计算仓单“开平”字段函数

Author: gamcing

tqsdk/tafunc_tick_msg.py

@@ -0,0 +1,74 @@
+__author__ = 'y.m.wong'
+import pandas as pd
+"""
+由https://github.com/shinnytech/shinny-futures-web/blob/master/src/components/TicksList.vue改写而来
+"""
+
+def cal_pc(tick:dict) -> int:
+    """
+    计算tick前后的pc值（可能是判断多空方向）
+    :param tick: 当前tick
+    :return: pc值
+    """
+    
+    pc = 0
+    if tick['trade_ask_spread'] >= 0:
+        pc = 1
+    elif tick['trade_bid_spread'] >= 0:
+        pc = -1
+    else:
+        if tick['price_diff'] > 0:
+            pc = 1
+        elif tick['price_diff'] < 0:
+            pc = -1
+        else:
+            pc = 0
+    return pc
+
+def cal_msg(tick:dict) -> str:
+    """
+    根据tick信息计算成交性质
+    :param tick: 当前tick
+    :return: string 成交性质
+    """
+    msg = ''
+    if tick['oi_diff'] > 0 and tick['oi_diff'] == tick['vol_diff']:
+        msg = '双开'
+    elif tick['oi_diff'] < 0 and tick['oi_diff'] + tick['vol_diff'] == 0:
+        msg = '双平'
+    else:
+        if tick['pc'] == 0:
+            msg = '换手'
+        else:
+            msg = ('多' if tick['pc'] > 0 else '空') + ('开' if tick['oi_diff'] > 0 else '平' if tick['oi_diff'] < 0 else '换')
+    return msg
+
+def cal_ticks_msg(ticks: pd.Series) -> pd.DataFrame:
+    """
+        主函数：计算仓单“开平”字段
+    """
+    ticks_ex = pd.DataFrame()
+    ticks_ex['datetime'] = ticks["datetime"]
+    ticks_ex['vol_diff'] = ticks["volume"].diff()
+    ticks_ex['oi_diff'] = ticks["open_interest"].diff()
+    ticks_ex['price_diff'] = ticks["last_price"].diff()
+    ticks_ex['trade_ask_spread'] = ticks["last_price"] - ticks["ask_price1"].shift(1)
+    ticks_ex['trade_bid_spread'] = ticks["last_price"] - ticks["bid_price1"].shift(1)
+    ticks_ex['pc'] = ticks_ex.apply(lambda tick: cal_pc(dict(tick)), axis=1)
+    ticks_ex['msg'] = ticks_ex.apply(lambda tick: cal_msg(dict(tick)), axis=1)
+    return ticks_ex
+
+"""
+demo:
+from tqsdk import TqApi, TqBacktest, BacktestFinished
+from datetime import datetime
+from tqsdk.tafunc import time_to_datetime
+
+api = TqApi(backtest=TqBacktest(datetime(2020,2,11,14,59,55),datetime(2020,2,11,15,00)))
+ticks = api.get_tick_serial('DCE.j2005')
+test = pd.DataFrame()
+while True:
+    api.wait_update()
+    test = cal_ticks_ex(ticks)
+    print(test)
+"""