ENH: Adds last_available_{session, minute} args to DataPortal (#1528)

This allows optionally setting the last available dts in the DataPortal
explicitly. If these args aren't provided, we fall back to inferring
these from the underlying readers, which was the previous behavior.

tests/test_data_portal.py

@@ -23,13 +23,14 @@ from zipline.assets import Equity
 from zipline.testing.fixtures import (
     ZiplineTestCase,
     WithTradingSessions,
-    WithDataPortal
+    WithDataPortal,
+    alias,
 )
 
 
-class TestDataPortal(WithDataPortal,
-                     WithTradingSessions,
-                     ZiplineTestCase):
+class DataPortalTestBase(WithDataPortal,
+                         WithTradingSessions,
+                         ZiplineTestCase):
 
     ASSET_FINDER_EQUITY_SIDS = (1,)
     START_DATE = pd.Timestamp('2016-08-01')
@@ -374,3 +375,13 @@ class TestDataPortal(WithDataPortal,
         self.assertEqual(minutes[4], result,
                          "Asset 10000 had a trade on fourth minute, so should "
                          "return that as the last trade on the fifth.")
+
+
+class TestDataPortal(DataPortalTestBase):
+    DATA_PORTAL_LAST_AVAILABLE_SESSION = None
+    DATA_PORTAL_LAST_AVAILABLE_MINUTE = None
+
+
+class TestDataPortalExplicitLastAvailable(DataPortalTestBase):
+    DATA_PORTAL_LAST_AVAILABLE_SESSION = alias('START_DATE')
+    DATA_PORTAL_LAST_AVAILABLE_MINUTE = alias('END_DATE')

zipline/data/data_portal.py

@@ -104,6 +104,10 @@ class DataPortal(object):
     adjustment_reader : SQLiteAdjustmentWriter, optional
         The adjustment reader. This is used to apply splits, dividends, and
         other adjustment data to the raw data from the readers.
+    last_available_session : pd.Timestamp, optional
+        The last session to make available in session-level data.
+    last_available_minute : pd.Timestamp, optional
+        The last minute to make available in minute-level data.
     """
     def __init__(self,
                  asset_finder,
@@ -113,7 +117,9 @@ class DataPortal(object):
                  equity_minute_reader=None,
                  future_daily_reader=None,
                  future_minute_reader=None,
-                 adjustment_reader=None):
+                 adjustment_reader=None,
+                 last_available_session=None,
+                 last_available_minute=None):
 
         self.trading_calendar = trading_calendar
         self.asset_finder = asset_finder
@@ -169,12 +175,14 @@ class DataPortal(object):
             self.trading_calendar,
             self.asset_finder,
             aligned_minute_readers,
+            last_available_minute,
         )
 
         _dispatch_session_reader = AssetDispatchSessionBarReader(
             self.trading_calendar,
             self.asset_finder,
             aligned_session_readers,
+            last_available_session,
         )
 
         self._pricing_readers = {

zipline/data/dispatch_bar_reader.py

@@ -35,11 +35,21 @@ class AssetDispatchBarReader(with_metaclass(ABCMeta)):
     - readers : dict
         A dict mapping Asset type to the corresponding
         [Minute|Session]BarReader
+    - last_available_dt : pd.Timestamp or None, optional
+        If not provided, infers it by using the min of the
+        last_available_dt values of the underlying readers.
     """
-    def __init__(self, trading_calendar, asset_finder, readers):
+    def __init__(
+        self,
+        trading_calendar,
+        asset_finder,
+        readers,
+        last_available_dt=None,
+    ):
         self._trading_calendar = trading_calendar
         self._asset_finder = asset_finder
         self._readers = readers
+        self._last_available_dt = last_available_dt
 
         for t, r in iteritems(self._readers):
             assert trading_calendar == r.trading_calendar, \
@@ -72,7 +82,10 @@ class AssetDispatchBarReader(with_metaclass(ABCMeta)):
 
     @lazyval
     def last_available_dt(self):
-        return min(r.last_available_dt for r in self._readers.values())
+        if self._last_available_dt is not None:
+            return self._last_available_dt
+        else:
+            return min(r.last_available_dt for r in self._readers.values())
 
     @lazyval
     def first_trading_day(self):

zipline/testing/fixtures.py

@@ -1265,6 +1265,9 @@ class WithDataPortal(WithAdjustmentReader,
 
     DATA_PORTAL_FIRST_TRADING_DAY = None
 
+    DATA_PORTAL_LAST_AVAILABLE_SESSION = None
+    DATA_PORTAL_LAST_AVAILABLE_MINUTE = None
+
     def make_data_portal(self):
         if self.DATA_PORTAL_FIRST_TRADING_DAY is None:
             if self.DATA_PORTAL_USE_MINUTE_DATA:
@@ -1300,6 +1303,8 @@ class WithDataPortal(WithAdjustmentReader,
                 if self.DATA_PORTAL_USE_MINUTE_DATA else
                 None
             ),
+            last_available_session=self.DATA_PORTAL_LAST_AVAILABLE_SESSION,
+            last_available_minute=self.DATA_PORTAL_LAST_AVAILABLE_MINUTE,
         )
 
     def init_instance_fixtures(self):