def prepare_calculate(self, trade_date):
self.trade_date = trade_date
tp_price_return, temp_price_sets = self.get_basic_data(trade_date)
# tp_price_return.set_index('symbol', inplace=True)
# tp_price_return['symbol'] = tp_price_return.index
# symbol_sets = list(set(tp_price_return['symbol']))
# tp_price_return_list = pd.DataFrame()
#
# for symbol in symbol_sets:
# if len(tp_price_return[tp_price_return['symbol'] == symbol]) < 3:
# continue
# tp_price_return_list = tp_price_return_list.append(
# tp_price_return.loc[symbol].sort_values(by='trade_date', ascending=True))
if len(tp_price_return) <= 0:
print("%s has no data" % trade_date)
return
else:
session = str(int(time.time() * 1000000 + datetime.now().microsecond))
data = {
'total_data': tp_price_return.to_json(orient='records'),
'index_daily_price_sets': temp_price_sets.to_json(orient='records')
}
cache_data.set_cache(session, 'volatility' + str(trade_date),
json.dumps(data))
# cache_data.set_cache(session, 'volatility' + str(trade_date) + '_a',
# tp_price_return_list.to_json(orient='records'))
# cache_data.set_cache(session, 'volatility' + str(trade_date) + '_b',
# temp_price_sets.to_json(orient='records'))
factor_volatility_value_task.calculate.delay(factor_name='volatility' + str(trade_date), trade_date=trade_date,
session=session)
def remote_run(self, trade_date):
total_data = self.loadon_data(trade_date)
# 存储数据
session = str(int(time.time() * 1000000 + datetime.now().microsecond))
cache_data.set_cache(session, 'factor_earning_expectation',
total_data.to_json(orient='records'))
distributed_factor.delay(session, trade_date,
json.dumps(self._methods), self._name)
def growth_calculate(trade_date):
# cash flow
ttm_factor_sets, balance_sets = get_basic_growth_data(trade_date)
growth_sets = pd.merge(ttm_factor_sets, balance_sets, on='symbol')
if len(growth_sets) <= 0:
print("%s has no data" % trade_date)
return
else:
tic = time.time()
session = str(int(time.time() * 1000000 + datetime.now().microsecond))
cache_data.set_cache(session + str(trade_date), trade_date,
growth_sets.to_json(orient='records'))
factor_growth.factor_calculate.delay(date_index=trade_date,
session=session)
time1 = time.time()
print('growth_cal_time:{}'.format(time1 - tic))
def cash_flow_calculate(trade_date):
# cash flow
tp_cash_flow, ttm_cash_flow_sets = get_basic_cash_flow(trade_date)
if len(tp_cash_flow) <= 0 or len(ttm_cash_flow_sets) <= 0:
print("%s has no data" % trade_date)
return
else:
tic = time.time()
session = str(int(time.time() * 1000000 + datetime.now().microsecond))
cache_data.set_cache(session + str(trade_date) + "1", trade_date,
tp_cash_flow.to_json(orient='records'))
cache_data.set_cache(session + str(trade_date) + "2", trade_date,
ttm_cash_flow_sets.to_json(orient='records'))
factor_cash_flow.factor_calculate.delay(date_index=trade_date,
session=session)
time4 = time.time()
print('cash_flow_cal_time:{}'.format(time4 - tic))
def constrain_calculate(trade_date):
# cash flow
balance_sets, ttm_factors_sets = get_basic_constrain(trade_date)
if len(balance_sets) <= 0 or len(ttm_factors_sets) <= 0:
print("%s has no data" % trade_date)
return
else:
tic = time.time()
session = str(int(time.time() * 1000000 + datetime.now().microsecond))
cache_data.set_cache(session + str(trade_date) + '1', trade_date,
balance_sets.to_json(orient='records'))
cache_data.set_cache(session + str(trade_date) + '2', trade_date,
ttm_factors_sets.to_json(orient='records'))
factor_constrain.factor_calculate.delay(date_index=trade_date,
session=session)
time5 = time.time()
print('constrain_cal_time:{}'.format(time5 - tic))
def prepaer_calculate(self, trade_date):
valuation_sets, ttm_factor_sets, income_sets, balance_set = self.get_basic_data(
trade_date)
# valuation_sets = pd.merge(valuation_sets, income_sets, on='symbol')
valuation_sets = pd.merge(valuation_sets, ttm_factor_sets, on='symbol')
valuation_sets = pd.merge(valuation_sets, balance_set, on='symbol')
if len(valuation_sets) <= 0:
print("%s has no data" % trade_date)
return
else:
session = str(
int(time.time() * 1000000 + datetime.now().microsecond))
cache_data.set_cache(session, 'scale' + str(trade_date),
valuation_sets.to_json(orient='records'))
factor_scale_value_task.calculate.delay(factor_name='scale' +
str(trade_date),
trade_date=trade_date,
session=session)
def historical_value_calculate(trade_date):
# history_value
valuation_sets, ttm_factor_sets, cash_flow_sets, income_sets = get_basic_history_value_data(
trade_date)
valuation_sets = pd.merge(valuation_sets, income_sets, on='symbol')
valuation_sets = pd.merge(valuation_sets, ttm_factor_sets, on='symbol')
valuation_sets = pd.merge(valuation_sets, cash_flow_sets, on='symbol')
if len(valuation_sets) <= 0:
print("%s has no data" % trade_date)
return
else:
tic = time.time()
session = str(int(time.time() * 1000000 + datetime.now().microsecond))
cache_data.set_cache(session + str(trade_date), trade_date,
valuation_sets.to_json(orient='records'))
historical_value.factor_calculate.delay(date_index=trade_date,
session=session)
time2 = time.time()
print('history_cal_time:{}'.format(time2 - tic))
def prepare_calculate(trade_date):
# per share indicators
valuation_sets, ttm_factor_sets, cash_flow_sets, income_sets, balance_sets = get_basic_scale_data(
trade_date)
valuation_sets = pd.merge(valuation_sets, income_sets, on='symbol')
valuation_sets = pd.merge(valuation_sets, ttm_factor_sets, on='symbol')
valuation_sets = pd.merge(valuation_sets, cash_flow_sets, on='symbol')
valuation_sets = pd.merge(valuation_sets, balance_sets, on='symbol')
if len(valuation_sets) <= 0:
print("%s has no data" % trade_date)
return
else:
tic = time.time()
session = str(int(time.time() * 1000000 + datetime.now().microsecond))
cache_data.set_cache(session + str(trade_date), trade_date,
valuation_sets.to_json(orient='records'))
factor_per_share_indicators.factor_calculate.delay(
date_index=trade_date, session=session)
time3 = time.time()
print('per_share_cal_time:{}'.format(time3 - tic))
def earning_calculate(trade_date):
# cash flow
tp_earning, ttm_earning_5y, ttm_earning = get_basic_earning(trade_date)
if len(tp_earning) <= 0 or len(ttm_earning_5y) <= 0 or len(
ttm_earning) <= 0:
print("%s has no data" % trade_date)
return
else:
tic = time.time()
session = str(int(time.time() * 1000000 + datetime.now().microsecond))
cache_data.set_cache(session + str(trade_date) + "1", trade_date,
tp_earning.to_json(orient='records'))
cache_data.set_cache(session + str(trade_date) + "2", trade_date,
ttm_earning_5y.to_json(orient='records'))
cache_data.set_cache(session + str(trade_date) + "3", trade_date,
ttm_earning.to_json(orient='records'))
factor_earning.factor_calculate.delay(date_index=trade_date,
session=session)
time6 = time.time()
print('earning_cal_time:{}'.format(time6 - tic))