0.9.31 update

czsc/utils/bar_generator.py

@@ -53,6 +53,9 @@ def check_freq_and_market(time_seq: List[AnyStr], freq: Optional[AnyStr] = None)
     if freq in ['日线', '周线', '月线', '季线', '年线']:
         return freq, "默认"
 
+    if freq == '1分钟':
+        time_seq.extend(['14:57', '14:58', '14:59', '15:00'])
+
     time_seq = sorted(list(set(time_seq)))
     assert len(time_seq) >= 2, "time_seq长度必须大于等于2"
 
@@ -127,8 +130,8 @@ def freq_end_time(dt: datetime, freq: Union[Freq, AnyStr], market="A股") -> dat
         edt = dt.replace(hour=int(h), minute=int(m))
         return edt
 
-    if not ("15:00" > hm > "09:00") and market == "期货":
-        dt = next_trading_date(dt, n=1)
+    # if not ("15:00" > hm > "09:00") and market == "期货":
+    #     dt = next_trading_date(dt, n=1)
 
     return freq_end_date(dt.date(), freq)
 

examples/develop/index_composition.py

@@ -57,13 +57,15 @@ def index_composition_by_base_point(klines, weights=None, base_point=1000):
     """
     klines["dt"] = pd.to_datetime(klines["dt"])
 
-    data = []
-    for _, kline in klines.groupby("symbol"):
-        kline = kline.sort_values("dt", ascending=True).reset_index(drop=True)
-        kline["returns"] = kline["close"].pct_change()
-        kline["returns"] = kline["returns"].fillna(0)
-        data.append(kline)
-    klines = pd.concat(data, ignore_index=True)
+    if 'returns' not in klines.columns.tolist():
+        data = []
+        for _, kline in klines.groupby("symbol"):
+            kline = kline.sort_values("dt", ascending=True).reset_index(drop=True)
+            kline["returns"] = kline["close"].pct_change()
+            kline["returns"] = kline["returns"].fillna(0)
+            data.append(kline)
+        klines = pd.concat(data, ignore_index=True)
+
     returns = klines.pivot_table(index="dt", columns="symbol", values="returns")
 
     if weights is None:
@@ -93,6 +95,8 @@ def test_index_composition_by_base_point():
     for symbol in symbols:
         bars = get_raw_bars(symbol=symbol, freq="1分钟", sdt="20210104 09:31", edt="20210107 13:10")
         df = pd.DataFrame(bars)
+        df['returns'] = df['close'].pct_change()
+        df = df.dropna(subset=['returns'])
         df.drop(columns=["freq", "cache", 'id'], inplace=True)
         data.append(df)
     klines = pd.concat(data, ignore_index=True)
@@ -106,6 +110,12 @@ def test_index_composition_by_base_point():
     index2 = index_composition_by_base_point(klines=klines.copy(), weights=weights)
     assert int(index2['vol'].iloc[-1]) == 200930300
 
+    weights = pd.DataFrame({"dt": ['2020-01-03 10:10:00', '2021-01-04 10:54:00', '2021-01-05 12:10:00']})
+    for symbol in symbols:
+        weights[symbol] = [random.uniform(0.1, 0.3) for _ in range(3)]
+    index2 = index_composition_by_base_point(klines=klines.copy(), weights=weights)
+    assert int(index2['vol'].iloc[-1]) == 200930300
+
     try:
         weights = pd.DataFrame({"dt": ['2021-01-04 10:10:00', '2021-01-04 10:54:00', '2021-01-05 12:10:00']})
         for symbol in symbols:

examples/test_offline/test_resample_bar.py

@@ -13,14 +13,16 @@ def test_read():
 
 def test_check_freq_and_market():
     from czsc.utils.bar_generator import check_freq_and_market
-    gruop_name = "期货主力"
-    # gruop_name = "中证500成分股"
+    # gruop_name = "期货主力"
+    gruop_name = "中证500成分股"
     files = Path(fr"D:\CZSC投研数据\{gruop_name}").glob("*.parquet")
     for file in files:
         df = pd.read_parquet(file)
         time_seq = sorted(list({x.strftime("%H:%M") for x in df['dt']}))
-        x_freq, market = check_freq_and_market(time_seq=time_seq)
+        x_freq, market = check_freq_and_market(time_seq=time_seq, freq='1分钟')
         print(file.stem, x_freq, market)
+        if market == "默认":
+            print(time_seq)
 
 
 def get_future_times():

test/test_bar_generator.py

@@ -15,8 +15,6 @@ def test_check_freq_and_market():
     assert check_freq_and_market(time_seq) == ('120分钟', '期货')
 
     time_seq = [
-        '09:25',
-        '09:30',
         '09:31',
         '09:32',
         '09:33',
@@ -257,7 +255,7 @@ def test_check_freq_and_market():
         '14:58',
         '15:00',
     ]
-    assert check_freq_and_market(time_seq) == ('1分钟', 'A股')
+    assert check_freq_and_market(time_seq, freq='1分钟') == ('1分钟', 'A股')
 
     for key, values in freq_market_times.items():
         assert check_freq_and_market(values) == (key.split("_")[0], key.split("_")[1])