fix(data): dbbardata 日线双行根治(统一纯日期+helper,留ROWID max去重)
根因: dbbardata UNIQUE(symbol,exchange,interval,datetime) 按字符串字面比较, 多写入路径混用 'YYYY-MM-DD' 与 'YYYY-MM-DD 00:00:00' -> 同一交易日双行, INSERT OR REPLACE 不去重 -> 回测交易日翻倍/pivot duplicate/信号异常。 方案A (统一纯日期, 详见 Main Agent 诊断): - 新增 scripts/data_platform/dbbardata_utils.py: normalize_daily_dt(s) 取前 10 字符, None/短串安全 - 4 个日线写入脚本写入前调 helper: - bs_eod.py (sanguo-bs-eod 个股日线 baostock) - migrate_daily_baostock.py (历史迁移) - xt_eod.py (sanguo-xt-eod ETF/基金 xtata) - import_vnpy_daily_fast.py (NAS 日线 parquet 导入, 加防御) - TDD: tests/data_platform/test_dbbardata_utils.py 9 cases 全过 - 回归: tests/data_platform + tests/portfolio 199 passed 12 skipped peewee DateTimeField formats 含 '%Y-%m-%d' (阶段0 VPS 实测确认), 读纯日期不崩, 方案A 前提成立。 15min 干净, 不动 (分钟必须带时分)。只改日线 interval='d'。 数据层根治, 不在 provider 适配兜底 (用户铁律)。
This commit is contained in:
@@ -31,6 +31,8 @@ except (AttributeError, ValueError):
|
||||
import baostock as bs
|
||||
import pandas as pd
|
||||
|
||||
from dbbardata_utils import normalize_daily_dt
|
||||
|
||||
BASE = Path(r"C:\sanguo_vnpy_v2")
|
||||
DB = BASE / "data" / "quant_trading.db"
|
||||
VAL_DIR = BASE / "data" / "valuation_baostock"
|
||||
@@ -121,10 +123,10 @@ def upsert_daily(conn, code, prefix, rows):
|
||||
"turn", "pctChg", "peTTM", "psTTM", "pcfNcfTTM", "pbMRQ"]:
|
||||
df[c] = pd.to_numeric(df[c], errors="coerce")
|
||||
exc = EXC_MAP[prefix]
|
||||
# OHLCV -> dbbardata('d')
|
||||
# OHLCV -> dbbardata('d') — datetime 归一纯日期 (dbbardata 双行根治方案A)
|
||||
db = pd.DataFrame({
|
||||
"symbol": code, "exchange": exc,
|
||||
"datetime": df["date"].astype(str) + " 00:00:00",
|
||||
"datetime": df["date"].astype(str).map(normalize_daily_dt),
|
||||
"interval": "d", "volume": df["volume"], "turnover": df["amount"],
|
||||
"open_interest": 0.0,
|
||||
"open_price": df["open"], "high_price": df["high"],
|
||||
|
||||
@@ -0,0 +1,28 @@
|
||||
# -*- coding: utf-8 -*-
|
||||
"""dbbardata 日线 datetime 归一化 helper (双行根治方案A)。
|
||||
|
||||
背景:
|
||||
dbbardata (DbBarData) UNIQUE(symbol,exchange,interval,datetime) 按字符串字面比较。
|
||||
历史多写入路径混用 "YYYY-MM-DD" 与 "YYYY-MM-DD 00:00:00" 两种格式,
|
||||
同一交易日 INSERT OR REPLACE 不去重 -> 双行, 致回测翻倍/pivot duplicate。
|
||||
|
||||
归一规则:
|
||||
日线 (interval='d') 写入前必调 normalize_daily_dt, 统一为纯日期 'YYYY-MM-DD'。
|
||||
分钟 (interval='15m' 等) 不要调 — 分钟必须带时分。
|
||||
|
||||
入口: 所有写 dbbardata interval='d' 的脚本 import 并在 datetime 列写入前调用。
|
||||
"""
|
||||
from typing import Optional
|
||||
|
||||
|
||||
def normalize_daily_dt(s: Optional[str]) -> Optional[str]:
|
||||
"""日线 datetime 归一为纯日期 'YYYY-MM-DD' (取前 10 字符)。
|
||||
|
||||
- 纯日期 'YYYY-MM-DD' -> no-op
|
||||
- 'YYYY-MM-DD 00:00:00' / 'YYYY-MM-DD HH:MM:SS' -> 截断为 'YYYY-MM-DD'
|
||||
- None -> None (保留语义不崩溃)
|
||||
- 短串/空串 -> 原样返回(前 10 字符)不崩溃
|
||||
"""
|
||||
if s is None:
|
||||
return None
|
||||
return str(s)[:10]
|
||||
@@ -12,6 +12,8 @@ import sys
|
||||
import time
|
||||
from pathlib import Path
|
||||
|
||||
from dbbardata_utils import normalize_daily_dt
|
||||
|
||||
DB_PATH = os.environ.get('VNPY_DB_PATH', '/tmp/quant_trading_import.db')
|
||||
DAILY_DIR = os.environ.get('DAILY_DIR', '/Volumes/stock/A股数据/日线数据/daily/')
|
||||
|
||||
@@ -57,8 +59,8 @@ def import_year(conn, year):
|
||||
|
||||
combined = pd.concat(all_dfs, ignore_index=True)
|
||||
|
||||
# Vectorized conversion
|
||||
combined['datetime'] = combined['date'].astype(str)
|
||||
# Vectorized conversion — datetime 归一纯日期 (dbbardata 双行根治方案A, 防御未来混入时分)
|
||||
combined['datetime'] = combined['date'].astype(str).map(normalize_daily_dt)
|
||||
combined['interval'] = 'd'
|
||||
combined['open_interest'] = 0.0
|
||||
combined = combined.rename(columns={
|
||||
|
||||
@@ -13,6 +13,8 @@ from pathlib import Path
|
||||
|
||||
import pandas as pd
|
||||
|
||||
from dbbardata_utils import normalize_daily_dt
|
||||
|
||||
DB = Path(r"C:\sanguo_vnpy_v2\data\quant_trading.db")
|
||||
VAL_DIR = Path(r"C:\sanguo_vnpy_v2\data\valuation_baostock")
|
||||
EXC_MAP = {"SH": "SSE", "SZ": "SZSE"}
|
||||
@@ -52,7 +54,8 @@ for chunk in pd.read_sql("SELECT * FROM daily_baostock_full", c, chunksize=20000
|
||||
odb = pd.DataFrame({
|
||||
"symbol": chunk["symbol"].values,
|
||||
"exchange": chunk["exchange"].map(EXC_MAP).values,
|
||||
"datetime": (chunk["date"].astype(str) + " 00:00:00").values,
|
||||
# datetime 归一纯日期 (dbbardata 双行根治方案A)
|
||||
"datetime": chunk["date"].astype(str).map(normalize_daily_dt).values,
|
||||
"interval": "d",
|
||||
"volume": chunk["volume"].values,
|
||||
"turnover": chunk["amount"].values,
|
||||
|
||||
@@ -21,6 +21,8 @@ import time
|
||||
from xtquant import xtdata as xd
|
||||
import pandas as pd
|
||||
|
||||
from dbbardata_utils import normalize_daily_dt
|
||||
|
||||
DB = r"C:\sanguo_vnpy_v2\data\quant_trading.db"
|
||||
LOOKBACK = int(__import__("os").environ.get("LOOKBACK_DAYS", "30"))
|
||||
T0 = time.time()
|
||||
@@ -88,8 +90,10 @@ def main():
|
||||
db = pd.DataFrame({
|
||||
"symbol": sym,
|
||||
"exchange": exc_of(sym),
|
||||
"datetime": [str(idx)[:4]+"-"+str(idx)[4:6]+"-"+str(idx)[6:8]+" 00:00:00"
|
||||
for idx in df.index],
|
||||
# datetime 归一纯日期 (dbbardata 双行根治方案A)
|
||||
"datetime": [normalize_daily_dt(
|
||||
f"{str(idx)[:4]}-{str(idx)[4:6]}-{str(idx)[6:8]}")
|
||||
for idx in df.index],
|
||||
"interval": "d",
|
||||
"volume": (df["volume"].astype(float).values * 100),
|
||||
"turnover": df["amount"].astype(float).values,
|
||||
|
||||
@@ -0,0 +1,46 @@
|
||||
# -*- coding: utf-8 -*-
|
||||
"""TDD for dbbardata daily datetime normalizer (dbbardata 双行根治方案A)."""
|
||||
import pytest
|
||||
|
||||
from scripts.data_platform.dbbardata_utils import normalize_daily_dt
|
||||
|
||||
|
||||
@pytest.mark.parametrize("s,expected", [
|
||||
("2026-07-20", "2026-07-20"),
|
||||
("2010-01-04", "2010-01-04"),
|
||||
])
|
||||
def test_pure_date_noop(s, expected):
|
||||
# 纯日期应原样返回
|
||||
assert normalize_daily_dt(s) == expected
|
||||
|
||||
|
||||
@pytest.mark.parametrize("s", [
|
||||
"2026-07-20 00:00:00",
|
||||
"2026-07-20 00:00:00.000000",
|
||||
])
|
||||
def test_truncate_midnight(s):
|
||||
# 带时间(00:00:00 历史迁移路径) 截断为纯日期
|
||||
assert normalize_daily_dt(s) == "2026-07-20"
|
||||
|
||||
|
||||
@pytest.mark.parametrize("s", [
|
||||
"2026-07-20 15:00:00",
|
||||
"2026-07-20 09:30:01",
|
||||
])
|
||||
def test_truncate_real_time(s):
|
||||
# 带真实时分(异常情况) 也截断为纯日期 — 日线无时分语义
|
||||
assert normalize_daily_dt(s) == "2026-07-20"
|
||||
|
||||
|
||||
def test_none_safe():
|
||||
# None 不应崩溃 — 返回 None 保留语义
|
||||
assert normalize_daily_dt(None) is None
|
||||
|
||||
|
||||
def test_empty_string_safe():
|
||||
assert normalize_daily_dt("") == ""
|
||||
|
||||
|
||||
def test_short_string_safe():
|
||||
# 短于 10 字符的异常输入不崩溃,返回本身(取前 10)
|
||||
assert normalize_daily_dt("2026") == "2026"
|
||||
Reference in New Issue
Block a user