aszerW
4f9e0231bd
- yfinance_source.py: 用 tz_localize(None) 替代 pd.to_datetime(utc=True)， 避免亚洲/欧洲市场因UTC转换导致日期回退一天（如日经225 5/25→5/24） - yfinance_source.py: 新增 _normalize_index() 静态方法统一处理时区剥除 - yfinance_source.py: fetch() 增加 close=NaN 行过滤（yfinance未收盘日返回不完整数据） - flask_api_source.py: 客户端同步增加 close=NaN 过滤防御 验证结果：N225 5/25-6/3 返回7个交易日数据，日期无偏移
240 lines
8.6 KiB
Python
240 lines
8.6 KiB
Python
"""
|
||
YFinance数据源
|
||
|
||
获取港股、美股数据（通过SSH隧道）
|
||
"""
|
||
|
||
import os
|
||
import time
|
||
from typing import Optional
|
||
from datetime import datetime, timedelta
|
||
import pandas as pd
|
||
import urllib3
|
||
|
||
# 禁用SSL警告
|
||
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)
|
||
|
||
|
||
class YFinanceSource:
|
||
"""YFinance数据源"""
|
||
|
||
# 代码映射（项目代码 -> YFinance格式）
|
||
CODE_MAP = {
|
||
# 港股
|
||
"HSTECH.HK": "3033.HK", # 恒生科技指数
|
||
"HSI": "^HSI", # 恒生指数
|
||
# 美股指数
|
||
"NDX": "^NDX", # 纳斯达克100
|
||
"SPX": "^GSPC", # 标普500
|
||
"DJI": "^DJI", # 道琼斯
|
||
# 日本/欧洲
|
||
"N225": "^N225", # 日经225
|
||
"GDAXI": "^GDAXI", # 德国DAX
|
||
# 商品
|
||
"CL.NYM": "CL=F", # WTI原油期货
|
||
}
|
||
|
||
def __init__(self, use_ssh_tunnel: bool = False):
|
||
"""
|
||
初始化YFinance数据源
|
||
|
||
Args:
|
||
use_ssh_tunnel: 是否使用SSH隧道（需先启动SSHTunnelManager）
|
||
"""
|
||
self.use_ssh_tunnel = use_ssh_tunnel
|
||
self._delay = 0.5 # 请求延迟（避免限流）
|
||
|
||
@staticmethod
|
||
def _normalize_index(index: pd.DatetimeIndex) -> pd.DatetimeIndex:
|
||
"""
|
||
标准化日期索引，保留交易所本地日期
|
||
|
||
yfinance 返回的时间戳带有交易所本地时区：
|
||
- 美股: 00:00-04:00 (US/Eastern) → 剥 tz 后 00:00，日期不变
|
||
- 日本: 00:00+09:00 (Asia/Tokyo) → 剥 tz 后 00:00，日期不变
|
||
- 欧洲: 00:00+02:00 (CET) → 剥 tz 后 00:00，日期不变
|
||
|
||
关键：直接 tz_localize(None) 剥除时区，不做 UTC 转换。
|
||
错误示范：pd.to_datetime(idx, utc=True) 会先把日本 00:00+09:00 转成
|
||
前一天 15:00 UTC，导致日期回退一天。
|
||
"""
|
||
if index.tz is not None:
|
||
# tz_localize(None) 直接剥除时区，保留本地时间部分
|
||
index = index.tz_localize(None)
|
||
return index.normalize()
|
||
|
||
def fetch(self, code: str, start_date: str, end_date: str, adj: str = 'raw') -> Optional[pd.DataFrame]:
|
||
"""
|
||
获取数据（支持 adj 参数）
|
||
|
||
Args:
|
||
code: 代码（如 'NDX', 'N225', 'HSI', 'AAPL'）
|
||
start_date: 开始日期 'YYYY-MM-DD'
|
||
end_date: 结束日期 'YYYY-MM-DD'
|
||
adj: 复权类型 'raw'(原始) / 'qfq'(前复权) / 'hfq'(后复权)，默认 'raw'
|
||
|
||
Returns:
|
||
DataFrame with columns: date, open, high, low, close, volume
|
||
股票元信息存储在 df.attrs['info'] 中
|
||
adj='qfq/hfq' 时 df.attrs['adj'] 会标记复权类型
|
||
"""
|
||
# 校验 adj 参数
|
||
if adj not in ['raw', 'qfq', 'hfq']:
|
||
raise ValueError(f"adj 参数必须是 'raw', 'qfq' 或 'hfq'，当前: {adj}")
|
||
|
||
# 复权数据：调用 fetch_adj
|
||
if adj in ['qfq', 'hfq']:
|
||
return self.fetch_adj(code, start_date, end_date, adj)
|
||
|
||
# 原始数据：以下为原有逻辑
|
||
import yfinance as yf
|
||
|
||
# 添加延迟避免限流
|
||
time.sleep(self._delay)
|
||
|
||
# 转换代码格式
|
||
yf_code = self.CODE_MAP.get(code, code)
|
||
|
||
try:
|
||
ticker = yf.Ticker(yf_code)
|
||
|
||
# 获取股票信息（仅对股票/ETF有效，指数可能没有）
|
||
stock_info = {}
|
||
try:
|
||
stock_info = ticker.info or {}
|
||
except Exception:
|
||
pass # 指数可能没有info
|
||
|
||
# end_date 需要加一天（yfinance的end是排他的）
|
||
end_dt = datetime.strptime(end_date, "%Y-%m-%d") + timedelta(days=1)
|
||
|
||
# auto_adjust=False 获取不复权价格
|
||
df = ticker.history(
|
||
start=start_date,
|
||
end=end_dt.strftime("%Y-%m-%d"),
|
||
auto_adjust=False
|
||
)
|
||
|
||
if df is None or len(df) == 0:
|
||
return None
|
||
|
||
# 标准化列名
|
||
df = df.rename(columns={
|
||
"Open": "open",
|
||
"High": "high",
|
||
"Low": "low",
|
||
"Close": "close",
|
||
"Volume": "volume",
|
||
})
|
||
|
||
# 确保索引是日期格式（保留交易所本地日期，避免 UTC 转换导致跨日偏移）
|
||
df.index = self._normalize_index(df.index)
|
||
df.index.name = "date"
|
||
|
||
# 过滤 yfinance 返回的不完整数据（未收盘日 close=NaN, volume=0）
|
||
nan_mask = df['close'].isna()
|
||
if nan_mask.any():
|
||
df = df[~nan_mask]
|
||
if len(df) == 0:
|
||
return None
|
||
|
||
# 添加代码列
|
||
df["code"] = code
|
||
|
||
# 将股票信息存储到 DataFrame.attrs 中（最外层结构）
|
||
df.attrs['info'] = stock_info
|
||
df.attrs['code'] = code
|
||
df.attrs['adj'] = 'raw'
|
||
|
||
return df[['code', 'open', 'high', 'low', 'close', 'volume']]
|
||
|
||
except Exception as e:
|
||
print(f"YFinance下载 {code} ({yf_code}) 失败: {e}")
|
||
return None
|
||
|
||
def fetch_adj(self, code: str, start_date: str, end_date: str, adj: str = 'qfq') -> Optional[pd.DataFrame]:
|
||
"""
|
||
获取复权价格数据
|
||
|
||
统一 adj 参数设计：
|
||
- 'qfq': 前复权 → yfinance auto_adjust=True (当前价不变)
|
||
- 'hfq': 后复权 → yfinance back_adjust=True (历史价不变)
|
||
|
||
Args:
|
||
code: 代码（如 'AAPL', 'TSLA', 'QQQ', '00700.HK'）
|
||
start_date: 开始日期 'YYYY-MM-DD'
|
||
end_date: 结束日期 'YYYY-MM-DD'
|
||
adj: 复权类型 'qfq'(前复权) 或 'hfq'(后复权)，默认 'qfq'
|
||
|
||
Returns:
|
||
DataFrame with columns: date, code, open, high, low, close, volume (复权后)
|
||
"""
|
||
import yfinance as yf
|
||
|
||
if adj not in ['qfq', 'hfq']:
|
||
raise ValueError(f"adj 参数必须是 'qfq' 或 'hfq'，当前: {adj}")
|
||
|
||
# 添加延迟避免限流
|
||
time.sleep(self._delay)
|
||
|
||
# 转换代码格式
|
||
yf_code = self.CODE_MAP.get(code, code)
|
||
|
||
# adj 参数映射到 yfinance 参数
|
||
# qfq(前复权) = auto_adjust=True, back_adjust=False (当前价不变)
|
||
# hfq(后复权) = auto_adjust=False, back_adjust=True (历史价不变)
|
||
adjust_params = {
|
||
'qfq': {'auto_adjust': True, 'back_adjust': False},
|
||
'hfq': {'auto_adjust': False, 'back_adjust': True},
|
||
}
|
||
|
||
try:
|
||
ticker = yf.Ticker(yf_code)
|
||
|
||
# end_date 需要加一天（yfinance的end是排他的）
|
||
end_dt = datetime.strptime(end_date, "%Y-%m-%d") + timedelta(days=1)
|
||
|
||
# 根据 adj 参数设置复权方式
|
||
params = adjust_params[adj]
|
||
df = ticker.history(
|
||
start=start_date,
|
||
end=end_dt.strftime("%Y-%m-%d"),
|
||
auto_adjust=params['auto_adjust'],
|
||
back_adjust=params['back_adjust']
|
||
)
|
||
|
||
if df is None or len(df) == 0:
|
||
return None
|
||
|
||
# 标准化列名
|
||
df = df.rename(columns={
|
||
"Open": "open",
|
||
"High": "high",
|
||
"Low": "low",
|
||
"Close": "close",
|
||
"Volume": "volume",
|
||
})
|
||
|
||
# 确保索引是日期格式（保留交易所本地日期，避免 UTC 转换导致跨日偏移）
|
||
df.index = self._normalize_index(df.index)
|
||
df.index.name = "date"
|
||
|
||
# 添加代码列和标记
|
||
df["code"] = code
|
||
df.attrs['code'] = code
|
||
df.attrs['adj'] = adj
|
||
|
||
return df[['code', 'open', 'high', 'low', 'close', 'volume']]
|
||
|
||
except Exception as e:
|
||
print(f"YFinance下载复权数据 {code} ({yf_code}) adj={adj} 失败: {e}")
|
||
return None
|
||
|
||
def is_yfinance_code(self, code: str) -> bool:
|
||
"""判断是否需要YFinance获取"""
|
||
# 非A股代码
|
||
china_suffixes = ['.SH', '.SZ', '.SS', '.CSI']
|
||
futures_suffixes = ['.SHF', '.NYM', '.DCE', '.CZC']
|
||
|
||
# A股或期货用Tushare，其他用YFinance
|
||
return not any(code.endswith(s) for s in china_suffixes + futures_suffixes) |