def get_futures_daily(start_day=None,
end_day=None,
market="CFFEX",
index_bar=False):
"""
获取交易所日交易数据
Parameters
------
start_day: 开始日期 format：YYYY-MM-DD 或 YYYYMMDD 或 datetime.date对象 为空时为当天
end_day: 结束数据 format：YYYY-MM-DD 或 YYYYMMDD 或 datetime.date对象 为空时为当天
market: 'CFFEX' 中金所, 'CZCE' 郑商所, 'SHFE' 上期所, 'DCE' 大商所 之一。默认为中金所
index_bar: bool 是否合成指数K线
Return
-------
DataFrame
中金所日交易数据(DataFrame):
symbol 合约代码
date 日期
open 开盘价
high 最高价
low 最低价
close 收盘价
volume 成交量
open_interest 持仓量
turnover 成交额
settle 结算价
pre_settle 前结算价
variety 合约类别
或 None(给定日期没有交易数据)
"""
if market.upper() == "CFFEX":
f = get_cffex_daily
elif market.upper() == "CZCE":
f = get_czce_daily
elif market.upper() == "SHFE":
f = get_shfe_daily
elif market.upper() == "DCE":
f = get_dce_daily
else:
print("Invalid Market Symbol")
return
start_day = (cons.convert_date(start_day)
if start_day is not None else datetime.date.today())
end_day = (cons.convert_date(end_day)
if end_day is not None else cons.convert_date(
cons.get_latest_data_date(datetime.datetime.now())))
df_list = list()
while start_day <= end_day:
df = f(start_day)
if df is not None:
df_list.append(df)
if index_bar:
df_list.append(get_futures_index(df))
start_day += datetime.timedelta(days=1)
if len(df_list) > 0:
return pd.concat(df_list).reset_index(drop=True)
def get_spot_price_daily(start_day=None,
end_day=None,
vars_list=cons.contract_symbols):
"""
获取某段时间大宗商品现货价格及相应基差
:param start_day: str 开始日期 format：YYYY-MM-DD 或 YYYYMMDD 或 datetime.date对象; 默认为当天
:param end_day: str 结束数据 format：YYYY-MM-DD 或 YYYYMMDD 或 datetime.date对象; 默认为当天
:param vars_list: list 合约品种如 [RB, AL]; 默认参数为所有商品
:return: pandas.DataFrame
展期收益率数据:
var 商品品种 string
sp 现货价格 float
near_symbol 临近交割合约 string
near_price 临近交割合约结算价 float
dom_symbol 主力合约 string
dom_price 主力合约结算价 float
near_basis 临近交割合约相对现货的基差 float
dom_basis 主力合约相对现货的基差 float
near_basis_rate 临近交割合约相对现货的基差率 float
dom_basis_rate 主力合约相对现货的基差率 float
date 日期 string YYYYMMDD
"""
start_day = (cons.convert_date(start_day)
if start_day is not None else datetime.date.today())
end_day = (cons.convert_date(end_day)
if end_day is not None else cons.convert_date(
cons.get_latest_data_date(datetime.datetime.now())))
df_list = []
while start_day <= end_day:
print(start_day)
temp_df = get_spot_price(start_day, vars_list)
if temp_df is False:
return pd.concat(df_list).reset_index(drop=True)
elif temp_df is not None:
df_list.append(temp_df)
start_day += datetime.timedelta(days=1)
if len(df_list) > 0:
temp_df = pd.concat(df_list)
temp_df.reset_index(drop=True, inplace=True)
return temp_df
def get_rank_sum_daily(start_day=None, end_day=None, vars_list=cons.contract_symbols):
"""
采集四个期货交易所前5、前10、前15、前20会员持仓排名数据
注1：由于上期所和中金所只公布每个品种内部的标的排名，没有公布品种的总排名;
所以函数输出的品种排名是由品种中的每个标的加总获得，并不是真实的品种排名列表
注2：大商所只公布了品种排名，未公布标的排名
:param start_day: 开始日期 format：YYYY-MM-DD 或 YYYYMMDD 或 datetime.date对象 为空时为当天
:param end_day: 结束数据 format：YYYY-MM-DD 或 YYYYMMDD 或 datetime.date对象 为空时为当天
:param vars_list: 合约品种如RB、AL等列表 为空时为所有商品
:return: pd.DataFrame
展期收益率数据(DataFrame):
symbol 标的合约 string
var 商品品种 string
vol_top5 成交量前5会员成交量总和 int
vol_chg_top5 成交量前5会员成交量变化总和 int
long_open_interest_top5 持多单前5会员持多单总和 int
long_open_interest_chg_top5 持多单前5会员持多单变化总和 int
short_open_interest_top5 持空单前5会员持空单总和 int
short_open_interest_chg_top5 持空单前5会员持空单变化总和 int
vol_top10 成交量前10会员成交量总和 int
...
date 日期 string YYYYMMDD
"""
start_day = cons.convert_date(start_day) if start_day is not None else datetime.date.today()
end_day = cons.convert_date(end_day) if end_day is not None else cons.convert_date(
cons.get_latest_data_date(datetime.datetime.now()))
records = pd.DataFrame()
while start_day <= end_day:
print(start_day)
if start_day.strftime('%Y%m%d') in calendar:
data = get_rank_sum(start_day, vars_list)
if data is False:
print(f"{start_day.strftime('%Y-%m-%d')}日交易所数据连接失败，已超过20次，您的地址被网站墙了，请保存好返回数据，稍后从该日期起重试")
return records.reset_index(drop=True)
records = records.append(data)
else:
warnings.warn(f"{start_day.strftime('%Y%m%d')}非交易日")
start_day += datetime.timedelta(days=1)
return records.reset_index(drop=True)
def get_receipt(start_day: str = None,
end_day: str = None,
vars_list: List = cons.contract_symbols):
"""
获取大宗商品注册仓单数量
:param start_day: 开始日期 format：YYYY-MM-DD 或 YYYYMMDD 或 datetime.date对象 为空时为当天
:type start_day: str
:param end_day: 结束数据 format：YYYY-MM-DD 或 YYYYMMDD 或 datetime.date对象 为空时为当天
:type end_day: str
:param vars_list: 合约品种如RB、AL等列表 为空时为所有商品
:type vars_list: str
:return: 展期收益率数据
:rtype: pandas.DataFrame
"""
start_day = cons.convert_date(
start_day) if start_day is not None else datetime.date.today()
end_day = cons.convert_date(
end_day) if end_day is not None else cons.convert_date(
cons.get_latest_data_date(datetime.datetime.now()))
records = pd.DataFrame()
while start_day <= end_day:
if start_day.strftime('%Y%m%d') not in calendar:
warnings.warn(f"{start_day.strftime('%Y%m%d')}非交易日")
else:
print(start_day)
for market, market_vars in cons.market_exchange_symbols.items():
if market == 'dce':
if start_day >= datetime.date(2009, 4, 7):
f = get_dce_receipt
else:
print('20090407起，dce每交易日更新仓单数据')
f = None
elif market == 'shfe':
if datetime.date(2008, 10,
6) <= start_day <= datetime.date(
2014, 5, 16):
f = get_shfe_receipt_1
elif start_day > datetime.date(2014, 5, 16):
f = get_shfe_receipt_2
else:
f = None
print('20081006起，shfe每交易日更新仓单数据')
elif market == 'czce':
if datetime.date(2008, 3, 3) <= start_day <= datetime.date(
2010, 8, 24):
f = get_czce_receipt_1
elif datetime.date(2010, 8,
24) < start_day <= datetime.date(
2015, 11, 11):
f = get_czce_receipt_2
elif start_day > datetime.date(2015, 11, 11):
f = get_czce_receipt_3
else:
f = None
print('20080303起，czce每交易日更新仓单数据')
get_vars = [var for var in vars_list if var in market_vars]
if market != 'cffex' and get_vars != []:
if f is not None:
records = records.append(f(start_day, get_vars))
start_day += datetime.timedelta(days=1)
records.reset_index(drop=True, inplace=True)
if "MA" in records["var"].to_list():
replace_index = records[records["var"] == "MA"]["receipt"].astype(
str).str.split("0", expand=True)[0].index
records.loc[replace_index, "receipt"] = records[
records["var"] == "MA"]["receipt"].astype(str).str.split(
"0", expand=True)[0]
return records
def get_roll_yield_bar(type_method="symbol",
var="RB",
date=None,
start_day=None,
end_day=None,
plot=False):
"""
获取展期收益率
:param type_method: 'symbol'：获取某天某品种所有交割月合约的收盘价, 'var'：获取某天所有品种两个主力合约的展期收益率（展期收益率横截面）, ‘date’：获取某品种每天的两个主力合约的展期收益率（展期收益率时间序列）
:param var: 合约品种如RB、AL等
:param date: 某一天日期 format： YYYYMMDD
:param start_day: 开始日期 format：YYYYMMDD
:param end_day: 结束数据 format：YYYYMMDD
:param plot: True or False作图
:return: pd.DataFrame
展期收益率数据(DataFrame):
ry 展期收益率
index 日期或品种
"""
date = cons.convert_date(
date) if date is not None else datetime.date.today()
start_day = (cons.convert_date(start_day)
if start_day is not None else datetime.date.today())
end_day = (cons.convert_date(end_day)
if end_day is not None else cons.convert_date(
cons.get_latest_data_date(datetime.datetime.now())))
if type_method == "symbol":
df = get_futures_daily(start_day=date,
end_day=date,
market=symbol_market(var))
df = df[df["variety"] == var]
if plot:
_plot_bar_2(df[["symbol", "close"]])
return df
if type_method == "var":
df = pd.DataFrame()
for market in ["dce", "cffex", "shfe", "czce"]:
df = df.append(
get_futures_daily(start_day=date, end_day=date, market=market))
var_list = list(set(df["variety"]))
df_l = pd.DataFrame()
for var in var_list:
ry = get_roll_yield(date, var, df=df)
if ry:
df_l = df_l.append(
pd.DataFrame([ry],
index=[var],
columns=["roll_yield", "near_by",
"deferred"]))
df_l["date"] = date
df_l = df_l.sort_values("roll_yield")
if plot:
_plot_bar(df_l["roll_yield"])
return df_l
if type_method == "date":
df_l = pd.DataFrame()
while start_day <= end_day:
try:
ry = get_roll_yield(start_day, var)
if ry:
df_l = df_l.append(
pd.DataFrame(
[ry],
index=[start_day],
columns=["roll_yield", "near_by", "deferred"],
))
except:
pass
start_day += datetime.timedelta(days=1)
if plot:
_plot(df_l["roll_yield"])
return df_l