REF: avoid special-casing Categorical astype

Author: jbrockmendel

pandas/core/arrays/integer.py

@@ -9,7 +9,7 @@
 from pandas.compat.numpy import function as nv
 from pandas.util._decorators import cache_readonly
 
-from pandas.core.dtypes.base import register_extension_dtype
+from pandas.core.dtypes.base import ExtensionDtype, register_extension_dtype
 from pandas.core.dtypes.common import (
     is_bool_dtype,
     is_datetime64_dtype,
@@ -409,6 +409,11 @@ def astype(self, dtype, copy: bool = True) -> ArrayLike:
         elif isinstance(dtype, StringDtype):
             return dtype.construct_array_type()._from_sequence(self, copy=False)
 
+        elif isinstance(dtype, ExtensionDtype):
+            # e.g. Categorical
+            cls = dtype.construct_array_type()
+            return cls._from_sequence(self, dtype=dtype, copy=copy)
+
         # coerce
         if is_float_dtype(dtype):
             # In astype, we consider dtype=float to also mean na_value=np.nan

pandas/core/indexes/category.py

@@ -370,11 +370,6 @@ def __contains__(self, key: Any) -> bool:
 
         return contains(self, key, container=self._engine)
 
-    @doc(Index.astype)
-    def astype(self, dtype, copy=True):
-        res_data = self._data.astype(dtype, copy=copy)
-        return Index(res_data, name=self.name)
-
     @doc(Index.fillna)
     def fillna(self, value, downcast=None):
         value = self._require_scalar(value)

pandas/core/indexes/extension.py

@@ -11,7 +11,7 @@
 from pandas.errors import AbstractMethodError
 from pandas.util._decorators import cache_readonly, doc
 
-from pandas.core.dtypes.common import is_dtype_equal, is_object_dtype
+from pandas.core.dtypes.common import is_dtype_equal, is_object_dtype, pandas_dtype
 from pandas.core.dtypes.generic import ABCDataFrame, ABCSeries
 
 from pandas.core.arrays import ExtensionArray
@@ -294,9 +294,12 @@ def map(self, mapper, na_action=None):
 
     @doc(Index.astype)
     def astype(self, dtype, copy=True):
-        if is_dtype_equal(self.dtype, dtype) and copy is False:
-            # Ensure that self.astype(self.dtype) is self
-            return self
+        dtype = pandas_dtype(dtype)
+        if is_dtype_equal(self.dtype, dtype):
+            if not copy:
+                # Ensure that self.astype(self.dtype) is self
+                return self
+            return self.copy()
 
         new_values = self._data.astype(dtype, copy=copy)
 

pandas/core/internals/blocks.py

@@ -642,15 +642,7 @@ def astype(self, dtype, copy: bool = False, errors: str = "raise"):
     def _astype(self, dtype: DtypeObj, copy: bool) -> ArrayLike:
         values = self.values
 
-        if is_categorical_dtype(dtype):
-
-            if is_categorical_dtype(values.dtype):
-                # GH#10696/GH#18593: update an existing categorical efficiently
-                return values.astype(dtype, copy=copy)
-
-            return Categorical(values, dtype=dtype)
-
-        elif is_datetime64tz_dtype(dtype) and is_datetime64_dtype(values.dtype):
+        if is_datetime64tz_dtype(dtype) and is_datetime64_dtype(values.dtype):
             # if we are passed a datetime64[ns, tz]
             if copy:
                 # this should be the only copy