats_play_value = 20.0
center_board_value = 0.5
#priotize the center
for i in range(9):
values[(4,i)] += center_board_value
# priortize the board that could cause us to win
masked_macrob = [0 if i < 0 else i for i in board.macroboard]
canwin = sb_logic.can_score(masked_macrob, self.myid)
if canwin is not False:
logging.info("CAN WIN BOARD {}".format(canwin))
for l in range(9):
values[canwin,l] += win_value
# depriortize the board that could cause us to win
canlose = sb_logic.can_score(masked_macrob, self.oppid)
if canlose is not False:
for b in range(9):
values[b,canlose] += lose_board_value
# priortize scoring
csm = board.canscorematrix()
for b in range(9):
if csm[self.myid][b] is not False:
values[b,csm[self.myid][b]] += score_value
# depriortize him scoring
for b in range(9):
if csm[self.oppid][b] is not False:
for b2 in range(9):
values[b2,b] += bad_score_value
# prioritize letting him play on cats
catsmatrix = board.catsmatrix()
for b in range(9):
if catsmatrix[b] is True:
for b2 in range(9):
values[b2,b] += cats_play_value
# don't give him free play
for b in range(9):
if board.macroboard[b] > 0:
for b2 in range(9):
values[b2,b] += free_play_value
legal_values = {k:v for k,v in values.iteritems() if k in board.legal_moves()}
return legal_values
def best_move(self, board, tleft):
logging.info("get_move {} {}".format(board, tleft))
# testb = [0, 1, 1, 0, 1, 2, 0, 2, 1]
# logging.info("testb?!\n{}".format(pprint(testb)))
# logging.info("test1?! {}".format(sb_logic.can_score(testb, 1)))
# logging.info("test2?! {}".format(sb_logic.can_score(testb, 2)))
# exit(-1)
logging.info("winmat {}".format(board.winmatrix()))
csm = board.canscorematrix()
logging.info("canscoremat {}".format(csm))
values = self.set_values(board)
logging.info("values {}".format(values))
best_moves = [
k for k, v in values.iteritems() if v == max(values.values())
]
logging.info("best moves {}".format(best_moves))
if len(best_moves) < 1:
logging.error("There is no legal moves")
for y in range(9):
for x in range(9):
if board.field[y * 9 + x] == 0:
return (x, y)
rm = randint(0, len(best_moves) - 1)
logging.info("rm {}".format(rm))
return board.locs_to_move(*best_moves[rm])
def best_move(self, board, tleft):
logging.info("get_move {} {}".format(board, tleft))
# testb = [0, 1, 1, 0, 1, 2, 0, 2, 1]
# logging.info("testb?!\n{}".format(pprint(testb)))
# logging.info("test1?! {}".format(sb_logic.can_score(testb, 1)))
# logging.info("test2?! {}".format(sb_logic.can_score(testb, 2)))
# exit(-1)
logging.info("winmat {}".format(board.winmatrix()))
csm = board.canscorematrix()
logging.info("canscoremat {}".format(csm))
values = self.set_values(board)
logging.info("values {}".format(values))
best_moves = [k for k,v in values.iteritems() if v == max(values.values())]
logging.info("best moves {}".format(best_moves))
if len(best_moves) < 1:
logging.error("There is no legal moves")
for y in range(9):
for x in range(9):
if board.field[y*9+x] == 0:
return (x, y)
rm = randint(0, len(best_moves)-1)
logging.info("rm {}".format(rm))
return board.locs_to_move(*best_moves[rm])