aszerW
02dbc7bd7d
TushareSource.fetch() 和 YFinanceSource.fetch() 新增 adj 参数支持 raw/qfq/hfq - TushareSource.fetch(adj='raw'): 内部路由到 fetch_index/fetch_stock_adj/fetch_etf_adj - YFinanceSource.fetch(adj='raw'): 内部路由到 fetch_adj() 或原始逻辑 - 添加 is_china_stock() 和 _is_etf_code() 方法用于资产类型判断
453 lines
16 KiB
Python
453 lines
16 KiB
Python
"""
|
||
Tushare数据源
|
||
|
||
获取A股指数、ETF、期货数据
|
||
"""
|
||
|
||
import os
|
||
from typing import Optional
|
||
from datetime import datetime
|
||
import pandas as pd
|
||
|
||
|
||
class TushareSource:
|
||
"""Tushare数据源"""
|
||
|
||
def __init__(self, token: Optional[str] = None):
|
||
"""
|
||
初始化Tushare数据源
|
||
|
||
Args:
|
||
token: Tushare Token（可选，默认从环境变量读取）
|
||
"""
|
||
self._token = token or os.getenv("TUSHARE_TOKEN")
|
||
if not self._token:
|
||
raise ValueError("请设置环境变量 TUSHARE_TOKEN")
|
||
|
||
def _get_pro_api(self):
|
||
"""获取Tushare Pro API"""
|
||
import tushare as ts
|
||
return ts.pro_api(self._token)
|
||
|
||
def fetch_index(self, code: str, start_date: str, end_date: str) -> Optional[pd.DataFrame]:
|
||
"""
|
||
获取A股指数数据
|
||
|
||
Args:
|
||
code: 指数代码，如 '000300.SH', '399006.SZ', 'H30269.CSI'
|
||
start_date: 开始日期 'YYYY-MM-DD'
|
||
end_date: 结束日期 'YYYY-MM-DD'
|
||
|
||
Returns:
|
||
DataFrame with columns: date, open, high, low, close, volume
|
||
"""
|
||
try:
|
||
pro = self._get_pro_api()
|
||
|
||
# 转换代码格式 (.SS -> .SH)
|
||
ts_code = code.replace(".SS", ".SH")
|
||
|
||
df = pro.index_daily(
|
||
ts_code=ts_code,
|
||
start_date=start_date.replace("-", ""),
|
||
end_date=end_date.replace("-", "")
|
||
)
|
||
|
||
if df is None or len(df) == 0:
|
||
return None
|
||
|
||
# 标准化列名
|
||
df = df.rename(columns={
|
||
"trade_date": "date",
|
||
"vol": "volume",
|
||
})
|
||
|
||
# 转换日期格式
|
||
df["date"] = pd.to_datetime(df["date"])
|
||
df = df.set_index("date")
|
||
df = df.sort_index()
|
||
df["code"] = code
|
||
|
||
return df[['code', 'open', 'high', 'low', 'close', 'volume']]
|
||
|
||
except Exception as e:
|
||
print(f"Tushare下载指数 {code} 失败: {e}")
|
||
return None
|
||
|
||
def fetch_futures(self, code: str, start_date: str, end_date: str) -> Optional[pd.DataFrame]:
|
||
"""
|
||
获取期货数据
|
||
|
||
Args:
|
||
code: 期货代码，如 'AU.SHF', 'CU.SHF'
|
||
start_date: 开始日期
|
||
end_date: 结束日期
|
||
"""
|
||
try:
|
||
pro = self._get_pro_api()
|
||
|
||
# 使用 fut_daily 接口
|
||
df = pro.fut_daily(
|
||
ts_code=code,
|
||
start_date=start_date.replace("-", ""),
|
||
end_date=end_date.replace("-", "")
|
||
)
|
||
|
||
if df is None or len(df) == 0:
|
||
return None
|
||
|
||
# 标准化列名
|
||
df = df.rename(columns={
|
||
"trade_date": "date",
|
||
"vol": "volume",
|
||
})
|
||
|
||
df["date"] = pd.to_datetime(df["date"])
|
||
df = df.set_index("date")
|
||
df = df.sort_index()
|
||
df["code"] = code
|
||
|
||
return df[['code', 'open', 'high', 'low', 'close', 'volume']]
|
||
|
||
except Exception as e:
|
||
print(f"Tushare下载期货 {code} 失败: {e}")
|
||
return None
|
||
|
||
def fetch_etf(self, code: str, start_date: str, end_date: str) -> Optional[pd.DataFrame]:
|
||
"""
|
||
获取ETF价格数据
|
||
|
||
Args:
|
||
code: ETF代码，如 '159915.SZ', '518880.SH'
|
||
"""
|
||
try:
|
||
pro = self._get_pro_api()
|
||
|
||
ts_code = code.replace(".SS", ".SH")
|
||
|
||
df = pro.fund_daily(
|
||
ts_code=ts_code,
|
||
start_date=start_date.replace("-", ""),
|
||
end_date=end_date.replace("-", "")
|
||
)
|
||
|
||
if df is None or len(df) == 0:
|
||
return None
|
||
|
||
df = df.rename(columns={
|
||
"trade_date": "date",
|
||
"vol": "volume",
|
||
})
|
||
|
||
df["date"] = pd.to_datetime(df["date"])
|
||
df = df.set_index("date")
|
||
df = df.sort_index()
|
||
df["code"] = code
|
||
|
||
return df[['code', 'open', 'high', 'low', 'close', 'volume']]
|
||
|
||
except Exception as e:
|
||
print(f"Tushare下载ETF {code} 失败: {e}")
|
||
return None
|
||
|
||
def fetch_etf_nav(self, code: str, start_date: str, end_date: str) -> Optional[pd.DataFrame]:
|
||
"""
|
||
获取ETF净值数据
|
||
|
||
Args:
|
||
code: ETF代码
|
||
"""
|
||
try:
|
||
pro = self._get_pro_api()
|
||
|
||
ts_code = code.replace(".SS", ".SH")
|
||
|
||
df = pro.fund_nav(
|
||
ts_code=ts_code,
|
||
start_date=start_date.replace("-", ""),
|
||
end_date=end_date.replace("-", "")
|
||
)
|
||
|
||
if df is None or len(df) == 0:
|
||
return None
|
||
|
||
df = df.rename(columns={
|
||
"nav_date": "date",
|
||
"unit_nav": "nav",
|
||
})
|
||
|
||
df["date"] = pd.to_datetime(df["date"])
|
||
df = df.set_index("date")
|
||
df = df.sort_index()
|
||
df["code"] = code
|
||
|
||
return df[['code', 'nav']]
|
||
|
||
except Exception as e:
|
||
print(f"Tushare下载ETF净值 {code} 失败: {e}")
|
||
return None
|
||
|
||
def is_china_index(self, code: str) -> bool:
|
||
"""判断是否为A股指数"""
|
||
return code.endswith(".SH") or code.endswith(".SZ") or code.endswith(".SS") or code.endswith(".CSI")
|
||
|
||
def is_futures(self, code: str) -> bool:
|
||
"""判断是否为中国期货（仅支持上期所、大商所、郑商所）"""
|
||
# 只支持中国交易所期货（.SHF上期所、.DCE大商所、.CZC郑商所）
|
||
# NYMEX (.NYM) 和 ICE (.ICE) 走 YFinance
|
||
return ".SHF" in code or ".DCE" in code or ".CZC" in code
|
||
|
||
def is_china_stock(self, code: str) -> bool:
|
||
"""判断是否为A股股票（6位数字代码 + .SZ/.SH/.SS）"""
|
||
# 股票代码：000001.SZ, 600000.SH 等
|
||
# 区分指数：指数代码通常是 000xxx.SH, 399xxx.SZ, H30xxx.CSI
|
||
# 股票代码通常是 00xxxx.SZ, 30xxxx.SZ, 60xxxx.SH, 000xxx.SH（部分）
|
||
import re
|
||
# 股票代码模式：6位数字 + .SZ/.SH/.SS
|
||
# 排除指数：000xxx.SH (指数), 399xxx.SZ (指数), Hxxxxx.CSI (指数)
|
||
if not re.match(r'^\d{6}\.(SZ|SH|SS)$', code):
|
||
return False
|
||
# 000xxx.SH 可能是指数也可能是股票，需要更细致判断
|
||
# 简化处理：000/001/002/003 开头 + .SZ 是股票，600/601/603 开头 + .SH 是股票
|
||
prefix = code[:3]
|
||
suffix = code.split('.')[1]
|
||
if suffix == 'SZ' and prefix in ['000', '001', '002', '003', '300']:
|
||
return True
|
||
if suffix == 'SH' and prefix in ['600', '601', '603', '605', '688']:
|
||
return True
|
||
return False
|
||
|
||
def fetch(self, code: str, start_date: str, end_date: str, adj: str = 'raw') -> Optional[pd.DataFrame]:
|
||
"""
|
||
通用数据获取（自动判断类型，支持 adj 参数）
|
||
|
||
Args:
|
||
code: 代码
|
||
start_date: 开始日期
|
||
end_date: 结束日期
|
||
adj: 复权类型 'raw'(原始) / 'qfq'(前复权) / 'hfq'(后复权)，默认 'raw'
|
||
|
||
Returns:
|
||
DataFrame with columns: date, open, high, low, close, volume
|
||
adj='hfq' 时 A股 ETF 会额外返回 adj_factor, close_hfq
|
||
"""
|
||
# 校验 adj 参数
|
||
if adj not in ['raw', 'qfq', 'hfq']:
|
||
raise ValueError(f"adj 参数必须是 'raw', 'qfq' 或 'hfq'，当前: {adj}")
|
||
|
||
# 原始数据
|
||
if adj == 'raw':
|
||
if self.is_china_index(code):
|
||
return self.fetch_index(code, start_date, end_date)
|
||
elif self.is_futures(code):
|
||
return self.fetch_futures(code, start_date, end_date)
|
||
elif self.is_china_stock(code):
|
||
return self.fetch_stock_adj(code, start_date, end_date, adj='raw')
|
||
else:
|
||
return None
|
||
|
||
# 复权数据
|
||
if adj in ['qfq', 'hfq']:
|
||
# A股股票复权
|
||
if self.is_china_stock(code):
|
||
return self.fetch_stock_adj(code, start_date, end_date, adj)
|
||
# A股 ETF 仅支持 hfq
|
||
elif self._is_etf_code(code):
|
||
if adj == 'hfq':
|
||
return self.fetch_etf_adj(code, start_date, end_date)
|
||
else:
|
||
raise ValueError(f"ETF 仅支持 adj='hfq'（后复权），当前: {adj}")
|
||
else:
|
||
# 指数/期货不支持复权
|
||
raise ValueError(f"指数/期货不支持复权，adj='{adj}' 仅适用于股票/ETF")
|
||
|
||
def _is_etf_code(self, code: str) -> bool:
|
||
"""判断是否为ETF代码"""
|
||
# ETF代码：51xxxx.SH, 52xxxx.SH, 15xxxx.SZ, 16xxxx.SZ
|
||
import re
|
||
if not re.match(r'^\d{6}\.(SZ|SH)$', code):
|
||
return False
|
||
prefix = code[:2]
|
||
return prefix in ['51', '52', '15', '16']
|
||
|
||
def fetch_etf_adj(self, code: str, start_date: str, end_date: str) -> Optional[pd.DataFrame]:
|
||
"""
|
||
获取 ETF 后复权价格数据
|
||
|
||
通过 fund_daily + fund_adj 手动计算后复权价格，消除份额折算（拆分）对收益率的影响。
|
||
fund_adj 单次限 2000 条，按 5 年分段请求再拼接。
|
||
|
||
Args:
|
||
code: ETF代码，如 '159915.SZ', '518880.SH'
|
||
start_date: 开始日期 'YYYY-MM-DD'
|
||
end_date: 结束日期 'YYYY-MM-DD'
|
||
|
||
Returns:
|
||
DataFrame with columns: date, open, close, adj_factor, close_hfq
|
||
"""
|
||
try:
|
||
pro = self._get_pro_api()
|
||
ts_code = code.replace('.SS', '.SH')
|
||
|
||
# 获取 fund_daily 数据
|
||
df_daily = pro.fund_daily(
|
||
ts_code=ts_code,
|
||
start_date=start_date.replace('-', ''),
|
||
end_date=end_date.replace('-', '')
|
||
)
|
||
|
||
if df_daily is None or len(df_daily) == 0:
|
||
return None
|
||
|
||
# 获取 fund_adj 数据（分段请求，单次限2000条）
|
||
# 按5年分段
|
||
start_dt = datetime.strptime(start_date, '%Y-%m-%d')
|
||
end_dt = datetime.strptime(end_date, '%Y-%m-%d')
|
||
|
||
adj_chunks = []
|
||
chunk_start = start_dt
|
||
while chunk_start < end_dt:
|
||
chunk_end = min(chunk_start.replace(year=chunk_start.year + 5), end_dt)
|
||
chunk_start_str = chunk_start.strftime('%Y%m%d')
|
||
chunk_end_str = chunk_end.strftime('%Y%m%d')
|
||
|
||
df_adj_chunk = pro.fund_adj(
|
||
ts_code=ts_code,
|
||
start_date=chunk_start_str,
|
||
end_date=chunk_end_str
|
||
)
|
||
|
||
if df_adj_chunk is not None and len(df_adj_chunk) > 0:
|
||
adj_chunks.append(df_adj_chunk)
|
||
|
||
chunk_start = chunk_end
|
||
|
||
if not adj_chunks:
|
||
# 无复权因子，返回原始数据
|
||
df = df_daily.rename(columns={'trade_date': 'date', 'vol': 'volume'})
|
||
df['date'] = pd.to_datetime(df['date'])
|
||
df = df.set_index('date').sort_index()
|
||
df['adj_factor'] = 1.0
|
||
df['close_hfq'] = df['close']
|
||
df['code'] = code
|
||
return df[['code', 'open', 'close', 'adj_factor', 'close_hfq']]
|
||
|
||
# 合并所有复权因子
|
||
df_adj = pd.concat(adj_chunks, ignore_index=True)
|
||
df_adj = df_adj.rename(columns={'trade_date': 'date'})
|
||
df_adj['date'] = pd.to_datetime(df_adj['date'])
|
||
df_adj = df_adj.set_index('date').sort_index()
|
||
|
||
# 合并 daily 和 adj
|
||
df_daily = df_daily.rename(columns={'trade_date': 'date', 'vol': 'volume'})
|
||
df_daily['date'] = pd.to_datetime(df_daily['date'])
|
||
df_daily = df_daily.set_index('date').sort_index()
|
||
|
||
# 复权因子对齐（用最新值）
|
||
df_adj_aligned = df_adj.reindex(df_daily.index, method='ffill')
|
||
df_adj_aligned['adj_factor'] = df_adj_aligned['adj_factor'].fillna(1.0)
|
||
|
||
# 计算后复权价格
|
||
df = df_daily.copy()
|
||
df['adj_factor'] = df_adj_aligned['adj_factor']
|
||
df['close_hfq'] = df['close'] * df['adj_factor']
|
||
df['code'] = code
|
||
|
||
return df[['code', 'open', 'close', 'adj_factor', 'close_hfq']]
|
||
|
||
except Exception as e:
|
||
print(f"Tushare下载ETF复权数据 {code} 失败: {e}")
|
||
return None
|
||
|
||
def fetch_trade_cal(self, start_date: str, end_date: str) -> pd.DatetimeIndex:
|
||
"""
|
||
获取 A 股（上交所 SSE）官方交易日历
|
||
|
||
Args:
|
||
start_date: 开始日期 'YYYY-MM-DD'
|
||
end_date: 结束日期 'YYYY-MM-DD'
|
||
|
||
Returns:
|
||
DatetimeIndex: A股交易日日期序列
|
||
"""
|
||
try:
|
||
pro = self._get_pro_api()
|
||
|
||
df = pro.trade_cal(
|
||
exchange='SSE',
|
||
start_date=start_date.replace('-', ''),
|
||
end_date=end_date.replace('-', ''),
|
||
is_open='1'
|
||
)
|
||
|
||
if df is None or len(df) == 0:
|
||
return pd.DatetimeIndex([])
|
||
|
||
# 提取交易日并转换为 DatetimeIndex
|
||
trade_dates = pd.to_datetime(df['cal_date'])
|
||
return pd.DatetimeIndex(trade_dates.sort_values())
|
||
|
||
except Exception as e:
|
||
print(f"Tushare下载交易日历失败: {e}")
|
||
return pd.DatetimeIndex([])
|
||
|
||
def fetch_stock_adj(self, code: str, start_date: str, end_date: str, adj: str = 'hfq') -> Optional[pd.DataFrame]:
|
||
"""
|
||
获取 A股股票复权价格数据
|
||
|
||
使用 pro_bar 接口获取前复权(qfq)或后复权(hfq)价格。
|
||
|
||
Args:
|
||
code: 股票代码，如 '000001.SZ', '600000.SH'
|
||
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 'hfq'
|
||
|
||
Returns:
|
||
DataFrame with columns: date, code, open, high, low, close, volume, adj_factor
|
||
"""
|
||
import tushare as ts
|
||
|
||
if adj not in ['qfq', 'hfq']:
|
||
raise ValueError(f"adj 参数必须是 'qfq' 或 'hfq'，当前: {adj}")
|
||
|
||
try:
|
||
ts_code = code.replace('.SS', '.SH')
|
||
|
||
# 使用 pro_bar 接口获取复权数据
|
||
df = ts.pro_bar(
|
||
ts_code=ts_code,
|
||
adj=adj,
|
||
start_date=start_date.replace('-', ''),
|
||
end_date=end_date.replace('-', ''),
|
||
adjfactor=True # 返回复权因子
|
||
)
|
||
|
||
if df is None or len(df) == 0:
|
||
return None
|
||
|
||
# 标准化列名
|
||
df = df.rename(columns={
|
||
'ts_code': 'code',
|
||
'trade_date': 'date',
|
||
'vol': 'volume',
|
||
})
|
||
|
||
# 转换日期格式
|
||
df['date'] = pd.to_datetime(df['date'])
|
||
df = df.set_index('date')
|
||
df = df.sort_index()
|
||
|
||
# 恢复原始代码格式（.SS -> .SH 反转）
|
||
df['code'] = code
|
||
|
||
# 标准化返回字段
|
||
columns = ['code', 'open', 'high', 'low', 'close', 'volume']
|
||
if 'adj_factor' in df.columns:
|
||
columns.append('adj_factor')
|
||
|
||
return df[columns]
|
||
|
||
except Exception as e:
|
||
print(f"Tushare下载股票复权数据 {code} 失败: {e}")
|
||
return None |