made risk_exp can be unused

alphamind/data/engines/sqlengine.py

@@ -323,7 +323,7 @@ class SqlEngine(object):
 
         query = select(
             [FullFactor.trade_date, FullFactor.code, FullFactor.isOpen] + list(factor_cols.keys())) \
-            .select_from(big_table)
+            .select_from(big_table).distinct()
 
         df = pd.read_sql(query, self.engine).sort_values(['trade_date', 'code'])
 
@@ -387,7 +387,7 @@ class SqlEngine(object):
 
         risk_exposure_cols = [FullFactor.__table__.columns[f] for f in total_risk_factors if f not in set(excluded)]
         query = select([FullFactor.code, special_risk_col] + risk_exposure_cols) \
-            .where(and_(FullFactor.trade_date == ref_date, FullFactor.code.in_(codes)))
+            .where(and_(FullFactor.trade_date == ref_date, FullFactor.code.in_(codes))).distinct()
 
         risk_exp = pd.read_sql(query, self.engine)
 
@@ -429,7 +429,7 @@ class SqlEngine(object):
 
         query = select(
             [FullFactor.trade_date, FullFactor.code, special_risk_col] + risk_exposure_cols) \
-            .select_from(big_table)
+            .select_from(big_table).distinct()
 
         risk_exp = pd.read_sql(query, self.engine)
 
@@ -450,7 +450,7 @@ class SqlEngine(object):
                 Industry.code.in_(codes),
                 Industry.industry == industry_category_name
             )
-        )
+        ).distinct()
 
         return pd.read_sql(query, self.engine)
 
@@ -481,7 +481,7 @@ class SqlEngine(object):
         query = select([Industry.trade_date,
                         Industry.code,
                         Industry.industryID1.label('industry_code'),
-                        Industry.industryName1.label('industry')]).select_from(big_table)
+                        Industry.industryName1.label('industry')]).select_from(big_table).distinct()
         return pd.read_sql(query, self.engine)
 
     def fetch_data(self, ref_date: str,

alphamind/examples/combined_model_training.py

@@ -17,11 +17,11 @@ plt.style.use('ggplot')
 Back test parameter settings
 """
 
-start_date = '2012-01-01'
-end_date = '2017-11-07'
+start_date = '2017-01-01'
+end_date = '2017-11-08'
 benchmark_code = 300
-universe_name = 'zz800'
-universe = Universe(universe_name, [universe_name])
+universe_name = ['zz500', 'hs300', 'sh50']
+universe = Universe(universe_name, universe_name)
 frequency = '1w'
 batch = 8
 method = 'risk_neutral'
@@ -85,7 +85,6 @@ const_model_factor_data = engine.fetch_data_range(universe,
                                                   total_features,
                                                   dates=ref_dates,
                                                   benchmark=benchmark_code)['factor']
-
 """
 Training phase
 """
@@ -101,7 +100,6 @@ for ref_date in ref_dates:
     models_series.loc[ref_date] = model
     alpha_logger.info('trade_date: {0} training finished'.format(ref_date))
 
-
 """
 Predicting and re-balance phase
 """