BUG: Correct Timestamp localization with tz near DST (#11481)

doc/source/whatsnew/v0.20.0.txt

@@ -1124,6 +1124,7 @@ Conversion
 - Bug in ``Timestamp.replace`` now raises ``TypeError`` when incorrect argument names are given; previously this raised ``ValueError`` (:issue:`15240`)
 - Bug in ``Timestamp.replace`` with compat for passing long integers (:issue:`15030`)
 - Bug in ``Timestamp`` returning UTC based time/date attributes when a timezone was provided (:issue:`13303`)
+- Bug in ``Timestamp`` incorrectly localizing timezones during construction (:issue:`11481`, :issue:`15777`)
 - Bug in ``TimedeltaIndex`` addition where overflow was being allowed without error (:issue:`14816`)
 - Bug in ``TimedeltaIndex`` raising a ``ValueError`` when boolean indexing with ``loc`` (:issue:`14946`)
 - Bug in catching an overflow in ``Timestamp`` + ``Timedelta/Offset`` operations (:issue:`15126`)

pandas/_libs/tslib.pyx

@@ -1569,7 +1569,9 @@ cpdef convert_str_to_tsobject(object ts, object tz, object unit,
                 ts = obj.value
                 if tz is not None:
                     # shift for _localize_tso
-                    ts = tz_convert_single(ts, tz, 'UTC')
+                    ts = tz_localize_to_utc(np.array([ts], dtype='i8'), tz,
+                                            ambiguous='raise',
+                                            errors='raise')[0]
         except ValueError:
             try:
                 ts = parse_datetime_string(
@@ -4073,7 +4075,23 @@ except:
     have_pytz = False
 
 
+@cython.boundscheck(False)
+@cython.wraparound(False)
 def tz_convert(ndarray[int64_t] vals, object tz1, object tz2):
+    """
+    Convert the values (in i8) from timezone1 to timezone2
+
+    Parameters
+    ----------
+    vals : int64 ndarray
+    tz1 : string / timezone object
+    tz2 : string / timezone object
+
+    Returns
+    -------
+    int64 ndarray of converted
+    """
+
     cdef:
         ndarray[int64_t] utc_dates, tt, result, trans, deltas
         Py_ssize_t i, j, pos, n = len(vals)
@@ -4175,6 +4193,23 @@ def tz_convert(ndarray[int64_t] vals, object tz1, object tz2):
 
 
 def tz_convert_single(int64_t val, object tz1, object tz2):
+    """
+    Convert the val (in i8) from timezone1 to timezone2
+
+    This is a single timezone versoin of tz_convert
+
+    Parameters
+    ----------
+    val : int64
+    tz1 : string / timezone object
+    tz2 : string / timezone object
+
+    Returns
+    -------
+    int64 converted
+
+    """
+
     cdef:
         ndarray[int64_t] trans, deltas
         Py_ssize_t pos
@@ -4374,7 +4409,7 @@ cpdef ndarray _unbox_utcoffsets(object transinfo):
 def tz_localize_to_utc(ndarray[int64_t] vals, object tz, object ambiguous=None,
                        object errors='raise'):
     """
-    Localize tzinfo-naive DateRange to given time zone (using pytz). If
+    Localize tzinfo-naive i8 to given time zone (using pytz). If
     there are ambiguities in the values, raise AmbiguousTimeError.
 
     Returns
@@ -4546,6 +4581,7 @@ def tz_localize_to_utc(ndarray[int64_t] vals, object tz, object ambiguous=None,
 
     return result
 
+
 cdef inline bisect_right_i8(int64_t *data, int64_t val, Py_ssize_t n):
     cdef Py_ssize_t pivot, left = 0, right = n
 

pandas/tests/series/test_indexing.py

@@ -1024,9 +1024,9 @@ def test_setitem_with_tz_dst(self):
         # scalar
         s = orig.copy()
         s[1] = pd.Timestamp('2011-01-01', tz=tz)
-        exp = pd.Series([pd.Timestamp('2016-11-06 00:00', tz=tz),
-                         pd.Timestamp('2011-01-01 00:00', tz=tz),
-                         pd.Timestamp('2016-11-06 02:00', tz=tz)])
+        exp = pd.Series([pd.Timestamp('2016-11-06 00:00-04:00', tz=tz),
+                         pd.Timestamp('2011-01-01 00:00-05:00', tz=tz),
+                         pd.Timestamp('2016-11-06 01:00-05:00', tz=tz)])
         tm.assert_series_equal(s, exp)
 
         s = orig.copy()

pandas/tests/tseries/test_timezones.py

@@ -1,4 +1,5 @@
 # pylint: disable-msg=E1101,W0612
+import pytest
 import pytz
 import numpy as np
 from distutils.version import LooseVersion
@@ -159,6 +160,52 @@ def test_timestamp_constructed_by_date_and_tz_explicit(self):
         self.assertEqual(result.hour, expected.hour)
         self.assertEqual(result, expected)
 
+    def test_timestamp_constructor_near_dst_boundary(self):
+        # GH 11481 & 15777
+        # Naive string timestamps were being localized incorrectly
+        # with tz_convert_single instead of tz_localize_to_utc
+
+        for tz in ['Europe/Brussels', 'Europe/Prague']:
+            result = Timestamp('2015-10-25 01:00', tz=tz)
+            expected = Timestamp('2015-10-25 01:00').tz_localize(tz)
+            assert result == expected
+
+            with pytest.raises(pytz.AmbiguousTimeError):
+                Timestamp('2015-10-25 02:00', tz=tz)
+
+        result = Timestamp('2017-03-26 01:00', tz='Europe/Paris')
+        expected = Timestamp('2017-03-26 01:00').tz_localize('Europe/Paris')
+        assert result == expected
+
+        with pytest.raises(pytz.NonExistentTimeError):
+            Timestamp('2017-03-26 02:00', tz='Europe/Paris')
+
+        # GH 11708
+        result = to_datetime("2015-11-18 15:30:00+05:30").tz_localize(
+            'UTC').tz_convert('Asia/Kolkata')
+        expected = Timestamp('2015-11-18 15:30:00+0530', tz='Asia/Kolkata')
+        assert result == expected
+
+        # GH 15823
+        result = Timestamp('2017-03-26 00:00', tz='Europe/Paris')
+        expected = Timestamp('2017-03-26 00:00:00+0100', tz='Europe/Paris')
+        assert result == expected
+
+        result = Timestamp('2017-03-26 01:00', tz='Europe/Paris')
+        expected = Timestamp('2017-03-26 01:00:00+0100', tz='Europe/Paris')
+        assert result == expected
+
+        with pytest.raises(pytz.NonExistentTimeError):
+            Timestamp('2017-03-26 02:00', tz='Europe/Paris')
+        result = Timestamp('2017-03-26 02:00:00+0100', tz='Europe/Paris')
+        expected = Timestamp(result.value).tz_localize(
+            'UTC').tz_convert('Europe/Paris')
+        assert result == expected
+
+        result = Timestamp('2017-03-26 03:00', tz='Europe/Paris')
+        expected = Timestamp('2017-03-26 03:00:00+0200', tz='Europe/Paris')
+        assert result == expected
+
     def test_timestamp_to_datetime_tzoffset(self):
         # tzoffset
         from dateutil.tz import tzoffset
@@ -517,8 +564,8 @@ def f():
                                              freq="H"))
         if dateutil.__version__ != LooseVersion('2.6.0'):
             # GH 14621
-            self.assertEqual(times[-1], Timestamp('2013-10-27 01:00', tz=tz,
-                                                  freq="H"))
+            self.assertEqual(times[-1], Timestamp('2013-10-27 01:00:00+0000',
+                                                  tz=tz, freq="H"))
 
     def test_ambiguous_nat(self):
         tz = self.tz('US/Eastern')