EA: implement+test EA.view (#27633)

Author: jbrockmendel

doc/source/reference/extensions.rst

@@ -44,6 +44,7 @@ objects.
       api.extensions.ExtensionArray.argsort
       api.extensions.ExtensionArray.astype
       api.extensions.ExtensionArray.copy
+      api.extensions.ExtensionArray.view
       api.extensions.ExtensionArray.dropna
       api.extensions.ExtensionArray.factorize
       api.extensions.ExtensionArray.fillna

pandas/core/arrays/base.py

@@ -64,6 +64,7 @@ class ExtensionArray:
     shift
     take
     unique
+    view
     _concat_same_type
     _formatter
     _from_factorized
@@ -146,7 +147,7 @@ class ExtensionArray:
     If implementing NumPy's ``__array_ufunc__`` interface, pandas expects
     that
 
-    1. You defer by raising ``NotImplemented`` when any Series are present
+    1. You defer by returning ``NotImplemented`` when any Series are present
        in `inputs`. Pandas will extract the arrays and call the ufunc again.
     2. You define a ``_HANDLED_TYPES`` tuple as an attribute on the class.
        Pandas inspect this to determine whether the ufunc is valid for the
@@ -861,6 +862,27 @@ def copy(self) -> ABCExtensionArray:
         """
         raise AbstractMethodError(self)
 
+    def view(self, dtype=None) -> Union[ABCExtensionArray, np.ndarray]:
+        """
+        Return a view on the array.
+
+        Parameters
+        ----------
+        dtype : str, np.dtype, or ExtensionDtype, optional
+            Default None
+
+        Returns
+        -------
+        ExtensionArray
+        """
+        # NB:
+        # - This must return a *new* object referencing the same data, not self.
+        # - The only case that *must* be implemented is with dtype=None,
+        #   giving a view with the same dtype as self.
+        if dtype is not None:
+            raise NotImplementedError(dtype)
+        return self[:]
+
     # ------------------------------------------------------------------------
     # Printing
     # ------------------------------------------------------------------------

pandas/core/arrays/categorical.py

@@ -516,19 +516,12 @@ def astype(self, dtype: Dtype, copy: bool = True) -> ArrayLike:
             return self._set_dtype(dtype)
         return np.array(self, dtype=dtype, copy=copy)
 
-    @cache_readonly
-    def ndim(self) -> int:
-        """
-        Number of dimensions of the Categorical
-        """
-        return self._codes.ndim
-
     @cache_readonly
     def size(self) -> int:
         """
         return the len of myself
         """
-        return len(self)
+        return self._codes.size
 
     @cache_readonly
     def itemsize(self) -> int:
@@ -1763,18 +1756,10 @@ def ravel(self, order="C"):
         )
         return np.array(self)
 
-    def view(self):
-        """
-        Return a view of myself.
-
-        For internal compatibility with numpy arrays.
-
-        Returns
-        -------
-        view : Categorical
-           Returns `self`!
-        """
-        return self
+    def view(self, dtype=None):
+        if dtype is not None:
+            raise NotImplementedError(dtype)
+        return self._constructor(values=self._codes, dtype=self.dtype, fastpath=True)
 
     def to_dense(self):
         """

pandas/core/arrays/datetimelike.py

@@ -544,18 +544,8 @@ def astype(self, dtype, copy=True):
             return np.asarray(self, dtype=dtype)
 
     def view(self, dtype=None):
-        """
-        New view on this array with the same data.
-
-        Parameters
-        ----------
-        dtype : numpy dtype, optional
-
-        Returns
-        -------
-        ndarray
-            With the specified `dtype`.
-        """
+        if dtype is None or dtype is self.dtype:
+            return type(self)(self._data, dtype=self.dtype)
         return self._data.view(dtype=dtype)
 
     # ------------------------------------------------------------------

pandas/core/arrays/interval.py

@@ -695,18 +695,14 @@ def isna(self):
         return isna(self.left)
 
     @property
-    def nbytes(self):
+    def nbytes(self) -> int:
         return self.left.nbytes + self.right.nbytes
 
     @property
-    def size(self):
+    def size(self) -> int:
         # Avoid materializing self.values
         return self.left.size
 
-    @property
-    def shape(self):
-        return self.left.shape
-
     def take(self, indices, allow_fill=False, fill_value=None, axis=None, **kwargs):
         """
         Take elements from the IntervalArray.

pandas/core/arrays/numpy_.py

@@ -245,11 +245,11 @@ def __setitem__(self, key, value):
         else:
             self._ndarray[key] = value
 
-    def __len__(self):
+    def __len__(self) -> int:
         return len(self._ndarray)
 
     @property
-    def nbytes(self):
+    def nbytes(self) -> int:
         return self._ndarray.nbytes
 
     def isna(self):

pandas/core/arrays/sparse.py

@@ -839,7 +839,7 @@ def fill_value(self, value):
         self._dtype = SparseDtype(self.dtype.subtype, value)
 
     @property
-    def kind(self):
+    def kind(self) -> str:
         """
         The kind of sparse index for this array. One of {'integer', 'block'}.
         """
@@ -854,7 +854,7 @@ def _valid_sp_values(self):
         mask = notna(sp_vals)
         return sp_vals[mask]
 
-    def __len__(self):
+    def __len__(self) -> int:
         return self.sp_index.length
 
     @property
@@ -868,7 +868,7 @@ def _fill_value_matches(self, fill_value):
             return self.fill_value == fill_value
 
     @property
-    def nbytes(self):
+    def nbytes(self) -> int:
         return self.sp_values.nbytes + self.sp_index.nbytes
 
     @property
@@ -886,7 +886,7 @@ def density(self):
         return r
 
     @property
-    def npoints(self):
+    def npoints(self) -> int:
         """
         The number of non- ``fill_value`` points.
 

pandas/tests/extension/arrow/test_bool.py

@@ -41,6 +41,10 @@ def test_copy(self, data):
         # __setitem__ does not work, so we only have a smoke-test
         data.copy()
 
+    def test_view(self, data):
+        # __setitem__ does not work, so we only have a smoke-test
+        data.view()
+
 
 class TestConstructors(BaseArrowTests, base.BaseConstructorsTests):
     def test_from_dtype(self, data):

pandas/tests/extension/base/interface.py

@@ -75,3 +75,18 @@ def test_copy(self, data):
 
         data[1] = data[0]
         assert result[1] != result[0]
+
+    def test_view(self, data):
+        # view with no dtype should return a shallow copy, *not* the same
+        #  object
+        assert data[1] != data[0]
+
+        result = data.view()
+        assert result is not data
+        assert type(result) == type(data)
+
+        result[1] = result[0]
+        assert data[1] == data[0]
+
+        # check specifically that the `dtype` kwarg is accepted
+        data.view(dtype=None)

pandas/tests/extension/decimal/array.py

@@ -137,11 +137,11 @@ def __setitem__(self, key, value):
             value = decimal.Decimal(value)
         self._data[key] = value
 
-    def __len__(self):
+    def __len__(self) -> int:
         return len(self._data)
 
     @property
-    def nbytes(self):
+    def nbytes(self) -> int:
         n = len(self)
         if n:
             return n * sys.getsizeof(self[0])

pandas/tests/extension/json/array.py

@@ -80,6 +80,9 @@ def __getitem__(self, item):
         elif isinstance(item, abc.Iterable):
             # fancy indexing
             return type(self)([self.data[i] for i in item])
+        elif isinstance(item, slice) and item == slice(None):
+            # Make sure we get a view
+            return type(self)(self.data)
         else:
             # slice
             return type(self)(self.data[item])
@@ -103,11 +106,11 @@ def __setitem__(self, key, value):
                     assert isinstance(v, self.dtype.type)
                     self.data[k] = v
 
-    def __len__(self):
+    def __len__(self) -> int:
         return len(self.data)
 
     @property
-    def nbytes(self):
+    def nbytes(self) -> int:
         return sys.getsizeof(self.data)
 
     def isna(self):

pandas/tests/extension/test_interval.py

@@ -95,7 +95,10 @@ class TestGrouping(BaseInterval, base.BaseGroupbyTests):
 
 
 class TestInterface(BaseInterval, base.BaseInterfaceTests):
-    pass
+    def test_view(self, data):
+        # __setitem__ incorrectly makes a copy (GH#27147), so we only
+        #  have a smoke-test
+        data.view()
 
 
 class TestReduce(base.BaseNoReduceTests):

pandas/tests/extension/test_sparse.py

@@ -103,6 +103,10 @@ def test_copy(self, data):
         # __setitem__ does not work, so we only have a smoke-test
         data.copy()
 
+    def test_view(self, data):
+        # __setitem__ does not work, so we only have a smoke-test
+        data.view()
+
 
 class TestConstructors(BaseSparseTests, base.BaseConstructorsTests):
     pass