class LocalLooper:
message_box = {-1: "超出下单限制", -2: "超出涨跌价格", -3: "未成交", -4: "资金不足"}
def __init__(self, app_signal, app):
""" 需要构建完整的成交回报以及发单报告,
在account里面需要存储大量的存储
在我们此处实现过程也通过调用事件引擎来进行调用
"""
self.app = app
# 活跃报单数量
self.change_month_record = {}
self.pending = []
self.sessionid = random.randint(1000, 10000)
self.frontid = random.randint(10001, 500000)
# 日志输出器
self.app_signal = app_signal
# 策略池子
self.strategy_mapping = dict()
# 覆盖里面的action和logger属性
# 涨跌停价格
self.upper_price = 99999
self.drop_price = 0
# 风控/risk control todo:完善
self.params = dict(deal_pattern="match",
single_order_limit=10,
single_day_limit=100,
today_exchange=['INE', "SHFE"])
# 账户属性
self.account = Account(self)
self.order_ref = 0
# 发单的ref集合
self.order_ref_set = set()
# 已经order_id ---- 成交单
self.traded_order_mapping = {}
# 已经order_id --- 报单
self.order_id_pending_mapping = {}
# 当日成交笔数, 需要如果是第二天的数据，那么需要被清空
self.today_volume = 0
self.pre_close_price = dict()
# 所有的报单数量
self.order_buffer = dict()
self.date = None
# 行情
self.data_entity = None
self.if_next_day = False
self.data_type = "bar"
self.price_mapping = dict()
# 仓位详细
self.position_detail = dict()
def get_trades(self):
return list(self.traded_order_mapping.values())
def on_event(self, type, data):
event = Event(type=type, data=data)
if type == EVENT_BAR or type == EVENT_TICK:
import ctpbee.signals as signals
signal = getattr(signals.common_signals, f"{type}_signal")
else:
signal = getattr(self.app_signal, f"{type}_signal")
signal.send(event)
def enable_extension(self, name):
if name in self.strategy_mapping.keys():
self.strategy_mapping.get(name).active = True
else:
return
def suspend_extension(self, name):
if name in self.strategy_mapping.keys():
self.strategy_mapping.get(name).active = False
else:
return
def update_risk(self, risk):
self.risk = risk
def _generate_order_data_from_req(self, req: OrderRequest):
""" 将发单请求转换为发单数据 """
self.order_ref += 1
order_id = f"{self.frontid}-{self.sessionid}-{self.order_ref}"
return req._create_order_data(gateway_name="looper",
order_id=order_id,
time=self.datetime)
def _generate_trade_data_from_order(self, order_data: OrderData):
""" 将orderdata转换成成交单 """
p = TradeData(price=order_data.price,
istraded=order_data.volume,
volume=order_data.volume,
tradeid=str(uuid.uuid1()),
offset=order_data.offset,
direction=order_data.direction,
gateway_name=order_data.gateway_name,
order_time=order_data.time,
time=self.datetime,
order_id=order_data.order_id,
symbol=order_data.symbol,
exchange=order_data.exchange)
return p
def send_order(self, order_req: OrderRequest):
""" 发单的操作 """
if order_req.volume == 0:
return 0
return self.intercept_gateway(order_req)
def _cancel(self, cancel_req):
""" 撤单机制 """
self.intercept_gateway(cancel_req)
def cancel_order(self, cancel_req: CancelRequest, **kwargs):
rx = []
for x in self.pending:
if x.order_id == cancel_req.order_id:
x.status = Status.CANCELLED
# 回调 on_order做处理
self.on_event(EVENT_ORDER, deepcopy(x))
# 移除掉冻结 使得成为可能
self.account.pop_order(x)
# 更新仓位信息
self.account.position_manager.update_order(x)
rx.append(x)
for x in rx:
self.pending.remove(x)
def cancel_all(self):
self.pending.clear()
return 1
def intercept_gateway(self, data):
""" 拦截网关 同时这里应该返回相应的水平"""
if isinstance(data, OrderRequest):
""" 发单请求处理 """
order_data = self._generate_order_data_from_req(data)
result, reason = self.account.is_traded(order=order_data)
if result:
self.pending.append(order_data)
order_data.status = Status.NOTTRADED
self.on_event(EVENT_ORDER, order_data)
self.account.update_account_from_order(order_data)
return 1
else:
self.on_event(
EVENT_ERROR, f"账户报单可用不足, 报单基础信息: "
f"{order_data.local_symbol} volume: {order_data.volume}"
f" price: {order_data.price} {order_data.offset.value}{order_data.direction.value}"
f" 出现原因: {reason} 当前报单队列存在 {len(self.pending)}")
return 0
else:
pass
def match_deal(self):
""" 撮合成交
维护一个返回状态
-1: 超出下单限制
-2: 超出涨跌价格
-3: 未成交
-4: 资金不足
p : 成交回报
todo: 处理冻结 ??
"""
rc = []
for data in self.pending:
px = "".join(filter(str.isalpha, data.local_symbol))
nx = "".join(filter(str.isalpha, self.data_entity.local_symbol))
if nx != px: # 针对多品种，实现拆分。 更新当前的价格，确保多个
continue
if self.params.get("deal_pattern") == "match":
""" 撮合成交 """
# todo: 是否可以模拟一定量的市场冲击响应？ 以靠近更加逼真的回测效果 ？？？？
""" 调用API生成成交单 """
# 同时这里需要处理是否要进行
trade = self._generate_trade_data_from_order(data)
""" 这里按照市价进行匹配成交 """
trade.price = self.data_entity.last_price
self.on_event(
EVENT_LOG,
f"--> {trade.local_symbol} 成交时间: {str(trade.time)}, 成交价格{str(trade.price)}, 成交笔数: {str(trade.volume)},"
f" 成交方向: {str(trade.direction.value)}，行为: {str(trade.offset.value)}"
)
self.account.update_trade(trade)
""" 调用strategy的on_trade """
rc.append(data)
data.status = Status.ALLTRADED
self.on_event(EVENT_ORDER, data=data)
self.on_event(EVENT_TRADE, data=trade)
self.traded_order_mapping[trade.order_id] = trade
self.today_volume += data.volume
continue
if self.params.get("deal_pattern") == "umatch":
""" 撮合成交 """
# todo: 是否可以模拟一定量的市场冲击响应？ 以靠近更加逼真的回测效果 ？？？？
""" 调用API生成成交单 """
# 同时这里需要处理是否要进行
trade = self._generate_trade_data_from_order(data)
""" 这里按照市价进行匹配成交 """
self.on_event(
EVENT_LOG,
f"--> {trade.local_symbol} 成交时间: {str(trade.time)}, 成交价格{str(trade.price)}, 成交笔数: {str(trade.volume)},"
f" 成交方向: {str(trade.direction.value)}，行为: {str(trade.offset.value)}"
)
self.account.update_trade(trade)
""" 调用strategy的on_trade """
rc.append(data)
data.status = Status.ALLTRADED
self.on_event(EVENT_ORDER, data=data)
self.on_event(EVENT_TRADE, data=trade)
self.traded_order_mapping[trade.order_id] = trade
self.today_volume += data.volume
continue
elif self.params.get("deal_pattern") == "price":
""" 限价成交 """
# 先判断价格和手数是否满足限制条件
# 进行成交判断
long_c = self.data_entity.low_price if self.data_type == "bar" else self.data_entity.ask_price_1
short_c = self.data_entity.high_price if self.data_type == "bar" is not None else self.data_entity.bid_price_1
long_cross = data.direction == Direction.LONG and 0 < long_c <= data.price
short_cross = data.direction == Direction.SHORT and data.price <= short_c and short_c > 0
if long_cross:
""" 判断账户资金是否足以支撑成交 """
# 同时这里需要处理是否要进行
trade = self._generate_trade_data_from_order(data)
self.on_event(
EVENT_LOG,
data=
f"成交时间: {str(trade.time)}, 成交价格{str(trade.price)}, 成交笔数: {str(trade.volume)},"
f" 成交方向: {str(trade.direction.value)}，行为: {str(trade.offset.value)}"
)
""" 调用strategy的on_trade """
rc.append(data)
data.status = Status.ALLTRADED
self.account.update_trade(trade)
self.on_event(EVENT_ORDER, data=deepcopy(data))
self.on_event(EVENT_TRADE, data=deepcopy(trade))
self.traded_order_mapping[trade.order_id] = trade
self.today_volume += data.volume
continue
elif short_cross:
""" 调用API生成成交单 """
# 同时这里需要处理是否要进行
trade = self._generate_trade_data_from_order(data)
self.on_event(
EVENT_LOG,
data=
f"成交时间: {str(trade.time)}, 成交价格{str(trade.price)}, 成交笔数: {str(trade.volume)},"
f" 成交方向: {str(trade.direction.value)}，行为: {str(trade.offset.value)}"
)
""" 调用strategy的on_trade """
rc.append(data)
data.status = Status.ALLTRADED
self.account.update_trade(trade)
self.on_event(EVENT_ORDER, data=deepcopy(data))
self.on_event(EVENT_TRADE, data=deepcopy(trade))
self.traded_order_mapping[trade.order_id] = trade
self.today_volume += data.volume
continue
continue
else:
raise TypeError("未支持的成交机制")
for data in rc:
self.pending.remove(data)
def _init_params(self, params):
""" 回测参数设置 """
self.params.update(params)
""" 更新接口参数设置 """
self.params.update(params)
""" 更新账户策略参数 """
self.account.update_params(params)
def init_params(self, params):
""" 初始化参数设置 """
if not isinstance(params, dict):
raise AttributeError("回测参数类型错误，请检查是否为字典")
self._init_params(params.get("LOOPER"))
@staticmethod
def auth_time(time):
if 15 < time.hour <= 20 or 3 <= time.hour <= 8:
return False
else:
return True
def __call__(self, *args, **kwargs):
""" 回测周期 """
entity = args[0]
# 日期不相等时, 更新前日结算价格
if self.account.date is None:
self.account.date = self.date
if self.date is None:
self.date = entity.datetime.date()
if not self.auth_time(entity.datetime):
return
self.data_type = entity.type
# 回测的时候自动更新策略的日期
try:
seconds = (entity.datetime - self.datetime).seconds
if seconds >= 60 * 60 * 4 and (entity.datetime.hour >= 21 or (
(14 <= self.datetime.hour <= 15)
and entity.datetime.date() != self.datetime.date())):
self.account.settle(entity.datetime.date())
# 针对于账户的实现 我们需要将昨仓转换为今仓
self.app.recorder.position_manager.covert_to_yesterday_holding(
)
for local_symbol, price in self.price_mapping.items():
self.pre_close_price[local_symbol] = price
# 结算完触发初始化函数
self.on_event(EVENT_INIT_FINISHED, True)
except KeyError:
pass
except AttributeError:
pass
self.data_entity = entity
self.change_month_record["".join(
filter(str.isalpha,
entity.local_symbol.split(".")[0]))] = entity
# 维护一个最新的价格
self.price_mapping[self.data_entity.local_symbol] = self.data_entity.close_price if entity.type == "bar" \
else self.data_entity.last_price
if self.pre_close_price.get(self.data_entity.local_symbol) is None:
self.pre_close_price[
self.data_entity.
local_symbol] = self.data_entity.last_price if entity.type == "tick" else self.data_entity.close_price
self.datetime = entity.datetime
self.match_deal()
if entity.type == "tick":
self.account.position_manager.update_tick(
self.data_entity,
self.pre_close_price[self.data_entity.local_symbol])
self.app.recorder.position_manager.update_tick(
self.data_entity,
self.pre_close_price[self.data_entity.local_symbol])
self.on_event(EVENT_TICK, TickData(**entity))
if entity.type == "bar":
self.account.position_manager.update_bar(
self.data_entity,
self.pre_close_price[self.data_entity.local_symbol])
self.app.recorder.position_manager.update_bar(
self.data_entity,
self.pre_close_price[self.data_entity.local_symbol])
self.on_event(EVENT_BAR, BarData(**entity))
if entity.datetime.hour >= 21:
"""if hour > 21, switch to next trade day"""
index = trade_dates.index(str(entity.datetime.date()))
self.date = datetime.strptime(trade_dates[index + 1],
"%Y-%m-%d").date()
else:
if str(entity.datetime.date()) not in trade_dates:
last_day = entity.datetime + timedelta(days=-1)
self.date = datetime.strptime(
trade_dates[trade_dates.index(str(last_day.date())) + 1],
"%Y-%m-%d").date()
else:
self.date = entity.datetime.date()
# 穿过接口日期检查
self.account.via_aisle()
def get_entity_from_alpha(self, alpha):
return self.change_month_record.get(alpha)
class LocalLooper():
message_box = {-1: "超出下单限制", -2: "超出涨跌价格", -3: "未成交", -4: "资金不足"}
def __init__(self, logger, risk=None):
""" 需要构建完整的成交回报以及发单报告,在account里面需要存储大量的存储 """
# 活跃报单数量
self.change_month_record = {}
self.pending = []
self.sessionid = random.randint(1000, 10000)
self.frontid = random.randint(10001, 500000)
# 日志输出器
self.logger = logger
# 策略池子
self.strategy_mapping = dict()
# 覆盖里面的action和logger属性
# 涨跌停价格
self.upper_price = 99999
self.drop_price = 0
# 风控/risk control todo:完善
self.risk = risk
self.params = dict(deal_pattern="match",
single_order_limit=10,
single_day_limit=100,
today_exchange=['INE', "SHFE"])
# 账户属性
self.account = Account(self)
self.order_ref = 0
# 发单的ref集合
self.order_ref_set = set()
# 已经order_id ---- 成交单
self.traded_order_mapping = {}
# 已经order_id --- 报单
self.order_id_pending_mapping = {}
# 当日成交笔数, 需要如果是第二天的数据，那么需要被清空
self.today_volume = 0
self.pre_close_price = dict()
# 所有的报单数量
self.order_buffer = dict()
self.date = None
# 行情
self.data_entity = None
self.if_next_day = False
self.data_type = "bar"
self.price_mapping = dict()
# 仓位详细
self.position_detail = dict()
self.action = Action(self)
def get_trades(self):
return list(self.traded_order_mapping.values())
def update_strategy(self, strategy):
setattr(strategy, "action", self.action)
setattr(strategy, "logger", self.logger)
setattr(strategy, "info", self.logger.info)
setattr(strategy, "debug", self.logger.debug)
setattr(strategy, "error", self.logger.error)
setattr(strategy, "warning", self.logger.warning)
setattr(strategy, "app", self)
self.strategy_mapping[strategy.name] = strategy
def enable_extension(self, name):
if name in self.strategy_mapping.keys():
self.strategy_mapping.get(name).active = True
else:
return
def suspend_extension(self, name):
if name in self.strategy_mapping.keys():
self.strategy_mapping.get(name).active = False
else:
return
def update_risk(self, risk):
self.risk = risk
def _generate_order_data_from_req(self, req: OrderRequest):
""" 将发单请求转换为发单数据 """
self.order_ref += 1
order_id = f"{self.frontid}-{self.sessionid}-{self.order_ref}"
return req._create_order_data(gateway_name="looper",
order_id=order_id,
time=self.datetime)
def _generate_trade_data_from_order(self, order_data: OrderData):
""" 将orderdata转换成成交单 """
p = TradeData(price=order_data.price,
istraded=order_data.volume,
volume=order_data.volume,
tradeid=str(uuid.uuid1()),
offset=order_data.offset,
direction=order_data.direction,
gateway_name=order_data.gateway_name,
order_time=order_data.time,
time=self.datetime,
order_id=order_data.order_id,
symbol=order_data.symbol,
exchange=order_data.exchange)
return p
def send_order(self, order_req: OrderRequest):
""" 发单的操作 """
if order_req.volume == 0:
return 0
return self.intercept_gateway(order_req)
def _cancel(self, cancel_req):
""" 撤单机制 """
self.intercept_gateway(cancel_req)
def cancel(self, order_id):
for x in self.pending:
if x.order_id == order_id:
self.pending.remove(x)
return 1
return 0
def cancel_all(self):
self.pending.clear()
return 1
def intercept_gateway(self, data):
""" 拦截网关 同时这里应该返回相应的水平"""
if isinstance(data, OrderRequest):
""" 发单请求处理 """
order_data = self._generate_order_data_from_req(data)
if self.account.is_traded(order=order_data):
self.pending.append(order_data)
self.account.update_account_from_order(order_data)
return 1
else:
self.logger.info("报单可用不足")
self.logger.debug(f"close_profit: {self.account.close_profit}")
self.logger.debug(f"margin: {self.account.margin}")
self.logger.debug(f"pre_balance: {self.account.pre_balance}")
self.logger.debug(f"float_pnl: {self.account.float_pnl}")
self.logger.debug(
f"frozen_margin: {self.account.frozen_margin}")
self.logger.debug(f"available: {self.account.available}")
return 0
if isinstance(data, CancelRequest):
""" 撤单请求处理
"""
for order in self.pending:
if data.order_id == order.order_id:
order = deepcopy(order)
[api(order) for api in self.strategy_mapping.values()]
self.pending.remove(order)
self.account.pop_order(order)
return 1
return 0
def match_deal(self):
""" 撮合成交
维护一个返回状态
-1: 超出下单限制
-2: 超出涨跌价格
-3: 未成交
-4: 资金不足
p : 成交回报
todo: 处理冻结 ??
"""
for data in self.pending:
px = "".join(filter(str.isalpha, data.local_symbol))
nx = "".join(filter(str.isalpha, self.data_entity.local_symbol))
if nx != px: # 针对多品种，实现拆分。 更新当前的价格，确保多个
continue
if self.params.get("deal_pattern") == "match":
""" 撮合成交 """
# todo: 是否可以模拟一定量的市场冲击响应？ 以靠近更加逼真的回测效果 ？？？？
if self.account.is_traded(data):
""" 调用API生成成交单 """
# 同时这里需要处理是否要进行
trade = self._generate_trade_data_from_order(data)
self.logger.info(
f"--> {trade.local_symbol} 成交时间: {str(trade.time)}, 成交价格{str(trade.price)}, 成交笔数: {str(trade.volume)},"
f" 成交方向: {str(trade.direction.value)}，行为: {str(trade.offset.value)}, "
f"账户净值: {self.account.balance} 保证金: {self.account.margin} 账户剩余可用: {self.account.available} 此时队列中的单子: {len(self.pending)}"
)
self.account.update_trade(trade)
""" 调用strategy的on_trade """
self.pending.remove(data)
data.status = Status.ALLTRADED
[
api(deepcopy(data))
for api in self.strategy_mapping.values()
]
[api(trade) for api in self.strategy_mapping.values()]
self.traded_order_mapping[trade.order_id] = trade
self.today_volume += data.volume
else:
print("单子没成交哦")
continue
elif self.params.get("deal_pattern") == "price":
""" 见价成交 """
# 先判断价格和手数是否满足限制条件
if data.price < self.drop_price or data.price > self.upper_price:
""" 超出涨跌价格 """
continue
# 进行成交判断
long_c = self.data_entity.low_price if self.data_type == "bar" else self.data_entity.ask_price_1
short_c = self.data_entity.high_price if self.data_type == "bar" is not None else self.data_entity.bid_price_1
# long_b = self.data_entity.open_price if self.data_type == "bar" is not None else long_c
# short_b = self.data_entity.open_price if self.data_type == "bar" is not None else short_c
long_cross = data.direction == Direction.LONG and 0 < long_c <= data.price
short_cross = data.direction == Direction.SHORT and data.price <= short_c and short_c > 0
if long_cross:
""" 判断账户资金是否足以支撑成交 """
if self.account.is_traded(data):
""" 调用API生成成交单 """
# 同时这里需要处理是否要进行
trade = self._generate_trade_data_from_order(data)
self.logger.info(
f"成交时间: {str(trade.time)}, 成交价格{str(trade.price)}, 成交笔数: {str(trade.volume)},"
f" 成交方向: {str(trade.direction.value)}，行为: {str(trade.offset.value)}"
)
self.account.update_trade(trade)
""" 调用strategy的on_trade """
self.pending.remove(data)
data.status = Status.ALLTRADED
[
api(deepcopy(data))
for api in self.strategy_mapping.values()
]
[api(trade) for api in self.strategy_mapping.values()]
self.traded_order_mapping[trade.order_id] = trade
self.today_volume += data.volume
if short_cross:
if self.account.is_traded(data):
""" 调用API生成成交单 """
# 同时这里需要处理是否要进行
trade = self._generate_trade_data_from_order(data)
self.logger.info(
f"成交时间: {str(trade.time)}, 成交价格{str(trade.price)}, 成交笔数: {str(trade.volume)},"
f" 成交方向: {str(trade.direction.value)}，行为: {str(trade.offset.value)}"
)
self.account.update_trade(trade)
""" 调用strategy的on_trade """
self.pending.remove(data)
data.status = Status.ALLTRADED
[
api(deepcopy(data))
for api in self.strategy_mapping.values()
]
[api(trade) for api in self.strategy_mapping.values()]
self.traded_order_mapping[trade.order_id] = trade
self.today_volume += data.volume
continue
else:
""" 当前账户不足以支撑成交 """
continue
else:
raise TypeError("未支持的成交机制")
def init_params(self, params):
""" 回测参数设置 """
self.params.update(params)
""" 更新接口参数设置 """
self.params.update(params)
""" 更新账户策略参数 """
self.account.update_params(params)
def __init_params(self, params):
""" 初始化参数设置 """
if not isinstance(params, dict):
raise AttributeError("回测参数类型错误，请检查是否为字典")
[
strategy.init_params(params.get("strategy"))
for strategy in self.strategy_mapping.values()
]
self.init_params(params.get("looper"))
@staticmethod
def auth_time(time):
if 15 < time.hour <= 20 or 3 <= time.hour <= 8:
return False
else:
return True
def __call__(self, *args, **kwargs):
""" 回测周期 """
entity, params = args
# 日期不相等时, 更新前日结算价格
if self.account.date is None:
self.account.date = self.date
if self.date is None:
self.date = entity.datetime.date()
if not self.auth_time(entity.datetime):
return
self.data_type = entity.type
# 回测的时候自动更新策略的日期
if entity.datetime.hour > 21:
dt = entity.datetime + timedelta(days=1)
else:
dt = entity.datetime
[
setattr(x, "date", str(dt.date()))
for x in self.strategy_mapping.values()
]
self.__init_params(params)
try:
seconds = (entity.datetime - self.datetime).seconds
if seconds >= 60 * 60 * 4 and (entity.datetime.hour >= 21 or (
(14 <= self.datetime.hour <= 15)
and entity.datetime.date() != self.datetime.date())):
self.logger.warning("结算数据: " + str(self.account.date))
self.account.settle(entity.datetime.date())
if self.data_entity is None:
self.pre_close_price[
self.data_entity.
local_symbol] = entity.close_price if entity.type == "bar" else entity.last_price
else:
self.pre_close_price[
self.data_entity.local_symbol] = self.data_entity.close_price if entity.type == "bar" \
else self.data_entity.last_price
# 结算完触发初始化函数
[x.on_init(entity) for x in self.strategy_mapping.values()]
except KeyError:
pass
except AttributeError:
pass
self.data_entity = entity
self.change_month_record["".join(
filter(str.isalpha,
entity.local_symbol.split(".")[0]))] = entity
# 维护一个最新的价格
self.price_mapping[self.data_entity.local_symbol] = self.data_entity.close_price if entity.type == "bar" \
else self.data_entity.last_price
if self.pre_close_price.get(self.data_entity.local_symbol) is None:
self.pre_close_price[
self.data_entity.
local_symbol] = self.data_entity.last_price if entity.type == "tick" else self.data_entity.close_price
self.datetime = entity.datetime
self.match_deal()
if entity.type == "tick":
[api(entity) for api in self.strategy_mapping.values()]
self.account.position_manager.update_tick(
self.data_entity,
self.pre_close_price[self.data_entity.local_symbol])
if entity.type == "bar":
[api(entity) for api in self.strategy_mapping.values()]
self.account.position_manager.update_bar(
self.data_entity,
self.pre_close_price[self.data_entity.local_symbol])
# 更新接口的日期
self.date = entity.datetime.date()
# 穿过接口日期检查
self.account.via_aisle()
def get_entity_from_alpha(self, alpha):
return self.change_month_record.get(alpha)