REF/API: Stricter extension checking.

doc/source/api.rst

@@ -2501,6 +2501,8 @@ Dtype introspection
     api.types.is_datetime64_ns_dtype
     api.types.is_datetime64tz_dtype
     api.types.is_extension_type
+    api.types.is_extension_array
+    api.types.is_extension_dtype
     api.types.is_float_dtype
     api.types.is_int64_dtype
     api.types.is_integer_dtype

doc/source/whatsnew/v0.24.0.txt

@@ -323,7 +323,8 @@ ExtensionType Changes
 - Bug in :meth:`Series.get` for ``Series`` using ``ExtensionArray`` and integer index (:issue:`21257`)
 - :meth:`Series.combine()` works correctly with :class:`~pandas.api.extensions.ExtensionArray` inside of :class:`Series` (:issue:`20825`)
 - :meth:`Series.combine()` with scalar argument now works for any function type (:issue:`21248`)
--
+- Added :func:`pandas.api.types.is_extension_array` for testing whether an array is an ExtensionArray and :func:`pandas.api.types.is_extension_dtype` for testing whether a dtype is an ExtensionDtype (:issue:`22021`)
+
 
 .. _whatsnew_0240.api.incompatibilities:
 

pandas/core/algorithms.py

@@ -18,7 +18,8 @@
     is_unsigned_integer_dtype, is_signed_integer_dtype,
     is_integer_dtype, is_complex_dtype,
     is_object_dtype,
-    is_extension_array_dtype,
+    is_extension_array,
+    is_extension_dtype,
     is_categorical_dtype, is_sparse,
     is_period_dtype,
     is_numeric_dtype, is_float_dtype,
@@ -153,7 +154,7 @@ def _reconstruct_data(values, dtype, original):
     Index for extension types, otherwise ndarray casted to dtype
     """
     from pandas import Index
-    if is_extension_array_dtype(dtype):
+    if is_extension_dtype(dtype):
         values = dtype.construct_array_type()._from_sequence(values)
     elif is_datetime64tz_dtype(dtype) or is_period_dtype(dtype):
         values = Index(original)._shallow_copy(values, name=None)
@@ -357,7 +358,7 @@ def unique(values):
 
     values = _ensure_arraylike(values)
 
-    if is_extension_array_dtype(values):
+    if is_extension_array(values):
         # Dispatch to extension dtype's unique.
         return values.unique()
 
@@ -610,7 +611,7 @@ def factorize(values, sort=False, order=None, na_sentinel=-1, size_hint=None):
     values = _ensure_arraylike(values)
     original = values
 
-    if is_extension_array_dtype(values):
+    if is_extension_array(values):
         values = getattr(values, '_values', values)
         labels, uniques = values.factorize(na_sentinel=na_sentinel)
         dtype = original.dtype
@@ -705,7 +706,7 @@ def value_counts(values, sort=True, ascending=False, normalize=False,
 
     else:
 
-        if is_extension_array_dtype(values) or is_sparse(values):
+        if is_extension_array(values) or is_sparse(values):
 
             # handle Categorical and sparse,
             result = Series(values)._values.value_counts(dropna=dropna)
@@ -1591,7 +1592,7 @@ def take_nd(arr, indexer, axis=0, out=None, fill_value=np.nan, mask_info=None,
 
     # TODO(EA): Remove these if / elifs as datetimeTZ, interval, become EAs
     # dispatch to internal type takes
-    if is_extension_array_dtype(arr):
+    if is_extension_array(arr):
         return arr.take(indexer, fill_value=fill_value, allow_fill=allow_fill)
     elif is_datetimetz(arr):
         return arr.take(indexer, fill_value=fill_value, allow_fill=allow_fill)

pandas/core/arrays/categorical.py

@@ -20,7 +20,7 @@
     ensure_int64,
     ensure_object,
     ensure_platform_int,
-    is_extension_array_dtype,
+    is_extension_array,
     is_dtype_equal,
     is_datetimelike,
     is_datetime64_dtype,
@@ -1244,7 +1244,7 @@ def __array__(self, dtype=None):
         ret = take_1d(self.categories.values, self._codes)
         if dtype and not is_dtype_equal(dtype, self.categories.dtype):
             return np.asarray(ret, dtype)
-        if is_extension_array_dtype(ret):
+        if is_extension_array(ret):
             # When we're a Categorical[ExtensionArray], like Interval,
             # we need to ensure __array__ get's all the way to an
             # ndarray.

pandas/core/base.py

@@ -15,7 +15,7 @@
     is_list_like,
     is_scalar,
     is_extension_type,
-    is_extension_array_dtype)
+    is_extension_array)
 
 from pandas.util._validators import validate_bool_kwarg
 from pandas.errors import AbstractMethodError
@@ -749,7 +749,7 @@ def _ndarray_values(self):
 
         - categorical -> codes
         """
-        if is_extension_array_dtype(self):
+        if is_extension_array(self):
             return self.values._ndarray_values
         return self.values
 
@@ -857,7 +857,7 @@ def tolist(self):
         """
         if is_datetimelike(self._values):
             return [com._maybe_box_datetimelike(x) for x in self._values]
-        elif is_extension_array_dtype(self._values):
+        elif is_extension_array(self._values):
             return list(self._values)
         else:
             return self._values.tolist()

pandas/core/dtypes/api.py

@@ -4,6 +4,8 @@
 
 from .common import (pandas_dtype,
                      is_dtype_equal,
+                     is_extension_dtype,
+                     is_extension_array,
                      is_extension_type,
 
                      # categorical

pandas/core/dtypes/base.py

@@ -6,9 +6,9 @@
 
 
 class _DtypeOpsMixin(object):
-    # Not all of pandas' extension dtypes are compatibile with
+    # Not all of pandas' extension dtypes are compatible with
     # the new ExtensionArray interface. This means PandasExtensionDtype
-    # can't subclass ExtensionDtype yet, as is_extension_array_dtype would
+    # can't subclass ExtensionDtype yet, as is_extension_dtype would
     # incorrectly say that these types are extension types.
     #
     # In the interim, we put methods that are shared between the two base

pandas/core/dtypes/cast.py

@@ -12,7 +12,7 @@
                      is_complex, is_datetimetz, is_categorical_dtype,
                      is_datetimelike,
                      is_extension_type,
-                     is_extension_array_dtype,
+                     is_extension_dtype,
                      is_object_dtype,
                      is_datetime64tz_dtype, is_datetime64_dtype,
                      is_datetime64_ns_dtype,
@@ -294,7 +294,7 @@ def maybe_promote(dtype, fill_value=np.nan):
     elif is_datetimetz(dtype):
         if isna(fill_value):
             fill_value = iNaT
-    elif is_extension_array_dtype(dtype) and isna(fill_value):
+    elif is_extension_dtype(dtype) and isna(fill_value):
         fill_value = dtype.na_value
     elif is_float(fill_value):
         if issubclass(dtype.type, np.bool_):
@@ -332,7 +332,7 @@ def maybe_promote(dtype, fill_value=np.nan):
         dtype = np.object_
 
     # in case we have a string that looked like a number
-    if is_extension_array_dtype(dtype):
+    if is_extension_dtype(dtype):
         pass
     elif is_datetimetz(dtype):
         pass
@@ -650,7 +650,7 @@ def astype_nansafe(arr, dtype, copy=True):
         need to be very careful as the result shape could change! """
 
     # dispatch on extension dtype if needed
-    if is_extension_array_dtype(dtype):
+    if is_extension_dtype(dtype):
         return dtype.construct_array_type()._from_sequence(
             arr, dtype=dtype, copy=copy)
 

pandas/core/dtypes/common.py

@@ -1688,38 +1688,69 @@ def is_extension_type(arr):
     return False
 
 
-def is_extension_array_dtype(arr_or_dtype):
-    """Check if an object is a pandas extension array type.
+def is_extension_array(arr):
+    """Check if an array object is a pandas extension array type.
 
     Parameters
     ----------
-    arr_or_dtype : object
+    arr : object
 
     Returns
     -------
     bool
 
     Notes
     -----
-    This checks whether an object implements the pandas extension
+    This checks whether an array object implements the pandas extension
     array interface. In pandas, this includes:
 
     * Categorical
+    * Interval
 
-    Third-party libraries may implement arrays or types satisfying
+    Third-party libraries may implement arrays satisfying
     this interface as well.
-    """
-    from pandas.core.arrays import ExtensionArray
 
-    if isinstance(arr_or_dtype, (ABCIndexClass, ABCSeries)):
-        arr_or_dtype = arr_or_dtype._values
+    See Also
+    --------
+    is_extension_dtype : Similar method for dtypes.
+    """
+    from pandas.core.dtypes.base import ExtensionDtype
 
     try:
-        arr_or_dtype = pandas_dtype(arr_or_dtype)
-    except TypeError:
-        pass
+        dtype = getattr(arr, 'dtype')
+    except AttributeError:
+        return False
+
+    return isinstance(dtype, ExtensionDtype)
 
-    return isinstance(arr_or_dtype, (ExtensionDtype, ExtensionArray))
+
+def is_extension_dtype(dtype):
+    """Check if a dtype object is a pandas extension dtype.
+
+    Parameters
+    ----------
+    dtype : dtype
+
+    Returns
+    -------
+    bool
+
+    Notes
+    -----
+    This checks whether a dtype object implements the pandas extension
+    array interface. In pandas, this includes:
+
+    * CategoricalDtype
+    * IntervalDtype
+
+    Third-party libraries may implement dtypes satisfying
+    this interface as well.
+
+    See Also
+    --------
+    is_extension_array : Similar method for arrays.
+    """
+    return isinstance(dtype, ExtensionDtype)
 
 
 def is_complex_dtype(arr_or_dtype):

pandas/core/dtypes/concat.py

@@ -8,7 +8,7 @@
 from pandas.core.dtypes.common import (
     is_categorical_dtype,
     is_sparse,
-    is_extension_array_dtype,
+    is_extension_array,
     is_datetimetz,
     is_datetime64_dtype,
     is_timedelta64_dtype,
@@ -177,7 +177,7 @@ def is_nonempty(x):
     elif 'sparse' in typs:
         return _concat_sparse(to_concat, axis=axis, typs=typs)
 
-    extensions = [is_extension_array_dtype(x) for x in to_concat]
+    extensions = [is_extension_array(x) for x in to_concat]
     if any(extensions) and axis == 1:
         to_concat = [np.atleast_2d(x.astype('object')) for x in to_concat]
 

pandas/core/dtypes/missing.py

@@ -15,7 +15,7 @@
                      is_complex_dtype,
                      is_string_like_dtype, is_bool_dtype,
                      is_integer_dtype, is_dtype_equal,
-                     is_extension_array_dtype,
+                     is_extension_array,
                      needs_i8_conversion, ensure_object,
                      pandas_dtype,
                      is_scalar,
@@ -190,7 +190,7 @@ def _isna_ndarraylike(obj):
     values = getattr(obj, 'values', obj)
     dtype = values.dtype
 
-    if is_extension_array_dtype(obj):
+    if is_extension_array(obj):
         if isinstance(obj, (ABCIndexClass, ABCSeries)):
             values = obj._values
         else:
@@ -502,7 +502,7 @@ def na_value_for_dtype(dtype, compat=True):
     """
     dtype = pandas_dtype(dtype)
 
-    if is_extension_array_dtype(dtype):
+    if is_extension_array(dtype):
         return dtype.na_value
     if (is_datetime64_dtype(dtype) or is_datetime64tz_dtype(dtype) or
             is_timedelta64_dtype(dtype) or is_period_dtype(dtype)):
@@ -522,7 +522,7 @@ def remove_na_arraylike(arr):
     """
     Return array-like containing only true/non-NaN values, possibly empty.
     """
-    if is_extension_array_dtype(arr):
+    if is_extension_array(arr):
         return arr[notna(arr)]
     else:
         return arr[notna(lib.values_from_object(arr))]

pandas/core/frame.py

@@ -40,7 +40,7 @@
     is_categorical_dtype,
     is_object_dtype,
     is_extension_type,
-    is_extension_array_dtype,
+    is_extension_array,
     is_datetimetz,
     is_datetime64_any_dtype,
     is_bool_dtype,
@@ -517,7 +517,7 @@ def _get_axes(N, K, index=index, columns=columns):
             index, columns = _get_axes(len(values), 1)
             return _arrays_to_mgr([values], columns, index, columns,
                                   dtype=dtype)
-        elif (is_datetimetz(values) or is_extension_array_dtype(values)):
+        elif (is_datetimetz(values) or is_extension_array(values)):
             # GH19157
             if columns is None:
                 columns = [0]
@@ -3507,7 +3507,7 @@ def reindexer(value):
             value = maybe_cast_to_datetime(value, value.dtype)
 
         # return internal types directly
-        if is_extension_type(value) or is_extension_array_dtype(value):
+        if is_extension_type(value) or is_extension_array(value):
             return value
 
         # broadcast across multiple columns if necessary

pandas/core/indexes/base.py

@@ -44,7 +44,8 @@
     is_datetime64_any_dtype,
     is_datetime64tz_dtype,
     is_timedelta64_dtype,
-    is_extension_array_dtype,
+    is_extension_array,
+    is_extension_dtype,
     is_hashable,
     is_iterator, is_list_like,
     is_scalar)
@@ -275,7 +276,7 @@ def __new__(cls, data=None, dtype=None, copy=False, name=None,
                                  closed=closed)
 
         # extension dtype
-        elif is_extension_array_dtype(data) or is_extension_array_dtype(dtype):
+        elif is_extension_array(data) or is_extension_dtype(dtype):
             data = np.asarray(data)
             if not (dtype is None or is_object_dtype(dtype)):
 
@@ -1191,7 +1192,7 @@ def astype(self, dtype, copy=True):
             return CategoricalIndex(self.values, name=self.name, dtype=dtype,
                                     copy=copy)
 
-        elif is_extension_array_dtype(dtype):
+        elif is_extension_dtype(dtype):
             return Index(np.asarray(self), dtype=dtype, copy=copy)
 
         try:

pandas/core/internals/__init__.py

@@ -23,7 +23,7 @@
     is_datetimelike_v_numeric,
     is_float_dtype, is_numeric_dtype,
     is_numeric_v_string_like, is_extension_type,
-    is_extension_array_dtype,
+    is_extension_array,
     is_scalar,
     _get_dtype)
 from pandas.core.dtypes.cast import (
@@ -1071,7 +1071,7 @@ def set(self, item, value, check=False):
         # TODO(EA): Remove an is_extension_ when all extension types satisfy
         # the interface
         value_is_extension_type = (is_extension_type(value) or
-                                   is_extension_array_dtype(value))
+                                   is_extension_array(value))
 
         # categorical/spares/datetimetz
         if value_is_extension_type: