add CSI500 data collector

scripts/data_collector/cn_index/README.md

@@ -1,4 +1,4 @@
-# CSI300/CSI100 History Companies Collection
+# CSI300/CSI100/CSI500 History Companies Collection
 
 ## Requirements
 
@@ -15,7 +15,7 @@ python collector.py --index_name CSI300 --qlib_dir ~/.qlib/qlib_data/cn_data --m
 # parse new companies
 python collector.py --index_name CSI300 --qlib_dir ~/.qlib/qlib_data/cn_data --method save_new_companies
 
-# index_name support: CSI300, CSI100
+# index_name support: CSI300, CSI100, CSI500
 # help
 python collector.py --help
 ```

scripts/data_collector/cn_index/collector.py

@@ -5,6 +5,7 @@ import re
 import abc
 import sys
 import importlib
+from tqdm import tqdm
 from io import BytesIO
 from typing import List, Iterable
 from pathlib import Path
@@ -12,6 +13,8 @@ from pathlib import Path
 import fire
 import requests
 import pandas as pd
+import baostock as bs
+from lxml import etree
 from loguru import logger
 
 CUR_DIR = Path(__file__).resolve().parent
@@ -44,6 +47,7 @@ def retry_request(url: str, method: str = "get", exclude_status: List = None):
 
 
 class CSIIndex(IndexBase):
+
     @property
     def calendar_list(self) -> List[pd.Timestamp]:
         """get history trading date
@@ -314,6 +318,7 @@ class CSIIndex(IndexBase):
 
 
 class CSI300(CSIIndex):
+
     @property
     def index_code(self):
         return "000300"
@@ -324,10 +329,11 @@ class CSI300(CSIIndex):
 
     @property
     def html_table_index(self):
-        return 1
+        return 0
 
 
 class CSI100(CSIIndex):
+
     @property
     def index_code(self):
         return "000903"
@@ -338,16 +344,54 @@ class CSI100(CSIIndex):
 
     @property
     def html_table_index(self):
-        return 2
+        return 1
+
+
+class CSI500(CSIIndex):
+
+    @property
+    def index_code(self):
+        return "000905"
+
+    @property
+    def bench_start_date(self) -> pd.Timestamp:
+        return pd.Timestamp("2007-01-15")
+
+    @property
+    def html_table_index(self):
+        return 0
+
+    def get_changes(self):
+        return self.get_changes_with_history_companies(self.get_history_companies())
+
+    def get_history_companies(self):
+        """
+		Data source：http://baostock.com/baostock/index.php/%E4%B8%AD%E8%AF%81500%E6%88%90%E5%88%86%E8%82%A1
+			Avoid a large number of parallel data acquisition,
+			such as 1000 times of concurrent data acquisition, because IP will be blocked
+		Returns
+		-------
+
+		"""
+        lg = bs.login()
+        today = pd.datetime.now()
+        date_range = pd.DataFrame(pd.date_range(start="2007-01-15", end=today, freq="7D"))[0].dt.date
+        ret_list = []
+        col = ["date", "symbol", "code_name"]
+        for date in tqdm(date_range, desc="Download CSI500"):
+            rs = bs.query_zz500_stocks(date=str(date))
+            zz500_stocks = []
+            while (rs.error_code == "0") & rs.next():
+                zz500_stocks.append(rs.get_row_data())
+            result = pd.DataFrame(zz500_stocks, columns=col)
+            result["symbol"] = result["symbol"].apply(lambda x: x.replace(".", "").upper())
+            ret_list.append(result[["date", "symbol"]])
+        bs.logout()
+        return pd.concat(ret_list, sort=False)
 
 
 def get_instruments(
-    qlib_dir: str,
-    index_name: str,
-    method: str = "parse_instruments",
-    freq: str = "day",
-    request_retry: int = 5,
-    retry_sleep: int = 3,
+    qlib_dir: str, index_name: str, method: str = "parse_instruments", request_retry: int = 5, retry_sleep: int = 3
 ):
     """
 

scripts/data_collector/cn_index/requirements.txt

@@ -1,5 +1,8 @@
+baostock
+logure
 fire
 requests
 pandas
 lxml
 loguru
+tqdm

scripts/data_collector/utils.py

@@ -28,15 +28,14 @@ SZSE_CALENDAR_URL = "http://www.szse.cn/api/report/exchange/onepersistenthour/mo
 
 CALENDAR_BENCH_URL_MAP = {
     "CSI300": CALENDAR_URL_BASE.format(market=1, bench_code="000300"),
+    "CSI500": CALENDAR_URL_BASE.format(market=1, bench_code="000905"),
     "CSI100": CALENDAR_URL_BASE.format(market=1, bench_code="000903"),
     # NOTE: Use the time series of SH600000 as the sequence of all stocks
     "ALL": CALENDAR_URL_BASE.format(market=1, bench_code="000905"),
     # NOTE: Use the time series of ^GSPC(SP500) as the sequence of all stocks
     "US_ALL": "^GSPC",
-    "IN_ALL": "^NSEI",
 }
 
-
 _BENCH_CALENDAR_LIST = None
 _ALL_CALENDAR_LIST = None
 _HS_SYMBOLS = None
@@ -234,13 +233,16 @@ def get_us_stock_symbols(qlib_data_path: [str, Path] = None) -> list:
         resp = requests.get(url)
         if resp.status_code != 200:
             raise ValueError("request error")
+
         try:
             _symbols = [_v["f12"].replace("_", "-P") for _v in resp.json()["data"]["diff"].values()]
         except Exception as e:
             logger.warning(f"request error: {e}")
             raise
+
         if len(_symbols) < 8000:
             raise ValueError("request error")
+
         return _symbols
 
     @deco_retry
@@ -273,6 +275,7 @@ def get_us_stock_symbols(qlib_data_path: [str, Path] = None) -> list:
         resp = requests.post(url, json=_parms)
         if resp.status_code != 200:
             raise ValueError("request error")
+
         try:
             _symbols = [_v["symbolTicker"].replace("-", "-P") for _v in resp.json()]
         except Exception as e:
@@ -449,7 +452,9 @@ def symbol_prefix_to_sufix(symbol: str, capital: bool = True) -> str:
 
 
 def deco_retry(retry: int = 5, retry_sleep: int = 3):
+
     def deco_func(func):
+
         @functools.wraps(func)
         def wrapper(*args, **kwargs):
             _retry = 5 if callable(retry) else retry
@@ -458,10 +463,12 @@ def deco_retry(retry: int = 5, retry_sleep: int = 3):
                 try:
                     _result = func(*args, **kwargs)
                     break
+
                 except Exception as e:
                     logger.warning(f"{func.__name__}: {_i} :{e}")
                     if _i == _retry:
                         raise
+
                 time.sleep(retry_sleep)
             return _result