added data preparing for model training stuff

alphamind/model/preparing.py

+# -*- coding: utf-8 -*-
+"""
+Created on 2017-8-24
+
+@author: cheng.li
+"""
+
+import pandas as pd
+from typing import Iterable
+from typing import Union
+from PyFin.api import makeSchedule
+from PyFin.api import BizDayConventions
+from alphamind.data.transformer import Transformer
+from alphamind.data.engines.sqlengine import SqlEngine
+from alphamind.data.engines.universe import Universe
+
+
+def _map_horizon(sampling: str) -> int:
+    if sampling == '1d':
+        return 0
+    elif sampling == '1w':
+        return 4
+    elif sampling == '1m':
+        return 21
+    elif sampling == '3m':
+        return 62
+    else:
+        raise ValueError('{0} is an unrecognized sampling rule'.format(sampling))
+
+
+def prepare_data(engine: SqlEngine,
+                 factors: Iterable[object],
+                 start_date: str,
+                 end_date: str,
+                 sampling: str,
+                 universe: Universe):
+    dates = makeSchedule(start_date, end_date, sampling, calendar='china.sse', dateRule=BizDayConventions.Following)
+
+    horizon = _map_horizon(sampling)
+
+    transformer = Transformer(factors)
+
+    factor_df = engine.fetch_factor_range(universe, factors=transformer, dates=dates).sort_values(['Date', 'Code'])
+    return_df = engine.fetch_dx_return_range(universe, dates=dates, horizon=horizon)
+
+    df = pd.merge(factor_df, return_df, on=['Date', 'Code'])
+
+    return df[['Date', 'Code', 'dx']], df[['Date', 'Code'] + transformer.names]
+
+
+if __name__ == '__main__':
+    from PyFin.api import *
+    engine = SqlEngine('postgresql+psycopg2://postgres:A12345678!@10.63.6.220/alpha')
+    universe = Universe('zz500', ['zz500'])
+    df1, df2 = prepare_data(engine,
+                            MA(10, 'EPS'),
+                            '2012-01-01',
+                            '2013-01-01',
+                            '1w',
+                            universe)
+
+    print(df1)
+    print(df2)