BUG: Series.ffill() with mixed dtypes containing tz-aware datetimes fails. (GH14956)

Author: Adrian

doc/source/whatsnew/v0.20.0.txt

@@ -300,3 +300,4 @@ Bug Fixes
 - Bug in ``Series.unique()`` in which unsigned 64-bit integers were causing overflow (:issue:`14721`)
 - Require at least 0.23 version of cython to avoid problems with character encodings (:issue:`14699`)
 - Bug in converting object elements of array-like objects to unsigned 64-bit integers (:issue:`4471`)
+- Bug in Series.ffill() with mixed dtypes containing tz-aware datetimes. (:issue:`14956`)

pandas/src/inference.pyx

@@ -859,7 +859,7 @@ def maybe_convert_objects(ndarray[object] objects, bint try_float=0,
 
     # we try to coerce datetime w/tz but must all have the same tz
     if seen_datetimetz:
-        if len(set([ getattr(val, 'tz', None) for val in objects ])) == 1:
+        if len(set([getattr(val, 'tzinfo', None) for val in objects])) == 1:
             from pandas import DatetimeIndex
             return DatetimeIndex(objects)
         seen_object = 1

pandas/tests/series/test_missing.py

@@ -386,6 +386,12 @@ def test_ffill(self):
         ts[2] = np.NaN
         assert_series_equal(ts.ffill(), ts.fillna(method='ffill'))
 
+    def test_ffill_mixed_dtypes_without_missing_data(self):
+        # GH14956
+        series = pd.Series([datetime(2015, 1, 1, tzinfo=pytz.utc), 1])
+        result = series.ffill()
+        assert_series_equal(series, result)
+
     def test_bfill(self):
         ts = Series([0., 1., 2., 3., 4.], index=tm.makeDateIndex(5))
         ts[2] = np.NaN

pandas/tests/types/test_inference.py

@@ -11,6 +11,7 @@
 import re
 from datetime import datetime, date, timedelta, time
 import numpy as np
+import pytz
 
 import pandas as pd
 from pandas import lib, tslib
@@ -275,6 +276,14 @@ def test_maybe_convert_objects_uint64(self):
         exp = np.array([2**63, -1], dtype=object)
         tm.assert_numpy_array_equal(lib.maybe_convert_objects(arr), exp)
 
+    def test_mixed_dtypes_remain_object_array(self):
+        # GH14956
+        array = np.array([datetime(2015, 1, 1, tzinfo=pytz.utc), 1],
+                         dtype=object)
+        result = lib.maybe_convert_objects(array)
+        tm.assert_numpy_array_equal(result, array)
+        self.assertTrue(result.dtype == object)
+
 
 class TestTypeInference(tm.TestCase):
     _multiprocess_can_split_ = True