def register_experience_to_sars(self, experiences: deque):
for experience in experiences:
# create dahai
if experience.action["type"] == MjMove.dahai.value:
dahai_actor = experience.action["actor"]
player_state = experience.state[dahai_actor]
# create dahai s_a_r
if player_state.dahai_observation is not None\
and not experience.board_state.reach[dahai_actor]:
label = Pai.str_to_id(experience.action["pai"])
self.dahai_queue.append(
tuple((
player_state.dahai_observation,
label,
experience.
reward, # already set to actor oriented align reward
)))
def think_on_tsumo(self, board_state: BoardState, candidates: List[Dict]):
my_tehais = board_state.tehais[self.id]
rest_in_view = board_state.restpai_in_view[self.id]
# if can hora, always do hora
hora_candidates = [
c for c in candidates if c['type'] == MjMove.hora.value
]
if len(hora_candidates) == 1:
return hora_candidates[0]
# if can reach, always do reach
reach_candiadtes = [
c for c in candidates if c['type'] == MjMove.reach.value
]
if len(reach_candiadtes) == 1:
return reach_candiadtes[0]
# dahai think
dahai_candiadtes = [
c for c in candidates if c['type'] == MjMove.dahai.value
]
# calclate the number of shanten reducing tsumo with each dahai
node = Node(my_tehais)
valid_nums = node.calc_ukeire_num(rest_num=rest_in_view)
max_valid_dahai_action = None
max_valid_num = -1
for candidate in dahai_candiadtes:
valid_dahai_id = Pai.str_to_id(candidate['pai'])
valid_num = valid_nums[valid_dahai_id]
if max_valid_num < valid_num:
max_valid_dahai_action = candidate
max_valid_num = valid_num
return max_valid_dahai_action
def restnum_update(self, action):
if action is None or "type" not in action:
return
if action["type"] == MjMove.start_kyoku.value:
# initialize
self.rest_pai_view = [[4] * 34 for _ in range(4)]
dora_marker_id = Pai.str_to_id(action["dora_marker"])
for i in range(4):
self.rest_pai_view[i][dora_marker_id] -= 1
for pai in action["tehais"][i]:
pai_id = Pai.str_to_id(pai)
self.rest_pai_view[i][pai_id] -= 1
elif action["type"] == MjMove.tsumo.value:
pai_id = Pai.from_str(action["pai"]).id
actor = action["actor"]
self.rest_pai_view[actor][pai_id] -= 1
elif action["type"] == MjMove.dahai.value:
pai_id = Pai.from_str(action["pai"]).id
actor = action["actor"]
for i in range(4):
if actor != i:
self.rest_pai_view[i][pai_id] -= 1
elif action["type"] == MjMove.dora.value:
pai_id = Pai.from_str(action["dora_marker"]).id
for i in range(4):
self.rest_pai_view[i][pai_id] -= 1
elif action["type"] == MjMove.chi.value:
actor = action["actor"]
for i in range(4):
if actor != i:
for consume in action["consumed"]:
self.rest_pai_view[i][Pai.str_to_id(consume)] -= 1
elif action["type"] == MjMove.pon.value:
actor = action["actor"]
for i in range(4):
if actor != i:
for consume in action["consumed"]:
self.rest_pai_view[i][Pai.str_to_id(consume)] -= 1
elif action["type"] == MjMove.daiminkan.value:
actor = action["actor"]
for i in range(4):
if actor != i:
for consume in action["consumed"]:
self.rest_pai_view[i][Pai.str_to_id(consume)] -= 1
elif action["type"] == MjMove.ankan.value:
actor = action["actor"]
for i in range(4):
if actor != i:
for consume in action["consumed"]:
self.rest_pai_view[i][Pai.str_to_id(consume)] -= 1
elif action["type"] == MjMove.kakan.value:
actor = action["actor"]
pai_id = Pai.str_to_id(action["pai"])
for i in range(4):
if actor != i:
self.rest_pai_view[i][pai_id] -= 1
def update_state(self, action):
if self.board.previous_action is not None and \
self.board.previous_action['type'] in [MjMove.dahai.value, MjMove.kakan.value] and \
'actor' in self.board.previous_action and \
self.board.previous_action['actor'] != self.id and \
action['type'] != MjMove.hora.value:
self.extra_anpais.append(Pai.from_str(self.board.previous_action['pai']))
action_type = action['type']
if action_type == MjMove.start_game.value:
if not self.id and 'id' in action:
self.id = action['id']
self.name = f"player{self.id}"
if 'names' in action:
self.name = action['names'][self.id]
self.score = 25000
self.tehais = []
self.furos = []
self.ho = []
self.sutehais = []
self.extra_anpais = []
self.reach_state = None
self.reach_ho_index = None
self.reach_sutehais_index = None
self.double_reach = False
self.ippatsu_chance = False
self.pao_for_id = None
self.rinshan = False
elif action_type == MjMove.start_kyoku.value:
self.tehais = sorted(Pai.from_list(action['tehais'][self.id]))
self.furos = []
self.ho = []
self.sutehais = []
self.extra_anpais = []
self.reach_state = "none"
self.reach_ho_index = None
self.reach_sutehais_index = None
self.double_reach = False
self.ippatsu_chance = False
self.pao_for_id = None
self.rinshan = False
elif action_type in [
MjMove.chi.value,
MjMove.pon.value,
MjMove.daiminkan.value,
MjMove.ankan.value
]:
self.ippatsu_chance = False
elif action_type == MjMove.tsumo.value:
if self.board.previous_action['type'] == MjMove.kakan.value:
self.ippatsu_chance = False
if 'actor' in action and action['actor'] == self.id:
if action_type == MjMove.tsumo.value:
pai = Pai.from_str(action['pai'])
self.tehais = sorted(self.tehais)
self.tehais.append(pai)
elif action_type == MjMove.dahai.value:
pai = Pai.from_str(action['pai'])
self.delete_tehai(pai)
self.tehais = sorted(self.tehais)
self.ho.append(pai)
self.sutehais.append(pai)
self.ippatsu_chance = False
self.rinshan = False
if self.reach == False:
self.extra_anpais.clear()
elif action_type in [
MjMove.chi.value,
MjMove.pon.value,
MjMove.daiminkan.value,
MjMove.ankan.value
]:
consumed_pais = Pai.from_list(action['consumed'])
for c in consumed_pais:
self.delete_tehai(c)
furo = {
'type':action['type'],
'consumed':consumed_pais,
}
if action_type != MjMove.ankan.value:
pai = Pai.from_str(action['pai'])
furo['taken'] = pai
if action_type == MjMove.chi.value:
furo['pai_id'] = min(
Pai.str_to_id(action['pai']),
min(
Pai.str_to_id(action['consumed'][0]),
Pai.str_to_id(action['consumed'][1]))
)
else:
furo['pai_id'] = Pai.str_to_id(action['consumed'][0])
if action_type == MjMove.ankan.value:
furo['target'] = self.id
else:
furo['target'] = action['target']
self.furos.append(Furo(furo))
if action_type in [MjMove.daiminkan.value, MjMove.ankan.value]:
self.rinshan = True
# pao
if action_type in [MjMove.daiminkan.value, MjMove.pon.value]:
pai = Pai.from_str(action['pai'])
if pai.is_sangenpai():
if self.is_daisangen_pao():
self.pao_for_id = action['target']
elif pai.is_wind():
if self.is_daisushi_pao():
self.pao_for_id = action['target']
elif action_type == MjMove.kakan.value:
pai = Pai.from_str(action['pai'])
self.delete_tehai(pai)
pon_index = -1
for i,f in enumerate(self.furos):
if f.type == MjMove.pon.value and pai.is_same_symbol(f.taken):
pon_index = i
if pon_index == -1:
raise Exception('not have same symbole pon')
self.furos[pon_index] = Furo({
'type':MjMove.kakan.value,
'taken':self.furos[pon_index].taken,
'consumed':self.furos[pon_index].consumed + [pai],
'target':self.furos[pon_index].target,
'pai_id':self.furos[pon_index].pai_id,
})
self.rinshan = True
elif action_type == MjMove.reach.value:
self.reach_state = 'declared'
self.double_reach = self.board.first_turn
elif action_type == MjMove.reach_accepted.value:
self.reach_state = 'accepted'
self.reach_ho_index = len(self.ho)-1
self.reach_sutehais_index = len(self.sutehais)-1
self.ippatsu_chance = True
if 'target' in action and action['target'] == self.id:
pai = Pai.from_str(action['pai'])
if action_type in [
MjMove.pon.value,
MjMove.chi.value,
MjMove.daiminkan.value
]:
taken = self.ho.pop()
# assert taken == pai
if 'scores' in action:
self.score = action['scores'][self.id]