CLN/BUG: Consolidate Index.astype and fix tz aware bugs

doc/source/whatsnew/v0.23.0.txt

@@ -288,6 +288,7 @@ Conversion
 - Bug in :class:`Series` constructor with an int or float list where specifying ``dtype=str``, ``dtype='str'`` or ``dtype='U'`` failed to convert the data elements to strings (:issue:`16605`)
 - Bug in :class:`Timestamp` where comparison with an array of ``Timestamp`` objects would result in a ``RecursionError`` (:issue:`15183`)
 - Bug in :class:`WeekOfMonth` and class:`Week` where addition and subtraction did not roll correctly (:issue:`18510`,:issue:`18672`,:issue:`18864`)
+- Bug in :meth:`DatetimeIndex.astype` when converting between timezone aware dtypes, and converting from timezone aware to naive (:issue:`18951`)
 
 
 Indexing

pandas/core/indexes/base.py

@@ -1065,12 +1065,18 @@ def _to_embed(self, keep_tz=False, dtype=None):
 
     @Appender(_index_shared_docs['astype'])
     def astype(self, dtype, copy=True):
-        if is_categorical_dtype(dtype):
+        if is_dtype_equal(self.dtype, dtype):
+            return self.copy() if copy else self
+        elif is_categorical_dtype(dtype):
             from .category import CategoricalIndex
             return CategoricalIndex(self.values, name=self.name, dtype=dtype,
                                     copy=copy)
-        return Index(self.values.astype(dtype, copy=copy), name=self.name,
-                     dtype=dtype)
+        try:
+            return Index(self.values.astype(dtype, copy=copy), name=self.name,
+                         dtype=dtype)
+        except (TypeError, ValueError):
+            msg = 'Cannot cast {name} to dtype {dtype}'
+            raise TypeError(msg.format(name=type(self).__name__, dtype=dtype))
 
     def _to_safe_for_reshape(self):
         """ convert to object if we are a categorical """

pandas/core/indexes/datetimelike.py

@@ -11,13 +11,22 @@
 
 import numpy as np
 from pandas.core.dtypes.common import (
-    is_integer, is_float,
-    is_bool_dtype, _ensure_int64,
-    is_scalar, is_dtype_equal,
-    is_list_like, is_timedelta64_dtype)
+    _ensure_int64,
+    is_dtype_equal,
+    is_float,
+    is_integer,
+    is_list_like,
+    is_scalar,
+    is_bool_dtype,
+    is_categorical_dtype,
+    is_datetime_or_timedelta_dtype,
+    is_float_dtype,
+    is_integer_dtype,
+    is_object_dtype,
+    is_string_dtype,
+    is_timedelta64_dtype)
 from pandas.core.dtypes.generic import (
-    ABCIndex, ABCSeries,
-    ABCPeriodIndex, ABCIndexClass)
+    ABCIndex, ABCSeries, ABCPeriodIndex, ABCIndexClass)
 from pandas.core.dtypes.missing import isna
 from pandas.core import common as com, algorithms
 from pandas.core.algorithms import checked_add_with_arr
@@ -859,6 +868,22 @@ def _concat_same_dtype(self, to_concat, name):
             new_data = np.concatenate([c.asi8 for c in to_concat])
         return self._simple_new(new_data, **attribs)
 
+    def astype(self, dtype, copy=True):
+        if is_object_dtype(dtype):
+            return self._box_values_as_index()
+        elif is_string_dtype(dtype) and not is_categorical_dtype(dtype):
+            return Index(self.format(), name=self.name, dtype=object)
+        elif is_integer_dtype(dtype):
+            return Index(self.values.astype('i8', copy=copy), name=self.name,
+                         dtype='i8')
+        elif (is_datetime_or_timedelta_dtype(dtype) and
+              not is_dtype_equal(self.dtype, dtype)) or is_float_dtype(dtype):
+            # disallow conversion between datetime/timedelta,
+            # and conversions for any datetimelike to float
+            msg = 'Cannot cast {name} to dtype {dtype}'
+            raise TypeError(msg.format(name=type(self).__name__, dtype=dtype))
+        return super(DatetimeIndexOpsMixin, self).astype(dtype, copy=copy)
+
 
 def _ensure_datetimelike_to_i8(other):
     """ helper for coercing an input scalar or array to i8 """

pandas/core/indexes/datetimes.py

@@ -10,17 +10,19 @@
 from pandas.core.base import _shared_docs
 
 from pandas.core.dtypes.common import (
-    _NS_DTYPE, _INT64_DTYPE,
-    is_object_dtype, is_datetime64_dtype,
-    is_datetimetz, is_dtype_equal,
+    _INT64_DTYPE,
+    _NS_DTYPE,
+    is_object_dtype,
+    is_datetime64_dtype,
+    is_datetimetz,
+    is_dtype_equal,
     is_timedelta64_dtype,
-    is_integer, is_float,
+    is_integer,
+    is_float,
     is_integer_dtype,
     is_datetime64_ns_dtype,
     is_period_dtype,
     is_bool_dtype,
-    is_string_dtype,
-    is_categorical_dtype,
     is_string_like,
     is_list_like,
     is_scalar,
@@ -36,20 +38,17 @@
 from pandas.core.algorithms import checked_add_with_arr
 
 from pandas.core.indexes.base import Index, _index_shared_docs
-from pandas.core.indexes.category import CategoricalIndex
 from pandas.core.indexes.numeric import Int64Index, Float64Index
 import pandas.compat as compat
-from pandas.tseries.frequencies import (
-    to_offset, get_period_alias,
-    Resolution)
+from pandas.tseries.frequencies import to_offset, get_period_alias, Resolution
 from pandas.core.indexes.datetimelike import (
     DatelikeOps, TimelikeOps, DatetimeIndexOpsMixin)
 from pandas.tseries.offsets import (
     DateOffset, generate_range, Tick, CDay, prefix_mapping)
 
 from pandas.core.tools.timedeltas import to_timedelta
-from pandas.util._decorators import (Appender, cache_readonly,
-                                     deprecate_kwarg, Substitution)
+from pandas.util._decorators import (
+    Appender, cache_readonly, deprecate_kwarg, Substitution)
 import pandas.core.common as com
 import pandas.tseries.offsets as offsets
 import pandas.core.tools.datetimes as tools
@@ -906,25 +905,16 @@ def _format_native_types(self, na_rep='NaT', date_format=None, **kwargs):
     @Appender(_index_shared_docs['astype'])
     def astype(self, dtype, copy=True):
         dtype = pandas_dtype(dtype)
-        if is_object_dtype(dtype):
-            return self._box_values_as_index()
-        elif is_integer_dtype(dtype):
-            return Index(self.values.astype('i8', copy=copy), name=self.name,
-                         dtype='i8')
-        elif is_datetime64_ns_dtype(dtype):
-            if self.tz is not None:
-                return self.tz_convert('UTC').tz_localize(None)
-            elif copy is True:
-                return self.copy()
-            return self
-        elif is_categorical_dtype(dtype):
-            return CategoricalIndex(self.values, name=self.name, dtype=dtype,
-                                    copy=copy)
-        elif is_string_dtype(dtype):
-            return Index(self.format(), name=self.name, dtype=object)
+        if (is_datetime64_ns_dtype(dtype) and
+                not is_dtype_equal(dtype, self.dtype)):
+            # GH 18951: datetime64_ns dtype but not equal means different tz
+            new_tz = getattr(dtype, 'tz', None)
+            if getattr(self.dtype, 'tz', None) is None:
+                return self.tz_localize(new_tz)
+            return self.tz_convert(new_tz)
         elif is_period_dtype(dtype):
             return self.to_period(freq=dtype.freq)
-        raise TypeError('Cannot cast DatetimeIndex to dtype %s' % dtype)
+        return super(DatetimeIndex, self).astype(dtype, copy=copy)
 
     def _get_time_micros(self):
         values = self.asi8

pandas/core/indexes/interval.py

@@ -12,8 +12,6 @@
     is_datetime_or_timedelta_dtype,
     is_datetime64tz_dtype,
     is_integer_dtype,
-    is_object_dtype,
-    is_categorical_dtype,
     is_float_dtype,
     is_interval_dtype,
     is_scalar,
@@ -29,7 +27,6 @@
     Interval, IntervalMixin, IntervalTree,
     intervals_to_interval_bounds)
 
-from pandas.core.indexes.category import CategoricalIndex
 from pandas.core.indexes.datetimes import date_range
 from pandas.core.indexes.timedeltas import timedelta_range
 from pandas.core.indexes.multi import MultiIndex
@@ -671,16 +668,8 @@ def copy(self, deep=False, name=None):
     @Appender(_index_shared_docs['astype'])
     def astype(self, dtype, copy=True):
         if is_interval_dtype(dtype):
-            if copy:
-                self = self.copy()
-            return self
-        elif is_object_dtype(dtype):
-            return Index(self.values, dtype=object)
-        elif is_categorical_dtype(dtype):
-            return CategoricalIndex(self.values, name=self.name, dtype=dtype,
-                                    copy=copy)
-        raise ValueError('Cannot cast IntervalIndex to dtype {dtype}'
-                         .format(dtype=dtype))
+            return self.copy() if copy else self
+        return super(IntervalIndex, self).astype(dtype, copy=copy)
 
     @cache_readonly
     def dtype(self):

pandas/core/indexes/numeric.py

@@ -4,10 +4,8 @@
 from pandas.core.dtypes.common import (
     is_dtype_equal,
     pandas_dtype,
-    is_float_dtype,
-    is_object_dtype,
+    needs_i8_conversion,
     is_integer_dtype,
-    is_categorical_dtype,
     is_bool,
     is_bool_dtype,
     is_scalar)
@@ -17,7 +15,6 @@
 from pandas.core import algorithms
 from pandas.core.indexes.base import (
     Index, InvalidIndexError, _index_shared_docs)
-from pandas.core.indexes.category import CategoricalIndex
 from pandas.util._decorators import Appender, cache_readonly
 import pandas.core.dtypes.concat as _concat
 import pandas.core.indexes.base as ibase
@@ -315,22 +312,14 @@ def inferred_type(self):
     @Appender(_index_shared_docs['astype'])
     def astype(self, dtype, copy=True):
         dtype = pandas_dtype(dtype)
-        if is_float_dtype(dtype):
-            values = self._values.astype(dtype, copy=copy)
-        elif is_integer_dtype(dtype):
-            if self.hasnans:
-                raise ValueError('cannot convert float NaN to integer')
-            values = self._values.astype(dtype, copy=copy)
-        elif is_object_dtype(dtype):
-            values = self._values.astype('object', copy=copy)
-        elif is_categorical_dtype(dtype):
-            return CategoricalIndex(self, name=self.name, dtype=dtype,
-                                    copy=copy)
-        else:
-            raise TypeError('Setting {cls} dtype to anything other than '
-                            'float64, object, or category is not supported'
-                            .format(cls=self.__class__))
-        return Index(values, name=self.name, dtype=dtype)
+        if needs_i8_conversion(dtype):
+            msg = ('Cannot convert Float64Index to dtype {dtype}; integer '
+                   'values are required for conversion').format(dtype=dtype)
+            raise TypeError(msg)
+        elif is_integer_dtype(dtype) and self.hasnans:
+            # GH 13149
+            raise ValueError('Cannot convert NA to integer')
+        return super(Float64Index, self).astype(dtype, copy=copy)
 
     @Appender(_index_shared_docs['_convert_scalar_indexer'])
     def _convert_scalar_indexer(self, key, kind=None):

pandas/core/indexes/period.py

@@ -7,24 +7,21 @@
 from pandas.core.dtypes.common import (
     is_integer,
     is_float,
-    is_object_dtype,
     is_integer_dtype,
     is_float_dtype,
     is_scalar,
     is_datetime64_dtype,
-    is_datetime64tz_dtype,
+    is_datetime64_any_dtype,
     is_timedelta64_dtype,
     is_period_dtype,
     is_bool_dtype,
-    is_categorical_dtype,
     pandas_dtype,
     _ensure_object)
 from pandas.core.dtypes.dtypes import PeriodDtype
 from pandas.core.dtypes.generic import ABCSeries
 
 import pandas.tseries.frequencies as frequencies
 from pandas.tseries.frequencies import get_freq_code as _gfc
-from pandas.core.indexes.category import CategoricalIndex
 from pandas.core.indexes.datetimes import DatetimeIndex, Int64Index, Index
 from pandas.core.indexes.timedeltas import TimedeltaIndex
 from pandas.core.indexes.datetimelike import DatelikeOps, DatetimeIndexOpsMixin
@@ -506,23 +503,14 @@ def asof_locs(self, where, mask):
     @Appender(_index_shared_docs['astype'])
     def astype(self, dtype, copy=True, how='start'):
         dtype = pandas_dtype(dtype)
-        if is_object_dtype(dtype):
-            return self._box_values_as_index()
-        elif is_integer_dtype(dtype):
-            if copy:
-                return self._int64index.copy()
-            else:
-                return self._int64index
-        elif is_datetime64_dtype(dtype):
-            return self.to_timestamp(how=how)
-        elif is_datetime64tz_dtype(dtype):
-            return self.to_timestamp(how=how).tz_localize(dtype.tz)
+        if is_integer_dtype(dtype):
+            return self._int64index.copy() if copy else self._int64index
+        elif is_datetime64_any_dtype(dtype):
+            tz = getattr(dtype, 'tz', None)
+            return self.to_timestamp(how=how).tz_localize(tz)
         elif is_period_dtype(dtype):
             return self.asfreq(freq=dtype.freq)
-        elif is_categorical_dtype(dtype):
-            return CategoricalIndex(self.values, name=self.name, dtype=dtype,
-                                    copy=copy)
-        raise TypeError('Cannot cast PeriodIndex to dtype %s' % dtype)
+        return super(PeriodIndex, self).astype(dtype, copy=copy)
 
     @Substitution(klass='PeriodIndex')
     @Appender(_shared_docs['searchsorted'])

pandas/core/indexes/timedeltas.py

@@ -4,23 +4,20 @@
 import numpy as np
 from pandas.core.dtypes.common import (
     _TD_DTYPE,
-    is_integer, is_float,
+    is_integer,
+    is_float,
     is_bool_dtype,
     is_list_like,
     is_scalar,
-    is_integer_dtype,
-    is_object_dtype,
     is_timedelta64_dtype,
     is_timedelta64_ns_dtype,
-    is_categorical_dtype,
     pandas_dtype,
     _ensure_int64)
 from pandas.core.dtypes.missing import isna
 from pandas.core.dtypes.generic import ABCSeries
 from pandas.core.common import _maybe_box, _values_from_object
 
 from pandas.core.indexes.base import Index
-from pandas.core.indexes.category import CategoricalIndex
 from pandas.core.indexes.numeric import Int64Index
 import pandas.compat as compat
 from pandas.compat import u
@@ -483,28 +480,14 @@ def to_pytimedelta(self):
     @Appender(_index_shared_docs['astype'])
     def astype(self, dtype, copy=True):
         dtype = pandas_dtype(dtype)
-
-        if is_object_dtype(dtype):
-            return self._box_values_as_index()
-        elif is_timedelta64_ns_dtype(dtype):
-            if copy is True:
-                return self.copy()
-            return self
-        elif is_timedelta64_dtype(dtype):
+        if is_timedelta64_dtype(dtype) and not is_timedelta64_ns_dtype(dtype):
             # return an index (essentially this is division)
             result = self.values.astype(dtype, copy=copy)
             if self.hasnans:
-                return Index(self._maybe_mask_results(result,
-                                                      convert='float64'),
-                             name=self.name)
+                values = self._maybe_mask_results(result, convert='float64')
+                return Index(values, name=self.name)
             return Index(result.astype('i8'), name=self.name)
-        elif is_integer_dtype(dtype):
-            return Index(self.values.astype('i8', copy=copy), dtype='i8',
-                         name=self.name)
-        elif is_categorical_dtype(dtype):
-            return CategoricalIndex(self.values, name=self.name, dtype=dtype,
-                                    copy=copy)
-        raise TypeError('Cannot cast TimedeltaIndex to dtype %s' % dtype)
+        return super(TimedeltaIndex, self).astype(dtype, copy=copy)
 
     def union(self, other):
         """