Update Version 3.5.5

Author: shinny-pack

PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: tqsdk
-Version: 3.5.4
+Version: 3.5.5
 Summary: TianQin SDK
 Home-page: https://www.shinnytech.com/tqsdk
 Author: TianQin

doc/conf.py

@@ -48,9 +48,9 @@
 # built documents.
 #
 # The short X.Y version.
-version = u'3.5.4'
+version = u'3.5.5'
 # The full version, including alpha/beta/rc tags.
-release = u'3.5.4'
+release = u'3.5.5'
 
 # The language for content autogenerated by Sphinx. Refer to documentation
 # for a list of supported languages.

doc/version.rst

@@ -2,6 +2,13 @@
 
 版本变更
 =============================
+3.5.5 (2024/03/27)
+
+* 修复：TqSim 在调用 set_margin 之后，使用 is_changing 判断某个对象是否更新，可能返回的结果不正确
+* 优化：多账户下使用 :py:meth:`~tqsdk.algorithm.time_table_generater.vwap_table`，
+    :py:meth:`~tqsdk.algorithm.time_table_generater.twap_table` 不需要用户多次指定账户
+
+
 3.5.4 (2024/03/01)
 
 * 修复：回测时，订阅多合约 K 线时，成交可能不符合预期的问题

setup.py

@@ -8,7 +8,7 @@
 
 setuptools.setup(
     name='tqsdk',
-    version="3.5.4",
+    version="3.5.5",
     description='TianQin SDK',
     author='TianQin',
     author_email='[email protected]',

tqsdk/__version__.py

@@ -1 +1 @@
-__version__ = '3.5.4'
+__version__ = '3.5.5'

tqsdk/algorithm/time_table_generater.py

@@ -13,6 +13,7 @@
 from tqsdk import utils
 from tqsdk.datetime import _get_trading_timestamp, _get_trade_timestamp, _get_trading_day_from_timestamp, \
     _datetime_to_timestamp_nano, _timestamp_nano_to_datetime
+from tqsdk.lib.time_table import TqTimeTable
 from tqsdk.rangeset import _rangeset_slice, _rangeset_head
 from tqsdk.tradeable import TqAccount, TqKq, TqSim
 
@@ -124,7 +125,7 @@ def twap_table(api: TqApi, symbol: str, target_pos: int, duration: int, min_volu
         interval_list = _gen_random_list(sum_val=duration, min_val=min_interval, max_val=max_interval,
                                          length=len(volume_list))
 
-    time_table = DataFrame(columns=['interval', 'volume', 'price'])
+    time_table = TqTimeTable(account=account)
     for index, volume in enumerate(volume_list):
         assert interval_list[index] >= 3
         active_interval = 2
@@ -253,7 +254,7 @@ def vwap_table(api: TqApi, symbol: str, target_pos: int, duration: float,
     predicted_percent = volume_percent.groupby(level=1).mean()  # 将历史上相同时间单元的成交量占比使用算数平均计算出预测值
 
     # 计算每个时间单元的成交量预测值
-    time_table = DataFrame(columns=['interval', 'volume', 'price'])
+    time_table = TqTimeTable(account=account)
     volume_left = target_volume  # 剩余手数
     percent_left = 1  # 剩余百分比
     for index, value in predicted_percent.items():

tqsdk/api.py

@@ -972,11 +972,12 @@ def _get_data_series(self, call_func: str, symbol_list: Union[str, List[str]], d
         if adj_type not in [None, "F", "B"]:
             raise Exception("adj_type 参数只支持 None (不复权) ｜ 'F' (前复权) ｜ 'B' (后复权) ")
         ds = DataSeries(self, symbol_list, dur_nano, start_dt_nano, end_dt_nano, adj_type)
-        while not self._loop.is_running() and not ds.is_ready:
-            deadline = time.time() + 30
-            if not self.wait_update(deadline=deadline):
-                raise TqTimeoutError(
-                    f"{call_func} 获取数据 ({symbol_list, duration_seconds, start_dt, end_dt}) 超时，请检查客户端及网络是否正常。")
+        if not self._loop.is_running():
+            while not ds._task.done():
+                deadline = time.time() + 30
+                if not self.wait_update(deadline=deadline, _task=ds._task):
+                    raise TqTimeoutError(
+                        f"{call_func} 获取数据 ({symbol_list, duration_seconds, start_dt, end_dt}) 超时，请检查客户端及网络是否正常。")
         return ds.df
 
     # ----------------------------------------------------------------------
@@ -1811,13 +1812,12 @@ def set_risk_management_rule(self, exchange_id: str, enable: bool, count_limit:
         rule = _get_obj(self._data, ["trade", self._account._get_account_key(account), "risk_management_rule", exchange_id], RiskManagementRule(self))
         if not self._loop.is_running():
             deadline = time.time() + 30
-            while not (rule_pack['enable'] == rule['enable']
-                       and rule_pack['self_trade'].items() <= rule['self_trade'].items()
-                       and rule_pack['frequent_cancellation'].items() <= rule['frequent_cancellation'].items()
-                       and rule_pack['trade_position_ratio'].items() <= rule['trade_position_ratio'].items()):
-                # @todo: merge diffs
-                if not self.wait_update(deadline=deadline):
-                    raise TqTimeoutError("设置风控规则超时请检查客户端及网络是否正常")
+            cond = lambda: (rule_pack['enable'] == rule['enable']
+                        and rule_pack['self_trade'].items() <= rule['self_trade'].items()
+                        and rule_pack['frequent_cancellation'].items() <= rule['frequent_cancellation'].items()
+                        and rule_pack['trade_position_ratio'].items() <= rule['trade_position_ratio'].items())
+            if not self._wait_update_until(cond=cond, deadline=deadline):
+                raise TqTimeoutError("设置风控规则超时请检查客户端及网络是否正常")
         return rule
 
     # ----------------------------------------------------------------------
@@ -1933,6 +1933,38 @@ def wait_update(self, deadline: Optional[float] = None, _task: Union[asyncio.Tas
                         else:  # 订阅多个合约
                             self._update_serial_multi(serial)
 
+    def _wait_update_until(self, cond: Callable[[], bool], deadline: Optional[float] = None) -> bool:
+        """
+        TqApi 内部使用，用于等待某个条件满足。持续调用 wait_update()，直到 cond() 返回 True。
+
+        Args:
+            cond (Callable[[], bool]): 条件函数
+            deadline (float): [可选]指定截止时间，自unix epoch(1970-01-01 00:00:00 GMT)以来的秒数(time.time())。默认没有超时(无限等待)
+
+        Returns:
+            bool: 当 cond() 为 True 时返回 True, 如果到截止时间 cond() 依然为 False 则返回 False
+
+        注：用于 tqsdk 内部，某些地方会用到 api.wait_update()，等待数据更新后再返回给用户，
+        * 简单调用 wait_update() 导致 api._sync_diffs 丢失变更
+        * 为了避免这种情况，内部调用 wait_update() 应该传入 _task 参数，这样 api._sync_diffs 不会丢失变更
+        """
+        if cond():
+            return True
+
+        async def _async_wait_task():
+            async with self.register_update_notify() as update_chan:
+                async for _ in update_chan:
+                    if cond():
+                        break
+
+        _task = self.create_task(_async_wait_task())
+
+        while not cond():
+            data_updated = self.wait_update(deadline=deadline, _task=_task)
+            if data_updated is False:
+                return False  # TimeoutError
+        return True
+
     # ----------------------------------------------------------------------
     def is_changing(self, obj: Any, key: Union[str, List[str], None] = None) -> bool:
         """
@@ -2226,9 +2258,9 @@ def query_graphql(self, query: str, variables: dict, query_id: Optional[str] = N
         })
         deadline = time.time() + 60
         if not self._loop.is_running():
-            while query_id not in symbols:
-                if not self.wait_update(deadline=deadline):
-                    raise TqTimeoutError("查询合约服务 %s 超时，请检查客户端及网络是否正常 %s" % (query, query_id))
+            if not self._wait_update_until(cond=lambda: query_id in symbols, deadline=deadline):
+                # 使用 _task 参数，确保不会丢掉 _sync_diffs 里的变更
+                raise TqTimeoutError("查询合约服务 %s 超时，请检查客户端及网络是否正常 %s" % (query, query_id))
             if isinstance(self._backtest, TqBacktest):
                 self._send_pack({
                     "aid": "ins_query",

tqsdk/data_series.py

@@ -68,9 +68,12 @@ def __init__(self, api, symbol_list, dur_nano, start_dt_nano, end_dt_nano, adj_t
         self._adj_type = adj_type
         self._dividend_cache = {}  # 缓存合约对应的复权系数矩阵，每个合约只计算一次
         self.df = pd.DataFrame()
-        self.is_ready = False
         DataSeries._ensure_cache_dir()  # 确认缓存文件夹存在
-        self._api.create_task(self._run())
+        self._task = self._api.create_task(self._run())
+
+    @property
+    def is_ready(self):
+        return self._task.done()
 
     async def _run(self):
         symbol = self._symbol_list[0]  # todo: 目前只处理一个合约的情况
@@ -105,7 +108,6 @@ async def _run(self):
             target_rangeset_dt = _rangeset_intersection([(self._start_dt_nano, self._end_dt_nano)], rangeset_dt)
             assert len(target_rangeset_dt) <= 1  # 用户请求应该落在一个时间段内，或者用户请求的时间段内没有任何数据
             if len(target_rangeset_dt) == 0:  # 用户请求的时间段内没有任何数据
-                self.is_ready = True
                 return
 
             # 此时用户请求时间范围，转化为 target_rangeset_dt[0]
@@ -160,8 +162,6 @@ async def _run(self):
                         adj_cols = DataSeries._get_adj_cols(symbol, self._dur_nano)
                         ge = self.df["datetime"].ge(dt)
                         self.df.loc[ge, adj_cols] = self.df.loc[ge, adj_cols] / factor
-            # 结束状态
-            self.is_ready = True
 
     async def _download_data_series(self, rangeset):
         symbol = self._symbol_list[0]

tqsdk/lib/target_pos_scheduler.py

@@ -11,6 +11,7 @@
 from tqsdk.channel import TqChan
 from tqsdk.datetime import _get_trade_timestamp
 from tqsdk.lib.target_pos_task import TargetPosTask
+from tqsdk.lib.time_table import TqTimeTable
 from tqsdk.lib.utils import _check_time_table, _get_deadline_from_interval
 from tqsdk.objs import Trade
 
@@ -93,6 +94,8 @@ def __init__(self, api: TqApi, symbol: str, time_table: DataFrame, offset_priori
             api.close()
         """
         self._api = api
+        if isinstance(time_table, TqTimeTable):
+            account = time_table.__dict__["_account"]
         self._account = api._account._check_valid(account)
 
         # 这些参数直接传给 TargetPosTask，由 TargetPosTask 来检查其合法性

tqsdk/lib/time_table.py

@@ -0,0 +1,14 @@
+#!/usr/bin/env python
+#  -*- coding: utf-8 -*-
+__author__ = 'mayanqiong'
+
+
+from pandas import DataFrame
+
+
+class TqTimeTable(DataFrame):
+
+    def __init__(self, account=None):
+        self.__dict__["_account"] = account
+        self.__dict__["_columns"] = ['interval', 'volume', 'price']
+        super(TqTimeTable, self).__init__(data=[], columns=self.__dict__["_columns"])