def Momentum021(cls,
data: pd.DataFrame,
price: str = PVN.CLOSE.value,
n: int = 1):
"""
最低价格时间因子(LT):1-index(Min(P, N))/L
"""
factor_name = sys._getframe().f_code.co_name + f'_{n}days'
data.set_index([KN.TRADE_DATE.value, KN.STOCK_ID.value], inplace=True)
data.sort_index(inplace=True)
data[factor_name] = data[price].groupby(KN.STOCK_ID.value,
group_keys=False).rolling(n).apply(
lambda x: 1 - (list(x).index(np.nanmin(x)) - 1) / len(x))
F = DataInfo()
F.data = data[factor_name]
F.data_type = 'MTM'
F.data_category = cls().__class__.__name__
F.data_name = factor_name
return F
def Momentum006(cls,
data: pd.DataFrame,
close_price: str = PVN.CLOSE.value,
n: int = 1):
"""
最高价格因子(HPTP):Max(P,N) / P
"""
factor_name = sys._getframe().f_code.co_name + f'_{n}days'
data.set_index([KN.TRADE_DATE.value, KN.STOCK_ID.value], inplace=True)
data.sort_index(inplace=True)
data[factor_name] = data[close_price].groupby(KN.STOCK_ID.value,
group_keys=False).apply(
lambda x: x.rolling(n, min_periods=1).max() / x)
F = DataInfo()
F.data = data[factor_name]
F.data_type = 'MTM'
F.data_category = cls().__class__.__name__
F.data_name = factor_name
return F
def Momentum013(cls,
data: pd.DataFrame,
close_price: str = PVN.CLOSE.value,
n: int = 1) -> DataInfo:
"""
动量CTC收益率(MTM_CTC):N日收盘价计算的收益率均值
:return:
"""
factor_name = sys._getframe().f_code.co_name + f'_{n}days'
data.set_index([KN.TRADE_DATE.value, KN.STOCK_ID.value], inplace=True)
data.sort_index(inplace=True)
data['return'] = data[close_price].groupby(KN.STOCK_ID.value).pct_change()
data[factor_name] = data['return'].groupby(KN.STOCK_ID.value, group_keys=False).rolling(n, min_periods=1).mean()
F = DataInfo()
F.data = data[factor_name]
F.data_type = 'MTM'
F.data_category = cls().__class__.__name__
F.data_name = factor_name
return F
def Momentum020(cls,
data: pd.DataFrame,
low_price: str = PVN.LOW.value,
high_price: str = PVN.HIGH.value,
open_price: str = PVN.OPEN.value,
close_price: str = PVN.CLOSE.value,
n: int = 1) -> DataInfo:
"""
动量CTHL收益率绝对值均值(MTM_CTHL_abs)
:return:
"""
factor_name = sys._getframe().f_code.co_name + f'_{n}days'
data.set_index([KN.TRADE_DATE.value, KN.STOCK_ID.value], inplace=True)
data.sort_index(inplace=True)
data['return_H'] = data.groupby(KN.STOCK_ID.value,
group_keys=False).apply(lambda x: (x[high_price] / x[close_price].shift(1) - 1))
data['return_L'] = data.groupby(KN.STOCK_ID.value,
group_keys=False).apply(lambda x: (x[low_price] / x[close_price].shift(1) - 1))
data['return_O'] = data.groupby(KN.STOCK_ID.value,
group_keys=False).apply(lambda x: (x[open_price] / x[close_price].shift(1) - 1))
data_sub_H = data[data['return_O'] > 0]['return_H']
data_sub_L = data[data['return_O'] < 0]['return_L']
data_sub_O = data[data['return_O'] == 0]['return_O']
data["CTHL"] = pd.concat([abs(data_sub_H), abs(data_sub_L), abs(data_sub_O)])
data[factor_name] = data['CTHL'].groupby(KN.STOCK_ID.value,
group_keys=False).rolling(n, min_periods=1).mean()
F = DataInfo()
F.data = data[factor_name]
F.data_type = 'MTM'
F.data_category = cls().__class__.__name__
F.data_name = factor_name
return F
def Momentum026(cls,
data: pd.DataFrame,
close_price: str = PVN.CLOSE.value,
n: int = 1) -> DataInfo:
"""
收益率rank标准差(MTM_RankPrice_std)
"""
factor_name = sys._getframe().f_code.co_name + f'_{n}days'
data.set_index([KN.TRADE_DATE.value, KN.STOCK_ID.value], inplace=True)
data.sort_index(inplace=True)
data['ret'] = data[close_price].groupby(KN.STOCK_ID.value).pct_change()
data['rank'] = data['ret'].groupby(KN.TRADE_DATE.value).rank()
data[factor_name] = data['rank'].groupby(KN.TRADE_DATE.value,
group_keys=False).rolling(n, min_periods=2).std(ddof=1)
F = DataInfo()
F.data = data[factor_name]
F.data_type = 'MTM'
F.data_category = cls().__class__.__name__
F.data_name = factor_name
return F
def strategyLabel(self) -> DataInfo:
func_name = sys._getframe().f_code.co_name
result_path = os.path.join(self.local_path, func_name + '.pkl')
if os.path.exists(result_path):
with open(result_path, 'rb') as f:
category_label = pickle.load(f)
else:
# read data
print(
f"{dt.datetime.now().strftime('%X')}: Construction the label pool"
)
data_dict = self.read_data()
price_data = data_dict['AStockData.csv'][self.Mapping["price"]
['columns']]
ind_exp = data_dict['AStockData.csv'][self.Mapping["industry"]
['columns']]
stock_mv_data = data_dict['AStockData.csv'][
self.Mapping["mv"]['columns']] * 10000 # wan yuan->yuan
# composition_data = data_dict['IndexMember.csv'][self.Mapping['composition']['columns']]
stock_w_data = data_dict['StockPool.csv'][self.Mapping['stock_w1']
['columns']]
up_down_limit = data_dict['AStockData.csv'][
self.Mapping['priceLimit']['columns']]
price_data = price_data.rename(
columns={
PVN.CLOSE_ADJ.value: PVN.CLOSE.value,
PVN.OPEN_ADJ.value: PVN.OPEN.value,
PVN.HIGH_ADJ.value: PVN.HIGH.value,
PVN.LOW_ADJ.value: PVN.LOW.value
})
print(
f"{dt.datetime.now().strftime('%X')}: Calculate stock daily return label"
)
stock_ret_c = self.api.stock_ret(price_data,
return_type=PVN.CLOSE.value)
stock_ret_o = self.api.stock_ret(price_data,
return_type=PVN.OPEN.value)
print(f"{dt.datetime.now().strftime('%X')}: Set price limit label")
up_down_limit = up_down_limit.fillna(1) == 0
############################################################################################################
# merge labels
print(f"{dt.datetime.now().strftime('%X')}: Merge labels")
category_label = self.merge_labels(
data_ret_close=stock_ret_c,
data_ret_open=stock_ret_o,
# composition=composition_data,
ind_exp=ind_exp,
mv=stock_mv_data[PVN.LIQ_MV.value],
stock_w=stock_w_data,
price_limit=up_down_limit)
# sort
category_label = category_label.sort_index()
category_label.to_pickle(result_path)
dataClass = DataInfo(data=category_label,
data_category=self.__class__.__name__,
data_name=func_name)
return dataClass
def StockPoolZD(self) -> DataInfo:
"""
1.剔除ST股：是ST为True
2.剔除成立年限小于6个月的股票：成立年限小于6个月为False
3.过去5天成交额占比排名最后5%：成交额占比在最后5%为False
4.过去5天停牌天数超过3天:停牌数超过阈值为False
注意：函数名需要与数据源文件名对应，保持一致防止出错，可自行修改
:return:
"""
func_name = sys._getframe().f_code.co_name
result_path = os.path.join(self.local_path, func_name + '.pkl')
if os.path.exists(result_path):
with open(result_path, 'rb') as f:
stock_judge = pickle.load(f)
else:
# read data
print(
f"{dt.datetime.now().strftime('%X')}: Read the data of stock pool"
)
data_input = self.read_data(func_name)
# get filter condition
print(f"{dt.datetime.now().strftime('%X')}: Weed out ST stock")
data_input['judge1'] = self.api.ST(data_input[PVN.ISST.value])
# print(f"{dt.datetime.now().strftime('%X')}: Weed out Price Up_Down limit stock")
# data_input['judge2'] = self.price_limit(data_input[PVN.Up_Down.value])
print(
f"{dt.datetime.now().strftime('%X')}: Weed out stock Established in less than 6 months"
)
data_input['judge3'] = self.api.established(
data=data_input[PVN.LIST_DAYS_NUM.value])
print(
f"{dt.datetime.now().strftime('%X')}: Weed out stock illiquidity"
)
data_input['judge4'] = self.api.liquidity(
data_input[PVN.AMOUNT.value])
print(
f"{dt.datetime.now().strftime('%X')}: Weed out Suspension stock"
)
data_input['judge5'] = self.api.suspension(
data_input[PVN.AMOUNT.value])
print(
f"{dt.datetime.now().strftime('%X')}: Weed out liq_mv less than 0.02 stock"
)
data_input['judge6'] = self.api.liq_mv(
data_input[PVN.LIQ_MV.value])
print(
f"{dt.datetime.now().strftime('%X')}: Weed out science and technology innovation board stock"
)
data_input['judge7'] = self.api.list_board(
data_input[PVN.LIST_BOARD.value])
Judge = [
column for column in data_input.columns if 'judge' in column
]
# Filter
stock_judge = data_input[Judge].all(axis=1).sort_index()
stock_judge.name = func_name
# to_csv
stock_judge.to_pickle(result_path)
# switch judge to label
stock_pool = stock_judge.groupby(
KN.STOCK_ID.value).shift(1).fillna(True)
dataClass = DataInfo(data=stock_pool[stock_pool],
data_category=self.__class__.__name__,
data_name=func_name)
return dataClass
