Skip to content

A
alpha-mind
Commit 9f2cbc30 authored by Dr.李's avatar Dr.李
Browse files
Options

using full factor table

parent 88a49650
Hide whitespace changes
Inline Side-by-side
Showing with 46 additions and 37 deletions
alphamind/data/engines/sqlengine.py
View file @ 9f2cbc30
...@@ -26,7 +26,7 @@ from alphamind.data.dbmodel.models import Experimental ...@@ -26,7 +26,7 @@ from alphamind.data.dbmodel.models import Experimental
from alphamind.data.dbmodel.models import RiskCovDay from alphamind.data.dbmodel.models import RiskCovDay
from alphamind.data.dbmodel.models import RiskCovShort from alphamind.data.dbmodel.models import RiskCovShort
from alphamind.data.dbmodel.models import RiskCovLong from alphamind.data.dbmodel.models import RiskCovLong
from alphamind.data.dbmodel.models import FullFactorView from alphamind.data.dbmodel.models import FullFactor
from alphamind.data.dbmodel.models import Models from alphamind.data.dbmodel.models import Models
from alphamind.data.dbmodel.models import Universe as UniverseTable from alphamind.data.dbmodel.models import Universe as UniverseTable
from alphamind.data.transformer import Transformer from alphamind.data.transformer import Transformer
...@@ -79,18 +79,18 @@ macro_styles = ['COUNTRY'] ...@@ -79,18 +79,18 @@ macro_styles = ['COUNTRY']
total_risk_factors = risk_styles + industry_styles + macro_styles total_risk_factors = risk_styles + industry_styles + macro_styles
factor_tables = [FullFactorView, Experimental] factor_tables = [FullFactor]
DEFAULT_URL = 'postgresql+psycopg2://postgres:A12345678!@10.63.6.220/alpha' DEFAULT_URL = 'postgresql+psycopg2://postgres:A12345678!@10.63.6.220/alpha'
def _map_risk_model_table(risk_model: str) -> tuple: def _map_risk_model_table(risk_model: str) -> tuple:
if risk_model == 'day': if risk_model == 'day':
return RiskCovDay, FullFactorView.d_srisk return RiskCovDay, FullFactor.d_srisk
elif risk_model == 'short': elif risk_model == 'short':
return RiskCovShort, FullFactorView.s_srisk return RiskCovShort, FullFactor.s_srisk
elif risk_model == 'long': elif risk_model == 'long':
return RiskCovLong, FullFactorView.l_srisk return RiskCovLong, FullFactor.l_srisk
else: else:
raise ValueError("risk model name {0} is not recognized".format(risk_model)) raise ValueError("risk model name {0} is not recognized".format(risk_model))
...@@ -246,17 +246,17 @@ class SqlEngine(object): ...@@ -246,17 +246,17 @@ class SqlEngine(object):
start_date = advanceDateByCalendar('china.sse', ref_date, str(-warm_start) + 'b').strftime('%Y-%m-%d') start_date = advanceDateByCalendar('china.sse', ref_date, str(-warm_start) + 'b').strftime('%Y-%m-%d')
end_date = ref_date end_date = ref_date
big_table = FullFactorView big_table = FullFactor
for t in set(factor_cols.values()): for t in set(factor_cols.values()):
if t.__table__.name != FullFactorView.__table__.name: if t.__table__.name != FullFactor.__table__.name:
big_table = outerjoin(big_table, t, and_(FullFactorView.trade_date == t.trade_date, big_table = outerjoin(big_table, t, and_(FullFactor.trade_date == t.trade_date,
FullFactorView.code == t.code)) FullFactor.code == t.code))
query = select( query = select(
[FullFactorView.trade_date, FullFactorView.code, FullFactorView.isOpen] + list(factor_cols.keys())) \ [FullFactor.trade_date, FullFactor.code, FullFactor.isOpen] + list(factor_cols.keys())) \
.select_from(big_table).where(and_(FullFactorView.trade_date.between(start_date, end_date), .select_from(big_table).where(and_(FullFactor.trade_date.between(start_date, end_date),
FullFactorView.code.in_(codes))) FullFactor.code.in_(codes)))
df = pd.read_sql(query, self.engine).sort_values(['trade_date', 'code']).set_index('trade_date') df = pd.read_sql(query, self.engine).sort_values(['trade_date', 'code']).set_index('trade_date')
res = transformer.transform('code', df) res = transformer.transform('code', df)
...@@ -288,26 +288,26 @@ class SqlEngine(object): ...@@ -288,26 +288,26 @@ class SqlEngine(object):
cond = universe.query_range(start_date, end_date, dates) cond = universe.query_range(start_date, end_date, dates)
big_table = FullFactorView big_table = FullFactor
for t in set(factor_cols.values()): for t in set(factor_cols.values()):
if t.__table__.name != FullFactorView.__table__.name: if t.__table__.name != FullFactor.__table__.name:
if dates is not None: if dates is not None:
big_table = outerjoin(big_table, t, and_(FullFactorView.trade_date == t.trade_date, big_table = outerjoin(big_table, t, and_(FullFactor.trade_date == t.trade_date,
FullFactorView.code == t.code, FullFactor.code == t.code,
FullFactorView.trade_date.in_(dates))) FullFactor.trade_date.in_(dates)))
else: else:
big_table = outerjoin(big_table, t, and_(FullFactorView.trade_date == t.trade_date, big_table = outerjoin(big_table, t, and_(FullFactor.trade_date == t.trade_date,
FullFactorView.code == t.code, FullFactor.code == t.code,
FullFactorView.trade_date.between(start_date, end_date))) FullFactor.trade_date.between(start_date, end_date)))
big_table = join(big_table, UniverseTable, big_table = join(big_table, UniverseTable,
and_(FullFactorView.trade_date == UniverseTable.trade_date, and_(FullFactor.trade_date == UniverseTable.trade_date,
FullFactorView.code == UniverseTable.code, FullFactor.code == UniverseTable.code,
cond)) cond))
query = select( query = select(
[FullFactorView.trade_date, FullFactorView.code, FullFactorView.isOpen] + list(factor_cols.keys())) \ [FullFactor.trade_date, FullFactor.code, FullFactor.isOpen] + list(factor_cols.keys())) \
.select_from(big_table) .select_from(big_table)
df = pd.read_sql(query, self.engine).sort_values(['trade_date', 'code']) df = pd.read_sql(query, self.engine).sort_values(['trade_date', 'code'])
...@@ -370,9 +370,9 @@ class SqlEngine(object): ...@@ -370,9 +370,9 @@ class SqlEngine(object):
) )
risk_cov = pd.read_sql(query, self.engine).sort_values('FactorID') risk_cov = pd.read_sql(query, self.engine).sort_values('FactorID')
risk_exposure_cols = [FullFactorView.__table__.columns[f] for f in total_risk_factors if f not in set(excluded)] risk_exposure_cols = [FullFactor.__table__.columns[f] for f in total_risk_factors if f not in set(excluded)]
query = select([FullFactorView.code, special_risk_col] + risk_exposure_cols) \ query = select([FullFactor.code, special_risk_col] + risk_exposure_cols) \
.where(and_(FullFactorView.trade_date == ref_date, FullFactorView.code.in_(codes))) .where(and_(FullFactor.trade_date == ref_date, FullFactor.code.in_(codes)))
risk_exp = pd.read_sql(query, self.engine) risk_exp = pd.read_sql(query, self.engine)
...@@ -404,16 +404,16 @@ class SqlEngine(object): ...@@ -404,16 +404,16 @@ class SqlEngine(object):
if not excluded: if not excluded:
excluded = [] excluded = []
risk_exposure_cols = [FullFactorView.__table__.columns[f] for f in total_risk_factors if f not in set(excluded)] risk_exposure_cols = [FullFactor.__table__.columns[f] for f in total_risk_factors if f not in set(excluded)]
cond = universe.query_range(start_date, end_date, dates) cond = universe.query_range(start_date, end_date, dates)
big_table = join(FullFactorView, UniverseTable, big_table = join(FullFactor, UniverseTable,
and_(FullFactorView.trade_date == UniverseTable.trade_date, and_(FullFactor.trade_date == UniverseTable.trade_date,
FullFactorView.code == UniverseTable.code, FullFactor.code == UniverseTable.code,
cond)) cond))
query = select( query = select(
[FullFactorView.trade_date, FullFactorView.code, special_risk_col] + risk_exposure_cols) \ [FullFactor.trade_date, FullFactor.code, special_risk_col] + risk_exposure_cols) \
.select_from(big_table) .select_from(big_table)
risk_exp = pd.read_sql(query, self.engine) risk_exp = pd.read_sql(query, self.engine)
......
alphamind/model/data_preparing.py
View file @ 9f2cbc30
...@@ -348,12 +348,14 @@ def fetch_predict_phase(engine, ...@@ -348,12 +348,14 @@ def fetch_predict_phase(engine,
end = dates[-1] end = dates[-1]
start = dates[-batch] start = dates[-batch]
index = (date_label >= start) & (date_label <= end) # index = (date_label >= start) & (date_label <= end)
this_raw_x = x_values[index] left_index = bisect.bisect_left(date_label, start)
sub_dates = date_label[index] 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: if risk_exp is not None:
this_risk_exp = risk_exp[index] this_risk_exp = risk_exp[left_index:right_index]
else: else:
this_risk_exp = None this_risk_exp = None
...@@ -362,8 +364,15 @@ def fetch_predict_phase(engine, ...@@ -362,8 +364,15 @@ def fetch_predict_phase(engine,
risk_factors=this_risk_exp, risk_factors=this_risk_exp,
post_process=post_process) post_process=post_process)
ne_x = ne_x[sub_dates == end] inner_left_index = bisect.bisect_left(sub_dates, end)
codes = train_x.code.values[date_label == 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: else:
ne_x = None ne_x = None
codes = None codes = None
......
Markdown is supported
0% or
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment