def get_adj_factor(self, adjustment_mode):
folder_name = LocalDataFolderName.ADJ_FACTOR.value
path = LocalDataPath.path + folder_name + '/'
data = pd.DataFrame({})
if adjustment_mode == RightsAdjustment.BACKWARD.value:
data = get_local_data(
path, AdjustmentFactor.BACKWARD_ADJ_FACTOR.value + '.h5')
elif adjustment_mode == RightsAdjustment.FROWARD.value:
data = get_local_data(
path, AdjustmentFactor.FROWARD_ADJ_FACTOR.value + '.h5')
return data
def get_all_sws_index(self):
folder_name = LocalDataFolderName.SWS_INDEX.value
path = LocalDataPath.path + folder_name + '/'
data_name = folder_name + '.h5'
self.all_sws_index = get_local_data(
path, data_name).sort_values(by='time_tag')
return self.all_sws_index
def cache_all_stock_data(self,
period=Period.DAILY.value,
dividend_type=RightsAdjustment.NONE.value,
field=None):
if field is not None:
self.field = field
folder_name = LocalDataFolderName.MARKET_DATA.value
sub_folder_name = LocalDataFolderName.KLINE_DAILY.value
sub_sub_folder_name = LocalDataFolderName.A_SHARE.value
path = LocalDataPath.path + folder_name + '/' + sub_folder_name + '/' + sub_sub_folder_name + '/'
all_market_date = {}
adj_factor = 1
if dividend_type == RightsAdjustment.FROWARD.value:
adj_factor = self.adj_factor_obj.get_adj_factor(
RightsAdjustment.FROWARD.value)
elif dividend_type == RightsAdjustment.BACKWARD.value:
adj_factor = self.adj_factor_obj.get_adj_factor(
RightsAdjustment.BACKWARD.value)
for i in self.field:
data_name = i + '.h5'
data = get_local_data(path, data_name)
if isinstance(adj_factor, int):
all_market_date[i] = data
else:
if i in ['open', 'high', 'low', 'close']:
all_market_date[i] = data.multiply(adj_factor)
else:
all_market_date[i] = data
return all_market_date
def cache_all_index_data(self, period=Period.DAILY.value):
folder_name = LocalDataFolderName.MARKET_DATA.value
sub_folder_name = LocalDataFolderName.KLINE_DAILY.value
sub_sub_folder_name = LocalDataFolderName.INDEX.value
path = LocalDataPath.path + folder_name + '/' + sub_folder_name + '/' + sub_sub_folder_name + '/'
index_date = {}
for i in self.field:
data_name = i + '.h5'
index_date[i] = get_local_data(path, data_name)
return index_date
def get_share(self, field=None):
"""
:param field: 'total_share', 'float_a_share', 'total_share_value', 'float_a_share_value'
:return:
"""
if field is None:
field = 'total_share'
folder_name = LocalDataFolderName.INDICATOR_EVERYDAY.value
path = LocalDataPath.path + folder_name + '/'
data_name = field + '.h5'
self.share_data = get_local_data(path, data_name)
return self.share_data
acf_result=self.acf_result,
# 因子收益率，单利，复利, 日收益率
factor_return=factor_return,
# 单因子检测的T值, Series, index为时间
factor_t_value=factor_t_value,
# 单因子检测的T值的统计值，'t_value_mean': 绝对值均值, 't_value_greater_two':绝对值序列大于2的占比
factor_t_value_statistics=self.factor_t_value_statistics,
# 净值分析结果
net_analysis_result=self.net_analysis_result)
doc.save()
if __name__ == '__main__':
factor_name = 'factor_ma10'
path = LocalDataPath.path + LocalDataFolderName.FACTOR.value + '/'
factor_ma5 = get_local_data(path, factor_name + '.h5')
# 指数数据不全，需要删一部分因子数据
factor_ma5 = factor_ma5[factor_ma5.index < datetime.datetime(2020, 1, 1)]
kline_object = GetKlineData()
market_data = kline_object.cache_all_stock_data(
dividend_type=RightsAdjustment.BACKWARD.value, field=['close'])
market_close_data = kline_object.get_market_data(market_data,
field=['close'])
# 指数行情，沪深300代替
all_index_data = kline_object.cache_all_index_data()
benchmark_df = kline_object.get_market_data(
all_index_data, stock_code=['000300.SH'],
field=['close']).to_frame(name='close')
# 沪深300 的日线，有脏数据，后续单独处理
result = factor_ic_mean.apply(cal_weight,
args=(factor_ic_cov, ),
axis=1,
result_type="expand")
result.columns = factor_ic.keys()
return result.div(result.sum(1), axis=0)
if __name__ == '__main__':
factor_list = ['factor_ma5', 'factor_ma10']
path = LocalDataPath.path + LocalDataFolderName.FACTOR.value + '/'
factor_data = {}
for factor_name in factor_list:
factor_single_data = get_local_data(path, factor_name + '.h5')
# 指数数据不全，需要删一部分因子数据
factor_data[factor_name] = factor_single_data[
factor_single_data.index < datetime(2020, 1, 1)]
with MongoConnect(DatabaseName.MULTI_FACTOR_DATA.value):
factor_ic = {}
factor_return = {}
for factor_name in factor_list:
factor_regression_analysis_result = FactorRegressionAnalysisResult.objects(factor_name=factor_name) \
.only('factor_name') \
.only('begin_date') \
.only('end_date') \
.only('factor_return') \
.as_pymongo()
factor_return[factor_name] = pd.DataFrame(
close_price = close_price_all['close'][stock]
if np.isnan(close_price):
continue
buy_share = int(
self.single_stock_value / close_price / 100) * 100
Trade(self).order_shares(stock_code=stock,
shares=buy_share,
price_type='fix',
order_price=close_price,
account_id=self.account[0])
Environment.logger.info(self.time_tag, 'buy', stock, buy_share,
'fix', close_price, self.account[0])
if __name__ == '__main__':
path = LocalDataPath.path + LocalDataFolderName.FACTOR.value + '/'
factor_ma5 = get_local_data(path, 'factor_ma5.h5')
# 指数数据不全，需要删一部分因子数据
factor_ma5 = factor_ma5[factor_ma5.index < datetime(2020, 1, 1)]
stratification_analysis_obj = StratificationAnalysis(
factor_ma5, 'factor_ma5')
stratification_analysis_obj.add_group()
group_hold = stratification_analysis_obj.cal_group_hold(
stratification_analysis_obj.group_key[0])
stratification_strategy = StratificationStrategy(group_hold)
import time
a = time.time()
stratification_strategy.run(save_trade_record=True)
print(time.time() - a)
def get_all_index_members(self):
folder_name = LocalDataFolderName.INDEX_MEMBER.value
path = LocalDataPath.path + folder_name + '/'
data_name = folder_name + '.h5'
self.all_index_members_df = get_local_data(path, data_name)
return self.all_index_members_df
def get_calendar(self, market):
folder_name = LocalDataFolderName.CALENDAR.value
path = LocalDataPath.path + folder_name + '/'
data_name = 'calendar_' + market + '.h5'
data = get_local_data(path, data_name)
return list(data[0])
def get_index_class(self):
folder_name = LocalDataFolderName.INDUSTRY_CLASS.value
path = LocalDataPath.path + folder_name + '/'
data_name = 'sw_industry_one' + '.h5'
self.index_class_df = get_local_data(path, data_name)
return self.index_class_df
