def __init__(self, **kwargs):
super(DXDataCenter, self).__init__(kwargs['logger'],
kwargs['symbolList'])
self.fields = [
'productID', 'instrumentID', 'tradingDate', 'tradingTime',
'openPrice', 'highPrice', 'lowPrice', 'closePrice', 'volume',
'multiplier', 'openInterest'
]
self.startDate = kwargs['startDate']
self.endDate = kwargs['endDate']
self._freq = kwargs['freq']
self.baseDate = adjustDateByCalendar('China.SSE', self.endDate,
BizDayConventions.Preceding)
if not Settings.usingCache:
self.forceUpdate = True
else:
self.forceUpdate = False
if self._freq == FreqType.MIN5:
self.loadSchedule = makeSchedule(self.startDate, self.endDate,
'1m')
elif self._freq == FreqType.MIN1:
self.loadSchedule = makeSchedule(self.startDate, self.endDate,
'7d')
else:
self.loadSchedule = [self.startDate, self.endDate]
self.schCurrEnd = 1
while True:
self._getMinutesBars(
startDate=self.loadSchedule[self.schCurrEnd -
1].strftime("%Y-%m-%d"),
endDate=self.loadSchedule[self.schCurrEnd].strftime(
"%Y-%m-%d"),
freq=self._freq)
if self.symbolList:
break
self.logger.warning('There is no any valid data in the back-testing data range ({0} - {1})' \
.format(self.loadSchedule[self.schCurrEnd - 1], self.loadSchedule[self.schCurrEnd]))
if self.schCurrEnd == len(self.loadSchedule) - 1:
break
self.schCurrEnd += 1
if not self.symbolList:
raise ValueError(
'There is no any valid data in the back-testing whole data range'
)
if kwargs['benchmark']:
self._getBenchmarkData(kwargs['benchmark'],
self.startDate.strftime("%Y-%m-%d"),
self.endDate.strftime("%Y-%m-%d"))
def __init__(self, **kwargs):
super(DXDataCenter, self).__init__(kwargs['logger'], kwargs['symbolList'])
self.fields = ['productID',
'instrumentID',
'tradingDate',
'tradingTime',
'openPrice',
'highPrice',
'lowPrice',
'closePrice',
'volume',
'multiplier',
'openInterest']
self.startDate = kwargs['startDate']
self.endDate = kwargs['endDate']
self._freq = kwargs['freq']
self.baseDate = adjustDateByCalendar('China.SSE', self.endDate, BizDayConventions.Preceding)
if not Settings.usingCache:
self.forceUpdate = True
else:
self.forceUpdate = False
if self._freq == FreqType.MIN5:
self.loadSchedule = makeSchedule(self.startDate, self.endDate, '1m')
elif self._freq == FreqType.MIN1:
self.loadSchedule = makeSchedule(self.startDate, self.endDate, '7d')
else:
self.loadSchedule = [self.startDate, self.endDate]
self.schCurrEnd = 1
while True:
self._getMinutesBars(startDate=self.loadSchedule[self.schCurrEnd - 1].strftime("%Y-%m-%d"),
endDate=self.loadSchedule[self.schCurrEnd].strftime("%Y-%m-%d"),
freq=self._freq)
if self.symbolList:
break
self.logger.warning('There is no any valid data in the back-testing data range ({0} - {1})' \
.format(self.loadSchedule[self.schCurrEnd - 1], self.loadSchedule[self.schCurrEnd]))
if self.schCurrEnd == len(self.loadSchedule) - 1:
break
self.schCurrEnd += 1
if not self.symbolList:
raise ValueError('There is no any valid data in the back-testing whole data range')
if kwargs['benchmark']:
self._getBenchmarkData(kwargs['benchmark'],
self.startDate.strftime("%Y-%m-%d"),
self.endDate.strftime("%Y-%m-%d"))
def __init__(self,
alpha_model,
data_meta,
universe,
start_date,
end_date,
freq,
benchmark=905,
industry_cat='sw_adj',
industry_level=1,
dask_client=None):
self.alpha_model = alpha_model
self.data_meta = data_meta
self.universe = universe
self.benchmark = benchmark
self.dates = makeSchedule(start_date, end_date, freq, 'china.sse')
self.dates = [d.strftime('%Y-%m-%d') for d in self.dates]
self.industry_cat = industry_cat
self.industry_level = industry_level
self.freq = freq
self.horizon = map_freq(freq)
self.engine = SqlEngine(self.data_meta.data_source)
self.dask_client = dask_client
self.total_data = None
self.index_return = None
self.risk_models = None
self.alpha_models = None
def __init__(self,
universe,
start_date,
end_date,
freq,
benchmark=905,
weights_bandwidth=0.02,
industry_cat='sw_adj',
industry_level=1,
rebalance_method='risk_neutral',
bounds=None,
**kwargs):
self.universe = universe
self.dates = makeSchedule(start_date, end_date, freq, 'china.sse')
self.dates = [d.strftime('%Y-%m-%d') for d in self.dates]
self.benchmark = benchmark
self.weights_bandwidth = weights_bandwidth
self.freq = freq
self.horizon = map_freq(freq)
self.executor = NaiveExecutor()
self.industry_cat = industry_cat
self.industry_level = industry_level
self.rebalance_method = rebalance_method
self.bounds = bounds
self.more_opts = kwargs
def factor(self,
factor_category,
begin_date,
end_date,
factor_name=None,
freq=None):
if factor_name is None:
table = importlib.import_module(
'data.factor_model').__getattribute__(factor_category)
return self.base(table, begin_date, end_date, freq)
else:
table = importlib.import_module(
'data.factor_model').__getattribute__(factor_category)
key_sets = ['id', 'security_code', 'trade_date'] + factor_name
db_columns = []
for key in key_sets:
db_columns.append(table.__dict__[key])
if freq is None:
query = select(db_columns).where(
and_(
table.trade_date >= begin_date,
table.trade_date <= end_date,
))
else:
rebalance_dates = makeSchedule(begin_date, end_date, freq,
'china.sse',
BizDayConventions.Preceding)
query = select(db_columns).where(
table.trade_date.in_(rebalance_dates))
return pd.read_sql(query, self._engine.sql_engine()).drop(['id'],
axis=1)
def factor_combination(engine, factors, universe_name_list, start_date,
end_date, freq):
universe = None
for name in universe_name_list:
if universe is None:
universe = Universe(name)
else:
universe += Universe(name)
dates = makeSchedule(start_date, end_date, freq, calendar='china.sse')
factor_negMkt = engine.fetch_factor_range(universe,
"negMarketValue",
dates=dates)
risk_cov, risk_factors = engine.fetch_risk_model_range(universe,
dates=dates)
dx_returns = engine.fetch_dx_return_range(universe,
dates=dates,
horizon=map_freq(freq))
# data combination
total_data = pd.merge(factors, risk_factors, on=['trade_date', 'code'])
total_data = pd.merge(total_data, factor_negMkt, on=['trade_date', 'code'])
total_data = pd.merge(total_data, dx_returns, on=['trade_date', 'code'])
industry_category = engine.fetch_industry_range(universe, dates=dates)
total_data = pd.merge(total_data,
industry_category,
on=['trade_date', 'code']).dropna()
total_data.dropna(inplace=True)
return total_data
def test_sql_engine_fetch_dx_return_with_universe_adjustment(self):
ref_dates = makeSchedule(
advanceDateByCalendar('china.sse', '2017-01-26', '-6m'),
'2017-01-26', '60b', 'china.sse')
universe = Universe('zz500')
dx_return = self.engine.fetch_dx_return_range(universe,
dates=ref_dates,
horizon=4,
offset=1)
codes = self.engine.fetch_codes_range(universe, dates=ref_dates)
groups = codes.groupby('trade_date')
for ref_date, g in groups:
start_date = advanceDateByCalendar('china.sse', ref_date, '2b')
end_date = advanceDateByCalendar('china.sse', ref_date, '6b')
query = select([Market.code, Market.chgPct]).where(
and_(Market.trade_date.between(start_date, end_date),
Market.code.in_(g.code.unique().tolist())))
df = pd.read_sql(query, con=self.engine.engine)
res = df.groupby('code').apply(lambda x: np.log(1. + x).sum())
calculated_return = dx_return[dx_return.trade_date == ref_date]
np.testing.assert_array_almost_equal(calculated_return.dx.values,
res.chgPct.values)
def test_sql_engine_fetch_factor_range_forward(self):
ref_dates = makeSchedule(
advanceDateByCalendar('china.sse', self.ref_date, '-6m'),
self.ref_date, '60b', 'china.sse')
ref_dates = ref_dates + [
advanceDateByCalendar('china.sse', ref_dates[-1],
'60b').strftime('%Y-%m-%d')
]
universe = Universe('zz500') + Universe('zz1000')
factor = 'ROE'
factor_data = self.engine.fetch_factor_range_forward(universe,
factor,
dates=ref_dates)
codes = self.engine.fetch_codes_range(universe, dates=ref_dates[:-1])
groups = codes.groupby('trade_date')
for ref_date, g in groups:
forward_ref_date = advanceDateByCalendar(
'china.sse', ref_date, '60b').strftime('%Y-%m-%d')
query = select([Uqer.code, Uqer.ROE]).where(
and_(Uqer.trade_date == forward_ref_date,
Uqer.code.in_(g.code.unique().tolist())))
df = pd.read_sql(query, con=self.engine.engine)
calculated_factor = factor_data[factor_data.trade_date == ref_date]
calculated_factor.set_index('code', inplace=True)
calculated_factor = calculated_factor.loc[df.code]
np.testing.assert_array_almost_equal(calculated_factor.dx.values,
df.ROE.values)
def prepare_data(engine: SqlEngine,
factors: Union[Transformer, Iterable[object]],
start_date: str,
end_date: str,
frequency: str,
universe: Universe,
benchmark: int,
warm_start: int = 0):
if warm_start > 0:
p = Period(frequency)
p = Period(length=-warm_start * p.length(), units=p.units())
start_date = advanceDateByCalendar('china.sse', start_date,
p).strftime('%Y-%m-%d')
dates = makeSchedule(start_date,
end_date,
frequency,
calendar='china.sse',
dateRule=BizDayConventions.Following,
dateGenerationRule=DateGeneration.Forward)
dates = [d.strftime('%Y-%m-%d') for d in dates]
horizon = map_freq(frequency)
if isinstance(factors, Transformer):
transformer = factors
else:
transformer = Transformer(factors)
factor_df = engine.fetch_factor_range(universe,
factors=transformer,
dates=dates).sort_values(
['trade_date', 'code'])
alpha_logger.info("factor data loading finished")
return_df = engine.fetch_dx_return_range(universe,
dates=dates,
horizon=horizon)
alpha_logger.info("return data loading finished")
industry_df = engine.fetch_industry_range(universe, dates=dates)
alpha_logger.info("industry data loading finished")
benchmark_df = engine.fetch_benchmark_range(benchmark, dates=dates)
alpha_logger.info("benchmark data loading finished")
df = pd.merge(factor_df, return_df, on=['trade_date', 'code']).dropna()
df = pd.merge(df, benchmark_df, on=['trade_date', 'code'], how='left')
df = pd.merge(df, industry_df, on=['trade_date', 'code'])
df['weight'] = df['weight'].fillna(0.)
return dates, df[['trade_date', 'code', 'dx']], df[[
'trade_date', 'code', 'weight', 'isOpen', 'industry_code', 'industry'
] + transformer.names]
def base(self, table_name, begin_date, end_date, freq=None):
if freq is None:
query = select([table_name]).where(
and_(
table_name.trade_date >= begin_date,
table_name.trade_date <= end_date,
))
else:
rebalance_dates = makeSchedule(begin_date, end_date, freq,
'china.sse',
BizDayConventions.Preceding)
query = select([table_name]).where(
table_name.trade_date.in_(rebalance_dates))
return pd.read_sql(query, self._engine.sql_engine())
def index(self, benchmark, begin_date, end_date, freq=None):
table = importlib.import_module('data.rl_model').Index
if freq is None:
query = select([table]).where(
and_(table.trade_date >= begin_date,
table.trade_date <= end_date,
table.isymbol.in_(benchmark)))
else:
rebalance_dates = makeSchedule(begin_date, end_date, freq,
'china.sse',
BizDayConventions.Preceding)
query = select([table]).where(
and_(table.trade_date.in_(rebalance_dates),
table.isymbol.in_(benchmark)))
return pd.read_sql(query, self._engine.sql_engine()).drop(['id'],
axis=1)
def test_sql_engine_fetch_codes_range(self):
ref_dates = makeSchedule(
advanceDateByCalendar('china.sse', self.ref_date, '-6m'),
self.ref_date, '60b', 'china.sse')
universe = Universe('zz500') + Universe('zz1000')
codes = self.engine.fetch_codes_range(universe, dates=ref_dates)
query = select([UniverseTable.trade_date, UniverseTable.code]).where(
and_(UniverseTable.trade_date.in_(ref_dates),
or_(UniverseTable.zz500 == 1, UniverseTable.zz1000 == 1)))
df = pd.read_sql(query, con=self.engine.engine).sort_values('code')
for ref_date in ref_dates:
calculated_codes = list(
sorted(codes[codes.trade_date == ref_date].code.values))
expected_codes = list(
sorted(df[df.trade_date == ref_date].code.values))
self.assertListEqual(calculated_codes, expected_codes)
def test_sql_engine_fetch_benchmark_range(self):
ref_dates = makeSchedule(
advanceDateByCalendar('china.sse', self.ref_date, '-9m'),
self.ref_date, '60b', 'china.sse')
benchmark = 906
index_data = self.engine.fetch_benchmark_range(benchmark,
dates=ref_dates)
query = select([
IndexComponent.trade_date, IndexComponent.code,
(IndexComponent.weight / 100.).label('weight')
]).where(
and_(IndexComponent.trade_date.in_(ref_dates),
IndexComponent.indexCode == benchmark))
df = pd.read_sql(query, con=self.engine.engine)
for ref_date in ref_dates:
calculated_data = index_data[index_data.trade_date == ref_date]
expected_data = df[df.trade_date == ref_date]
np.testing.assert_array_almost_equal(calculated_data.weight.values,
expected_data.weight.values)
def fetch_factors(self, begin_date, end_date, freq=None):
if freq is None:
query = select([self._table]).where(
and_(
self._table.trade_date >= begin_date,
self._table.trade_date <= end_date,
))
else:
rebalance_dates = makeSchedule(begin_date, end_date, freq,
'china.sse',
BizDayConventions.Preceding)
query = select([self._table]).where(
and_(self._table.trade_date.in_(rebalance_dates), ))
data = pd.read_sql(query, self._engine)
for col in ['id', 'creat_time', 'update_time']:
if col in data.columns:
data = data.drop([col], axis=1)
return data
def test_sql_engine_fetch_factor_range(self):
ref_dates = makeSchedule(
advanceDateByCalendar('china.sse', self.ref_date, '-6m'),
self.ref_date, '60b', 'china.sse')
universe = Universe('zz500') + Universe('zz1000')
factor = 'ROE'
factor_data = self.engine.fetch_factor_range(universe,
factor,
dates=ref_dates)
codes = self.engine.fetch_codes_range(universe, dates=ref_dates)
groups = codes.groupby('trade_date')
for ref_date, g in groups:
query = select([Uqer.code, Uqer.ROE]).where(
and_(Uqer.trade_date == ref_date,
Uqer.code.in_(g.code.unique().tolist())))
df = pd.read_sql(query, con=self.engine.engine)
calculated_factor = factor_data[factor_data.trade_date == ref_date]
np.testing.assert_array_almost_equal(calculated_factor.ROE.values,
df.ROE.values)
def test_sql_engine_fetch_dx_return_index_range(self):
ref_dates = makeSchedule(
advanceDateByCalendar('china.sse', self.ref_date, '-6m'),
self.ref_date, '60b', 'china.sse')
index_code = 906
dx_return = self.engine.fetch_dx_return_index_range(index_code,
dates=ref_dates,
horizon=4,
offset=1)
for ref_date in ref_dates:
start_date = advanceDateByCalendar('china.sse', ref_date, '2b')
end_date = advanceDateByCalendar('china.sse', ref_date, '6b')
query = select([IndexMarket.indexCode, IndexMarket.chgPct]).where(
and_(IndexMarket.trade_date.between(start_date, end_date),
IndexMarket.indexCode == index_code))
df = pd.read_sql(query, con=self.engine.engine)
res = df.groupby('indexCode').apply(lambda x: np.log(1. + x).sum())
calculated_return = dx_return[dx_return.trade_date == ref_date]
np.testing.assert_array_almost_equal(calculated_return.dx.values,
res.chgPct.values)
def fetch_predict_phase(engine,
alpha_factors: Iterable[object],
ref_date,
frequency,
universe,
batch,
neutralized_risk: Iterable[str] = None,
risk_model: str = 'short',
pre_process: Iterable[object] = None,
post_process: Iterable[object] = None,
warm_start: int = 0):
transformer = Transformer(alpha_factors)
p = Period(frequency)
p = Period(length=-(warm_start + batch) * p.length(), units=p.units())
start_date = advanceDateByCalendar('china.sse', ref_date, p,
BizDayConventions.Following)
dates = makeSchedule(start_date,
ref_date,
frequency,
calendar='china.sse',
dateRule=BizDayConventions.Following,
dateGenerationRule=DateGeneration.Backward)
factor_df = engine.fetch_factor_range(universe,
factors=transformer,
dates=dates).dropna()
names = transformer.names
if neutralized_risk:
risk_df = engine.fetch_risk_model_range(universe,
dates=dates,
risk_model=risk_model)[1]
used_neutralized_risk = list(set(neutralized_risk).difference(names))
risk_df = risk_df[['trade_date', 'code'] +
used_neutralized_risk].dropna()
train_x = pd.merge(factor_df, risk_df, on=['trade_date', 'code'])
risk_exp = train_x[neutralized_risk].values.astype(float)
x_values = train_x[names].values.astype(float)
else:
train_x = factor_df.copy()
risk_exp = None
date_label = pd.DatetimeIndex(factor_df.trade_date).to_pydatetime()
dates = np.unique(date_label)
if dates[-1] == dt.datetime.strptime(ref_date, '%Y-%m-%d'):
end = dates[-1]
start = dates[-batch]
# index = (date_label >= start) & (date_label <= end)
left_index = bisect.bisect_left(date_label, start)
right_index = bisect.bisect_right(date_label, end)
this_raw_x = x_values[left_index:right_index]
sub_dates = date_label[left_index:right_index]
if risk_exp is not None:
this_risk_exp = risk_exp[left_index:right_index]
else:
this_risk_exp = None
ne_x = factor_processing(this_raw_x,
pre_process=pre_process,
risk_factors=this_risk_exp,
post_process=post_process)
inner_left_index = bisect.bisect_left(sub_dates, end)
inner_right_index = bisect.bisect_right(sub_dates, end)
ne_x = ne_x[inner_left_index:inner_right_index]
left_index = bisect.bisect_left(date_label, end)
right_index = bisect.bisect_right(date_label, end)
codes = train_x.code.values[left_index:right_index]
else:
ne_x = None
codes = None
ret = dict()
ret['x_names'] = transformer.names
ret['predict'] = {'x': ne_x, 'code': codes}
return ret
def fetch_train_phase(engine,
alpha_factors: Iterable[object],
ref_date,
frequency,
universe,
batch,
neutralized_risk: Iterable[str] = None,
risk_model: str = 'short',
pre_process: Iterable[object] = None,
post_process: Iterable[object] = None,
warm_start: int = 0) -> dict:
transformer = Transformer(alpha_factors)
p = Period(frequency)
p = Period(length=-(warm_start + batch + 1) * p.length(), units=p.units())
start_date = advanceDateByCalendar('china.sse', ref_date, p,
BizDayConventions.Following)
dates = makeSchedule(start_date,
ref_date,
frequency,
calendar='china.sse',
dateRule=BizDayConventions.Following,
dateGenerationRule=DateGeneration.Backward)
horizon = _map_horizon(frequency)
factor_df = engine.fetch_factor_range(universe,
factors=transformer,
dates=dates)
return_df = engine.fetch_dx_return_range(universe,
dates=dates,
horizon=horizon)
df = pd.merge(factor_df, return_df, on=['trade_date', 'code']).dropna()
return_df, factor_df = df[['trade_date', 'code',
'dx']], df[['trade_date', 'code', 'isOpen'] +
transformer.names]
return_df, dates, date_label, risk_exp, x_values, y_values, _, _ = \
_merge_df(engine, transformer.names, factor_df, return_df, universe, dates, risk_model, neutralized_risk)
if dates[-1] == dt.datetime.strptime(ref_date, '%Y-%m-%d'):
end = dates[-2]
start = dates[-batch - 1]
else:
end = dates[-1]
start = dates[-batch]
index = (date_label >= start) & (date_label <= end)
this_raw_x = x_values[index]
this_raw_y = y_values[index]
if risk_exp is not None:
this_risk_exp = risk_exp[index]
else:
this_risk_exp = None
ne_x = factor_processing(this_raw_x,
pre_process=pre_process,
risk_factors=this_risk_exp,
post_process=post_process)
ne_y = factor_processing(this_raw_y,
pre_process=pre_process,
risk_factors=this_risk_exp,
post_process=post_process)
ret = dict()
ret['x_names'] = transformer.names
ret['train'] = {'x': ne_x, 'y': ne_y}
return ret
if __name__ == "__main__":
from PyFin.api import makeSchedule
# db_url = "mysql+mysqldb://reader:Reader#[email protected]:13317/vision?charset=utf8"
db_url = "mysql+mysqldb://dxrw:[email protected]:13317/dxtest?charset=utf8"
sql_engine = SqlEngine(db_url=db_url, factor_tables=["factor_momentum"])
universe = Universe("hs300")
start_date = '2020-01-02'
end_date = '2020-02-21'
frequency = "10b"
benchmark = 300
factors = ["EMA5D", "EMV6D"]
ref_dates = makeSchedule(start_date, end_date, frequency, 'china.sse')
print(ref_dates)
df = sql_engine.fetch_factor("2020-02-21",
factors=factors,
codes=["2010031963"])
print(df)
df = sql_engine.fetch_factor_range(universe=universe,
dates=ref_dates,
factors=factors)
print(df)
df = sql_engine.fetch_codes_range(start_date=start_date,
end_date=end_date,
universe=Universe("hs300"))
print(df)
df = sql_engine.fetch_dx_return("2020-10-09",
codes=["2010031963"],
def prepare_data(engine: SqlEngine,
factors: Union[Transformer, Iterable[object]],
start_date: str,
end_date: str,
frequency: str,
universe: Universe,
benchmark: int,
warm_start: int = 0,
fit_target: Union[Transformer, object] = None):
if warm_start > 0:
p = Period(frequency)
p = Period(length=-warm_start * p.length(), units=p.units())
start_date = advanceDateByCalendar('china.sse', start_date,
p).strftime('%Y-%m-%d')
dates = makeSchedule(start_date,
end_date,
frequency,
calendar='china.sse',
dateRule=BizDayConventions.Following,
dateGenerationRule=DateGeneration.Forward)
dates = [d.strftime('%Y-%m-%d') for d in dates]
horizon = map_freq(frequency)
if isinstance(factors, Transformer):
transformer = factors
else:
transformer = Transformer(factors)
factor_df = engine.fetch_factor_range(universe,
factors=transformer,
dates=dates).sort_values(
['trade_date', 'code'])
alpha_logger.info("factor data loading finished")
if fit_target is None:
target_df = engine.fetch_dx_return_range(universe,
dates=dates,
horizon=horizon)
else:
one_more_date = advanceDateByCalendar('china.sse', dates[-1],
frequency)
target_df = engine.fetch_factor_range_forward(universe,
factors=fit_target,
dates=dates +
[one_more_date])
target_df = target_df[target_df.trade_date.isin(dates)]
target_df = target_df.groupby('code').apply(
lambda x: x.fillna(method='pad'))
alpha_logger.info("fit target data loading finished")
industry_df = engine.fetch_industry_range(universe, dates=dates)
alpha_logger.info("industry data loading finished")
benchmark_df = engine.fetch_benchmark_range(benchmark, dates=dates)
alpha_logger.info("benchmark data loading finished")
df = pd.merge(factor_df, target_df, on=['trade_date', 'code']).dropna()
df = pd.merge(df, benchmark_df, on=['trade_date', 'code'], how='left')
df = pd.merge(df, industry_df, on=['trade_date', 'code'])
df['weight'] = df['weight'].fillna(0.)
df.dropna(inplace=True)
return dates, df[[
'trade_date', 'code', 'dx'
]], df[['trade_date', 'code', 'weight', 'industry_code', 'industry'] +
transformer.names]
def fetch_predict_phase(engine,
alpha_factors: Union[Transformer, Iterable[object]],
ref_date,
frequency,
universe,
batch=1,
neutralized_risk: Iterable[str] = None,
risk_model: str = 'short',
pre_process: Iterable[object] = None,
post_process: Iterable[object] = None,
warm_start: int = 0,
fillna: str = None,
fit_target: Union[Transformer, object] = None):
if isinstance(alpha_factors, Transformer):
transformer = alpha_factors
else:
transformer = Transformer(alpha_factors)
p = Period(frequency)
p = Period(length=-(warm_start + batch - 1) * p.length(), units=p.units())
start_date = advanceDateByCalendar('china.sse', ref_date, p,
BizDayConventions.Following)
dates = makeSchedule(start_date,
ref_date,
frequency,
calendar='china.sse',
dateRule=BizDayConventions.Following,
dateGenerationRule=DateGeneration.Backward)
horizon = map_freq(frequency)
factor_df = engine.fetch_factor_range(universe,
factors=transformer,
dates=dates)
if fillna:
factor_df = factor_df.groupby('trade_date').apply(
lambda x: x.fillna(x.median())).reset_index(drop=True).dropna()
else:
factor_df = factor_df.dropna()
if fit_target is None:
target_df = engine.fetch_dx_return_range(universe,
dates=dates,
horizon=horizon)
else:
one_more_date = advanceDateByCalendar('china.sse', dates[-1],
frequency)
target_df = engine.fetch_factor_range_forward(universe,
factors=fit_target,
dates=dates +
[one_more_date])
target_df = target_df[target_df.trade_date.isin(dates)]
target_df = target_df.groupby('code').apply(
lambda x: x.fillna(method='pad'))
names = transformer.names
if neutralized_risk:
risk_df = engine.fetch_risk_model_range(universe,
dates=dates,
risk_model=risk_model)[1]
used_neutralized_risk = list(set(neutralized_risk).difference(names))
risk_df = risk_df[['trade_date', 'code'] +
used_neutralized_risk].dropna()
train_x = pd.merge(factor_df, risk_df, on=['trade_date', 'code'])
train_x = pd.merge(train_x,
target_df,
on=['trade_date', 'code'],
how='left')
risk_exp = train_x[neutralized_risk].values.astype(float)
else:
train_x = pd.merge(factor_df,
target_df,
on=['trade_date', 'code'],
how='left')
risk_exp = None
train_x.dropna(inplace=True, subset=train_x.columns[:-1])
x_values = train_x[names].values.astype(float)
y_values = train_x[['dx']].values.astype(float)
date_label = pd.DatetimeIndex(train_x.trade_date).to_pydatetime()
dates = np.unique(date_label)
if dates[-1] == dt.datetime.strptime(ref_date, '%Y-%m-%d'):
end = dates[-1]
start = dates[-batch] if batch <= len(dates) else dates[0]
left_index = bisect.bisect_left(date_label, start)
right_index = bisect.bisect_right(date_label, end)
this_raw_x = x_values[left_index:right_index]
this_raw_y = y_values[left_index:right_index]
sub_dates = date_label[left_index:right_index]
if risk_exp is not None:
this_risk_exp = risk_exp[left_index:right_index]
else:
this_risk_exp = None
ne_x = factor_processing(this_raw_x,
pre_process=pre_process,
risk_factors=this_risk_exp,
post_process=post_process)
ne_y = factor_processing(this_raw_y,
pre_process=pre_process,
risk_factors=this_risk_exp,
post_process=post_process)
inner_left_index = bisect.bisect_left(sub_dates, end)
inner_right_index = bisect.bisect_right(sub_dates, end)
ne_x = ne_x[inner_left_index:inner_right_index]
ne_y = ne_y[inner_left_index:inner_right_index]
left_index = bisect.bisect_left(date_label, end)
right_index = bisect.bisect_right(date_label, end)
codes = train_x.code.values[left_index:right_index]
else:
ne_x = None
ne_y = None
codes = None
ret = dict()
ret['x_names'] = transformer.names
ret['predict'] = {
'x': pd.DataFrame(ne_x, columns=transformer.names),
'code': codes,
'y': ne_y.flatten()
}
return ret
def fetch_train_phase(engine,
alpha_factors: Union[Transformer, Iterable[object]],
ref_date,
frequency,
universe,
batch=1,
neutralized_risk: Iterable[str] = None,
risk_model: str = 'short',
pre_process: Iterable[object] = None,
post_process: Iterable[object] = None,
warm_start: int = 0,
fit_target: Union[Transformer, object] = None) -> dict:
if isinstance(alpha_factors, Transformer):
transformer = alpha_factors
else:
transformer = Transformer(alpha_factors)
p = Period(frequency)
p = Period(length=-(warm_start + batch) * p.length(), units=p.units())
start_date = advanceDateByCalendar('china.sse', ref_date, p,
BizDayConventions.Following)
dates = makeSchedule(start_date,
ref_date,
frequency,
calendar='china.sse',
dateRule=BizDayConventions.Following,
dateGenerationRule=DateGeneration.Backward)
horizon = map_freq(frequency)
factor_df = engine.fetch_factor_range(universe,
factors=transformer,
dates=dates)
if fit_target is None:
target_df = engine.fetch_dx_return_range(universe,
dates=dates,
horizon=horizon)
else:
one_more_date = advanceDateByCalendar('china.sse', dates[-1],
frequency)
target_df = engine.fetch_factor_range_forward(universe,
factors=fit_target,
dates=dates +
[one_more_date])
target_df = target_df[target_df.trade_date.isin(dates)]
target_df = target_df.groupby('code').apply(
lambda x: x.fillna(method='pad'))
df = pd.merge(factor_df, target_df, on=['trade_date', 'code']).dropna()
target_df, factor_df = df[['trade_date', 'code',
'dx']], df[['trade_date', 'code'] +
transformer.names]
target_df, dates, date_label, risk_exp, x_values, y_values, _, _, codes = \
_merge_df(engine, transformer.names, factor_df, target_df, universe, dates, risk_model,
neutralized_risk)
if dates[-1] == dt.datetime.strptime(ref_date, '%Y-%m-%d'):
pyFinAssert(
len(dates) >= 2, ValueError,
"No previous data for training for the date {0}".format(ref_date))
end = dates[-2]
start = dates[-batch - 1] if batch <= len(dates) - 1 else dates[0]
else:
end = dates[-1]
start = dates[-batch] if batch <= len(dates) else dates[0]
index = (date_label >= start) & (date_label <= end)
this_raw_x = x_values[index]
this_raw_y = y_values[index]
this_code = codes[index]
if risk_exp is not None:
this_risk_exp = risk_exp[index]
else:
this_risk_exp = None
ne_x = factor_processing(this_raw_x,
pre_process=pre_process,
risk_factors=this_risk_exp,
post_process=post_process)
ne_y = factor_processing(this_raw_y,
pre_process=pre_process,
risk_factors=this_risk_exp,
post_process=post_process)
ret = dict()
ret['x_names'] = transformer.names
ret['train'] = {
'x': pd.DataFrame(ne_x, columns=transformer.names),
'y': ne_y,
'code': this_code
}
return ret
