def run_game(trial, diff_level=10):
state = (0, 0)
i = 1
total_pts = 0
while i <= 13:
logger.info("%s,diff %2d, trial %d, turn %d", '=' * 20, diff_level,
trial, i)
i += 1
roll_left = 2
bigstate = get_state_big_mask(state)
roll0 = dice_to_id[roll_dice(5)]
logger.info('dice0, %s', id_to_dice[roll0])
kid0 = get_option_by_decoding(bigstate, roll0, roll_left, diff_level)
keep0 = id_to_kept[kid0]
logger.info('keep0,%s', keep0)
roll_left -= 1
roll1 = dice_to_id[merge(keep0, roll_dice(5 - len(keep0)))]
logger.info('dice1,%s', id_to_dice[roll1])
kid1 = get_option_by_decoding(bigstate, roll1, roll_left, diff_level)
keep1 = id_to_kept[kid1]
logger.info('keep1,%s', keep1)
roll_left -= 1
roll2 = dice_to_id[merge(keep1, roll_dice(5 - len(keep1)))]
logger.info('dice2,%s', id_to_dice[roll2])
act = get_option_by_decoding(bigstate, roll2, roll_left, diff_level)
state = get_next_state(state, id_to_dice[roll2],
Category.CATEGORY_ID_TO_NAME[act])
pts = eval_points[Category.CATEGORY_ID_TO_NAME[act]][id_to_dice[roll2]]
logger.info('choice,%s, earn %f', Category.CATEGORY_ID_TO_NAME[act],
pts)
total_pts += pts
logger.info('this turn: %d, total pts: %d, upper score: %d', pts,
total_pts, state[1])
logger.info('state: %s', bin(state[0]))
if state[1] == 63:
total_pts += 35
logger.info('extra bonus added, total: %s', total_pts)
return total_pts
def run_game(trial,diff_level=10):
state = (0,0)
i=1
total_pts = 0
while i<=13:
logger.info("%s,diff %2d, trial %d, turn %d", '='*20, diff_level,trial,i)
i+=1
roll_left = 2
bigstate = get_state_big_mask(state)
roll0 = dice_to_id[roll_dice(5)]
logger.info('dice0, %s', id_to_dice[roll0])
kid0 = get_option_by_decoding(bigstate, roll0, roll_left, diff_level)
keep0 = id_to_kept[kid0]
logger.info('keep0,%s', keep0)
roll_left -= 1
roll1 = dice_to_id[merge(keep0, roll_dice(5-len(keep0)))]
logger.info('dice1,%s', id_to_dice[roll1])
kid1 = get_option_by_decoding(bigstate, roll1, roll_left, diff_level)
keep1 = id_to_kept[kid1]
logger.info('keep1,%s', keep1)
roll_left -= 1
roll2 = dice_to_id[merge(keep1, roll_dice(5-len(keep1)))]
logger.info('dice2,%s', id_to_dice[roll2])
act = get_option_by_decoding(bigstate, roll2, roll_left, diff_level)
state = get_next_state(state, id_to_dice[roll2], Category.CATEGORY_ID_TO_NAME[act])
pts = eval_points[Category.CATEGORY_ID_TO_NAME[act]][id_to_dice[roll2]]
logger.info('choice,%s, earn %f', Category.CATEGORY_ID_TO_NAME[act], pts)
total_pts += pts
logger.info('this turn: %d, total pts: %d, upper score: %d',pts, total_pts, state[1])
logger.info('state: %s', bin(state[0]))
if state[1] == 63:
total_pts += 35
logger.info( 'extra bonus added, total: %s', total_pts)
return total_pts
def run_game_two_players(trial, diff_level=50):
"""
the first player is the bot player with fixed difficulty level
the second player is the adaptive bot that tries to get a
final score as close as to the first play's.
"""
state = (0,0) # first player state
my_state = (0,0) # second player state
turn = 1
total_pts = 0
my_total_pts = 0
while turn<=13:
# first player begin
logger.info("%s,diff %2d, trial %d, p1-turn %d", '='*20, diff_level,trial,turn)
roll_left = 2
bigstate = get_state_big_mask(state)
roll0 = dice_to_id[roll_dice(5)]
logger.info('dice0, %s', id_to_dice[roll0])
kid0 = get_option_by_decoding(bigstate, roll0, roll_left, diff_level)
keep0 = id_to_kept[kid0]
logger.info('keep0,%s', keep0)
roll_left -= 1
roll1 = dice_to_id[merge(keep0, roll_dice(5-len(keep0)))]
logger.info('dice1,%s', id_to_dice[roll1])
kid1 = get_option_by_decoding(bigstate, roll1, roll_left, diff_level)
keep1 = id_to_kept[kid1]
logger.info('keep1,%s', keep1)
roll_left -= 1
roll2 = dice_to_id[merge(keep1, roll_dice(5-len(keep1)))]
logger.info('dice2,%s', id_to_dice[roll2])
act = get_option_by_decoding(bigstate, roll2, roll_left, diff_level)
state = get_next_state(state, id_to_dice[roll2], Category.CATEGORY_ID_TO_NAME[act])
pts = eval_points[Category.CATEGORY_ID_TO_NAME[act]][id_to_dice[roll2]]
my_diff_level = get_adaptive_diff(turn, my_total_pts - total_pts, diff_level)
total_pts += pts
logger.info('choice,%s, earn %f, exp-score %f', Category.CATEGORY_ID_TO_NAME[act], pts, total_pts + get_expected_score(state))
logger.info('this turn: %d, total pts: %d, upper score: %d',pts, total_pts, state[1])
logger.info("%s,diff %2d, trial %d, p2-turn %d", '='*20, diff_level,trial,turn)
# second player begin
# my intelligence is based on delta score and the current turn
my_roll_left = 2
my_bigstate = get_state_big_mask(my_state)
my_roll0 = dice_to_id[roll_dice(5)]
logger.info('dice0, %s', id_to_dice[my_roll0])
my_kid0 = get_option_by_decoding(my_bigstate, my_roll0, my_roll_left, my_diff_level)
my_keep0 = id_to_kept[my_kid0]
logger.info('keep0,%s', my_keep0)
my_roll_left -= 1
my_roll1 = dice_to_id[merge(my_keep0, roll_dice(5-len(my_keep0)))]
logger.info('dice1,%s', id_to_dice[my_roll1])
my_kid1 = get_option_by_decoding(my_bigstate, my_roll1, my_roll_left, my_diff_level)
my_keep1 = id_to_kept[my_kid1]
logger.info('keep1,%s', my_keep1)
my_roll_left -= 1
my_roll2 = dice_to_id[merge(my_keep1, roll_dice(5-len(my_keep1)))]
logger.info('dice2,%s', id_to_dice[my_roll2])
my_act = get_option_by_decoding(my_bigstate, my_roll2, my_roll_left, my_diff_level)
my_state = get_next_state(my_state, id_to_dice[my_roll2], Category.CATEGORY_ID_TO_NAME[my_act])
my_pts = eval_points[Category.CATEGORY_ID_TO_NAME[my_act]][id_to_dice[my_roll2]]
my_total_pts += my_pts
logger.info('choice,%s, earn %f, exp-score %f', Category.CATEGORY_ID_TO_NAME[my_act], my_pts, my_total_pts + get_expected_score(my_state))
logger.info('this turn: %d, total pts: %d, upper score: %d',my_pts, my_total_pts, my_state[1])
#logger.info('state: %s', bin(state[0]))
turn += 1
if state[1] == 63:
total_pts += 35
logger.info( 'p1 extra bonus added, total: %s', total_pts)
if my_state[1] == 63:
my_total_pts += 35
logger.info( 'p2 extra bonus added, total: %s', my_total_pts)
return my_total_pts - total_pts
def run_game_two_players(trial, diff_level=50):
"""
the first player is the bot player with fixed difficulty level
the second player is the adaptive bot that tries to get a
final score as close as to the first play's.
"""
state = (0, 0) # first player state
my_state = (0, 0) # second player state
turn = 1
total_pts = 0
my_total_pts = 0
while turn <= 13:
# first player begin
logger.info("%s,diff %2d, trial %d, p1-turn %d", '=' * 20, diff_level,
trial, turn)
roll_left = 2
bigstate = get_state_big_mask(state)
roll0 = dice_to_id[roll_dice(5)]
logger.info('dice0, %s', id_to_dice[roll0])
kid0 = get_option_by_decoding(bigstate, roll0, roll_left, diff_level)
keep0 = id_to_kept[kid0]
logger.info('keep0,%s', keep0)
roll_left -= 1
roll1 = dice_to_id[merge(keep0, roll_dice(5 - len(keep0)))]
logger.info('dice1,%s', id_to_dice[roll1])
kid1 = get_option_by_decoding(bigstate, roll1, roll_left, diff_level)
keep1 = id_to_kept[kid1]
logger.info('keep1,%s', keep1)
roll_left -= 1
roll2 = dice_to_id[merge(keep1, roll_dice(5 - len(keep1)))]
logger.info('dice2,%s', id_to_dice[roll2])
act = get_option_by_decoding(bigstate, roll2, roll_left, diff_level)
state = get_next_state(state, id_to_dice[roll2],
Category.CATEGORY_ID_TO_NAME[act])
pts = eval_points[Category.CATEGORY_ID_TO_NAME[act]][id_to_dice[roll2]]
my_diff_level = get_adaptive_diff(turn, my_total_pts - total_pts,
diff_level)
total_pts += pts
logger.info('choice,%s, earn %f, exp-score %f',
Category.CATEGORY_ID_TO_NAME[act], pts,
total_pts + get_expected_score(state))
logger.info('this turn: %d, total pts: %d, upper score: %d', pts,
total_pts, state[1])
logger.info("%s,diff %2d, trial %d, p2-turn %d", '=' * 20, diff_level,
trial, turn)
# second player begin
# my intelligence is based on delta score and the current turn
my_roll_left = 2
my_bigstate = get_state_big_mask(my_state)
my_roll0 = dice_to_id[roll_dice(5)]
logger.info('dice0, %s', id_to_dice[my_roll0])
my_kid0 = get_option_by_decoding(my_bigstate, my_roll0, my_roll_left,
my_diff_level)
my_keep0 = id_to_kept[my_kid0]
logger.info('keep0,%s', my_keep0)
my_roll_left -= 1
my_roll1 = dice_to_id[merge(my_keep0, roll_dice(5 - len(my_keep0)))]
logger.info('dice1,%s', id_to_dice[my_roll1])
my_kid1 = get_option_by_decoding(my_bigstate, my_roll1, my_roll_left,
my_diff_level)
my_keep1 = id_to_kept[my_kid1]
logger.info('keep1,%s', my_keep1)
my_roll_left -= 1
my_roll2 = dice_to_id[merge(my_keep1, roll_dice(5 - len(my_keep1)))]
logger.info('dice2,%s', id_to_dice[my_roll2])
my_act = get_option_by_decoding(my_bigstate, my_roll2, my_roll_left,
my_diff_level)
my_state = get_next_state(my_state, id_to_dice[my_roll2],
Category.CATEGORY_ID_TO_NAME[my_act])
my_pts = eval_points[Category.CATEGORY_ID_TO_NAME[my_act]][
id_to_dice[my_roll2]]
my_total_pts += my_pts
logger.info('choice,%s, earn %f, exp-score %f',
Category.CATEGORY_ID_TO_NAME[my_act], my_pts,
my_total_pts + get_expected_score(my_state))
logger.info('this turn: %d, total pts: %d, upper score: %d', my_pts,
my_total_pts, my_state[1])
#logger.info('state: %s', bin(state[0]))
turn += 1
if state[1] == 63:
total_pts += 35
logger.info('p1 extra bonus added, total: %s', total_pts)
if my_state[1] == 63:
my_total_pts += 35
logger.info('p2 extra bonus added, total: %s', my_total_pts)
return my_total_pts - total_pts