Merge pull request #7789 from sinhrks/normalize_pickle

BUG/COMPAT: pickled dtindex with freq raises AttributeError in normalize...

Author: jreback

doc/source/v0.15.0.txt

@@ -189,6 +189,7 @@ Bug Fixes
 - Bug in ``DataFrame.as_matrix()`` with mixed ``datetime64[ns]`` and ``timedelta64[ns]`` dtypes (:issue:`7778`)
 
 
+- Bug in pickles contains ``DateOffset`` may raise ``AttributeError`` when ``normalize`` attribute is reffered internally (:issue:`7748`)
 
 
 

pandas/io/tests/test_pickle.py

@@ -48,9 +48,12 @@ def compare(self, vf):
         # py3 compat when reading py2 pickle
         try:
             data = pandas.read_pickle(vf)
-        except (ValueError) as detail:
-            # trying to read a py3 pickle in py2
-            return
+        except (ValueError) as e:
+            if 'unsupported pickle protocol:' in str(e):
+                # trying to read a py3 pickle in py2
+                return
+            else:
+                raise
 
         for typ, dv in data.items():
             for dt, result in dv.items():
@@ -60,6 +63,7 @@ def compare(self, vf):
                     continue
 
                 self.compare_element(typ, result, expected)
+        return data
 
     def read_pickles(self, version):
         if not is_little_endian():
@@ -68,7 +72,14 @@ def read_pickles(self, version):
         pth = tm.get_data_path('legacy_pickle/{0}'.format(str(version)))
         for f in os.listdir(pth):
             vf = os.path.join(pth,f)
-            self.compare(vf)
+            data = self.compare(vf)
+
+            if data is None:
+                continue
+
+            if 'series' in data:
+                if 'ts' in data['series']:
+                    self._validate_timeseries(data['series']['ts'], self.data['series']['ts'])
 
     def test_read_pickles_0_10_1(self):
         self.read_pickles('0.10.1')
@@ -82,6 +93,9 @@ def test_read_pickles_0_12_0(self):
     def test_read_pickles_0_13_0(self):
         self.read_pickles('0.13.0')
 
+    def test_read_pickles_0_14_0(self):
+        self.read_pickles('0.14.0')
+
     def test_round_trip_current(self):
         for typ, dv in self.data.items():
 
@@ -94,6 +108,14 @@ def test_round_trip_current(self):
                     result = pd.read_pickle(path)
                     self.compare_element(typ, result, expected)
 
+    def _validate_timeseries(self, pickled, current):
+        # GH 7748
+        tm.assert_series_equal(pickled, current)
+        self.assertEqual(pickled.index.freq, current.index.freq)
+        self.assertEqual(pickled.index.freq.normalize, False)
+        self.assert_numpy_array_equal(pickled > 0, current > 0)
+
+
 if __name__ == '__main__':
     import nose
     nose.runmodule(argv=[__file__, '-vvs', '-x', '--pdb', '--pdb-failure'],

pandas/tseries/offsets.py

@@ -130,6 +130,9 @@ def __add__(date):
     _cacheable = False
     _normalize_cache = True
 
+    # default for prior pickles
+    normalize = False
+
     def __init__(self, n=1, normalize=False, **kwds):
         self.n = int(n)
         self.normalize = normalize

setup.py

@@ -578,6 +578,7 @@ def pxd(name):
                                   'tests/data/legacy_pickle/0.11.0/*.pickle',
                                   'tests/data/legacy_pickle/0.12.0/*.pickle',
                                   'tests/data/legacy_pickle/0.13.0/*.pickle',
+                                  'tests/data/legacy_pickle/0.14.0/*.pickle',
                                   'tests/data/*.csv',
                                   'tests/data/*.dta',
                                   'tests/data/*.txt',