API/DEPR: Change default skipna behaviour + deprecate numeric_only in…

… Categorical.min and max (#27929)

doc/source/whatsnew/v1.0.0.rst

@@ -304,6 +304,26 @@ The following methods now also correctly output values for unobserved categories
    df.groupby(["cat_1", "cat_2"], observed=False)["value"].count()
 
 
+By default :meth:`Categorical.min` now returns the minimum instead of np.nan
+^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
+
+When :class:`Categorical` contains ``np.nan``,
+:meth:`Categorical.min` no longer return ``np.nan`` by default (skipna=True) (:issue:`25303`)
+
+*pandas 0.25.x*
+
+.. code-block:: ipython
+
+   In [1]: pd.Categorical([1, 2, np.nan], ordered=True).min()
+   Out[1]: nan
+
+
+*pandas 1.0.0*
+
+.. ipython:: python
+
+   pd.Categorical([1, 2, np.nan], ordered=True).min()
+
 .. _whatsnew_1000.api_breaking.deps:
 
 Increased minimum versions for dependencies
@@ -410,6 +430,8 @@ Deprecations
 - :func:`is_extension_type` is deprecated, :func:`is_extension_array_dtype` should be used instead (:issue:`29457`)
 - :func:`eval` keyword argument "truediv" is deprecated and will be removed in a future version (:issue:`29812`)
 - :meth:`Categorical.take_nd` is deprecated, use :meth:`Categorical.take` instead (:issue:`27745`)
+- The parameter ``numeric_only`` of :meth:`Categorical.min` and :meth:`Categorical.max` is deprecated and replaced with ``skipna`` (:issue:`25303`)
+-
 
 .. _whatsnew_1000.prior_deprecations:
 

pandas/core/arrays/categorical.py

@@ -2123,7 +2123,8 @@ def _reduce(self, name, axis=0, **kwargs):
             raise TypeError(f"Categorical cannot perform the operation {name}")
         return func(**kwargs)
 
-    def min(self, numeric_only=None, **kwargs):
+    @deprecate_kwarg(old_arg_name="numeric_only", new_arg_name="skipna")
+    def min(self, skipna=True):
         """
         The minimum value of the object.
 
@@ -2139,17 +2140,18 @@ def min(self, numeric_only=None, **kwargs):
         min : the minimum of this `Categorical`
         """
         self.check_for_ordered("min")
-        if numeric_only:
-            good = self._codes != -1
-            pointer = self._codes[good].min(**kwargs)
-        else:
-            pointer = self._codes.min(**kwargs)
-        if pointer == -1:
-            return np.nan
+        good = self._codes != -1
+        if not good.all():
+            if skipna:
+                pointer = self._codes[good].min()
+            else:
+                return np.nan
         else:
-            return self.categories[pointer]
+            pointer = self._codes.min()
+        return self.categories[pointer]
 
-    def max(self, numeric_only=None, **kwargs):
+    @deprecate_kwarg(old_arg_name="numeric_only", new_arg_name="skipna")
+    def max(self, skipna=True):
         """
         The maximum value of the object.
 
@@ -2165,15 +2167,15 @@ def max(self, numeric_only=None, **kwargs):
         max : the maximum of this `Categorical`
         """
         self.check_for_ordered("max")
-        if numeric_only:
-            good = self._codes != -1
-            pointer = self._codes[good].max(**kwargs)
-        else:
-            pointer = self._codes.max(**kwargs)
-        if pointer == -1:
-            return np.nan
+        good = self._codes != -1
+        if not good.all():
+            if skipna:
+                pointer = self._codes[good].max()
+            else:
+                return np.nan
         else:
-            return self.categories[pointer]
+            pointer = self._codes.max()
+        return self.categories[pointer]
 
     def mode(self, dropna=True):
         """

pandas/core/series.py

@@ -3820,9 +3820,7 @@ def _reduce(
             self._get_axis_number(axis)
 
         if isinstance(delegate, Categorical):
-            # TODO deprecate numeric_only argument for Categorical and use
-            # skipna as well, see GH25303
-            return delegate._reduce(name, numeric_only=numeric_only, **kwds)
+            return delegate._reduce(name, skipna=skipna, **kwds)
         elif isinstance(delegate, ExtensionArray):
             # dispatch to ExtensionArray interface
             return delegate._reduce(name, skipna=skipna, **kwds)

pandas/tests/arrays/categorical/test_analytics.py

@@ -35,31 +35,43 @@ def test_min_max(self):
         assert _min == "d"
         assert _max == "a"
 
+    @pytest.mark.parametrize("skipna", [True, False])
+    def test_min_max_with_nan(self, skipna):
+        # GH 25303
         cat = Categorical(
             [np.nan, "b", "c", np.nan], categories=["d", "c", "b", "a"], ordered=True
         )
-        _min = cat.min()
-        _max = cat.max()
-        assert np.isnan(_min)
-        assert _max == "b"
+        _min = cat.min(skipna=skipna)
+        _max = cat.max(skipna=skipna)
 
-        _min = cat.min(numeric_only=True)
-        assert _min == "c"
-        _max = cat.max(numeric_only=True)
-        assert _max == "b"
+        if skipna is False:
+            assert np.isnan(_min)
+            assert np.isnan(_max)
+        else:
+            assert _min == "c"
+            assert _max == "b"
 
         cat = Categorical(
             [np.nan, 1, 2, np.nan], categories=[5, 4, 3, 2, 1], ordered=True
         )
-        _min = cat.min()
-        _max = cat.max()
-        assert np.isnan(_min)
-        assert _max == 1
-
-        _min = cat.min(numeric_only=True)
-        assert _min == 2
-        _max = cat.max(numeric_only=True)
-        assert _max == 1
+        _min = cat.min(skipna=skipna)
+        _max = cat.max(skipna=skipna)
+
+        if skipna is False:
+            assert np.isnan(_min)
+            assert np.isnan(_max)
+        else:
+            assert _min == 2
+            assert _max == 1
+
+    @pytest.mark.parametrize("method", ["min", "max"])
+    def test_deprecate_numeric_only_min_max(self, method):
+        # GH 25303
+        cat = Categorical(
+            [np.nan, 1, 2, np.nan], categories=[5, 4, 3, 2, 1], ordered=True
+        )
+        with tm.assert_produces_warning(expected_warning=FutureWarning):
+            getattr(cat, method)(numeric_only=True)
 
     @pytest.mark.parametrize(
         "values,categories,exp_mode",

pandas/tests/reductions/test_reductions.py

@@ -1043,7 +1043,7 @@ def test_min_max(self):
         )
         _min = cat.min()
         _max = cat.max()
-        assert np.isnan(_min)
+        assert _min == "c"
         assert _max == "b"
 
         cat = Series(
@@ -1053,30 +1053,24 @@ def test_min_max(self):
         )
         _min = cat.min()
         _max = cat.max()
-        assert np.isnan(_min)
+        assert _min == 2
         assert _max == 1
 
-    def test_min_max_numeric_only(self):
-        # TODO deprecate numeric_only argument for Categorical and use
-        # skipna as well, see GH25303
+    @pytest.mark.parametrize("skipna", [True, False])
+    def test_min_max_skipna(self, skipna):
+        # GH 25303
         cat = Series(
             Categorical(["a", "b", np.nan, "a"], categories=["b", "a"], ordered=True)
         )
+        _min = cat.min(skipna=skipna)
+        _max = cat.max(skipna=skipna)
 
-        _min = cat.min()
-        _max = cat.max()
-        assert np.isnan(_min)
-        assert _max == "a"
-
-        _min = cat.min(numeric_only=True)
-        _max = cat.max(numeric_only=True)
-        assert _min == "b"
-        assert _max == "a"
-
-        _min = cat.min(numeric_only=False)
-        _max = cat.max(numeric_only=False)
-        assert np.isnan(_min)
-        assert _max == "a"
+        if skipna is True:
+            assert _min == "b"
+            assert _max == "a"
+        else:
+            assert np.isnan(_min)
+            assert np.isnan(_max)
 
 
 class TestSeriesMode: