reorder the priority

alphamind/data/dbmodel/models.py

@@ -36,6 +36,9 @@ class Experimental(Base):
     DROE = Column(Float(53))
     IVR = Column(Float(53))
     ROEAfterNonRecurring = Column(Float(53))
+    EPAfterNonRecurring = Column(Float(53))
+    DROEAfterNonRecurring = Column(Float(53))
+    CFinc1 = Column(Float(53))
 
 
 class FactorMaster(Base):

alphamind/data/engines/sqlengine.py

@@ -84,7 +84,7 @@ macro_styles = ['COUNTRY']
 
 total_risk_factors = risk_styles + industry_styles + macro_styles
 
-factor_tables = [Uqer, Tiny, LegacyFactor, Experimental, RiskExposure, Market]
+factor_tables = [Uqer, LegacyFactor, Tiny,  Experimental, RiskExposure, Market]
 
 
 def append_industry_info(df):
@@ -543,7 +543,7 @@ if __name__ == '__main__':
     data2 = engine.fetch_industry(ref_date=ref_date,
                                   codes=codes)
     data2 = engine.fetch_data_range(universe,
-                                    factors=['EPS'],
+                                    factors=None,
                                     start_date=start_date,
                                     end_date=end_date)
     print(codes)

alphamind/examples/model_training.py

@@ -32,7 +32,8 @@ training     - every 4 week
 engine = SqlEngine('postgresql+psycopg2://postgres:A12345678!@10.63.6.220/alpha')
 universe = Universe('zz500', ['zz500'])
 neutralize_risk = industry_styles
-alpha_factors = ['BDTO', 'RVOL', 'CHV', 'VAL', 'CFinc1'] # risk_styles
+alpha_factors = ['RVOL', 'EPS', 'CFinc1', 'BDTO', 'VAL', 'GREV',
+                 'ROEDiluted']  # ['BDTO', 'RVOL', 'CHV', 'VAL', 'CFinc1'] # risk_styles
 benchmark = 905
 n_bins = 5
 frequency = '1w'
@@ -54,7 +55,8 @@ data_package = fetch_data_package(engine,
                                   batch=batch,
                                   neutralized_risk=neutralize_risk,
                                   pre_process=[winsorize_normal, standardize],
-                                  post_process=[standardize])
+                                  post_process=[standardize],
+                                  warm_start=20)
 
 '''
 training phase: using Linear - regression from scikit-learn
@@ -83,31 +85,73 @@ predicting phase: using trained model on the re-balance dates
 predict_x = data_package['predict']['x']
 settlement = data_package['settlement']
 
-final_res = np.zeros((len(dates), n_bins))
+# final_res = np.zeros((len(dates), n_bins))
+#
+# for i, predict_date in enumerate(dates):
+#     model = model_df[predict_date]
+#     x = predict_x[predict_date]
+#     benchmark_w = settlement[settlement.trade_date == predict_date]['weight'].values
+#     realized_r = settlement[settlement.trade_date == predict_date]['dx'].values
+#
+#     predict_y = model.predict(x)
+#
+#     res = er_quantile_analysis(predict_y,
+#                                n_bins,
+#                                dx_return=realized_r,
+#                                benchmark=benchmark_w)
+#
+#     final_res[i] = res / benchmark_w.sum()
+#     print('trade_date: {0} predicting finished'.format(train_date))
+#
+# last_date = advanceDateByCalendar('china.sse', dates[-1], frequency)
+#
+# df = pd.DataFrame(final_res, index=dates[1:] + [last_date])
+# df.sort_index(inplace=True)
+# df.cumsum().plot()
+# plt.title('Risk style factors model training with Linear Regression from 2012 - 2017')
+# plt.show()
 
+'''
+predicting phase: using trained model on the re-balance dates (optimizing with risk neutral)
+'''
+
+industry_dummies = pd.get_dummies(settlement['industry_code'].values)
+final_res = np.zeros(len(dates))
 
 for i, predict_date in enumerate(dates):
     model = model_df[predict_date]
     x = predict_x[predict_date]
-    benchmark_w = settlement[settlement.trade_date == predict_date]['weight'].values
-    realized_r = settlement[settlement.trade_date == predict_date]['dx'].values
+    cons = Constraints()
+    index = settlement.trade_date == predict_date
+    benchmark_w = settlement[index]['weight'].values
+    realized_r = settlement[index]['dx'].values
+    industry_names = settlement[index]['industry'].values
+    is_tradable = settlement[index]['isOpen'].values
+    ind_exp = industry_dummies[index]
 
-    predict_y = model.predict(x)
+    risk_tags = ind_exp.columns
+    cons.add_exposure(risk_tags, ind_exp.values)
+    benchmark_exp = benchmark_w @ ind_exp.values
 
-    res = er_quantile_analysis(predict_y,
-                               n_bins,
-                               dx_return=realized_r,
-                               benchmark=benchmark_w)
+    for k, name in enumerate(risk_tags):
+        cons.set_constraints(name, benchmark_exp[k], benchmark_exp[k])
 
-    final_res[i] = res / benchmark_w.sum()
-    print('trade_date: {0} predicting finished'.format(train_date))
+    predict_y = model.predict(x)
+    weights, analysis = er_portfolio_analysis(predict_y,
+                                              industry_names,
+                                              realized_r,
+                                              constraints=cons,
+                                              detail_analysis=True,
+                                              benchmark=benchmark_w,
+                                              is_tradable=is_tradable)
+
+    final_res[i] = analysis['er']['total'] / benchmark_w.sum()
 
 last_date = advanceDateByCalendar('china.sse', dates[-1], frequency)
 
-df = pd.DataFrame(final_res, index=dates[1:] + [last_date])
+df = pd.Series(final_res, index=dates[1:] + [last_date])
 df.sort_index(inplace=True)
 df.cumsum().plot()
 plt.title('Risk style factors model training with Linear Regression from 2012 - 2017')
 plt.show()
-df = df.cumsum()
-df.to_csv('d:/20120101_20170823_bt.csv')
+