DEPR: numeric_only=None in DataFrame.rank (#45059)

Author: jbrockmendel

doc/source/whatsnew/v0.18.0.rst

@@ -669,9 +669,9 @@ New signature
 
 .. ipython:: python
 
-   pd.Series([0,1]).rank(axis=0, method='average', numeric_only=None,
+   pd.Series([0,1]).rank(axis=0, method='average', numeric_only=False,
                          na_option='keep', ascending=True, pct=False)
-   pd.DataFrame([0,1]).rank(axis=0, method='average', numeric_only=None,
+   pd.DataFrame([0,1]).rank(axis=0, method='average', numeric_only=False,
                             na_option='keep', ascending=True, pct=False)
 
 

doc/source/whatsnew/v1.4.0.rst

@@ -547,6 +547,7 @@ Other Deprecations
 - Deprecated :meth:`Categorical.replace`, use :meth:`Series.replace` instead (:issue:`44929`)
 - Deprecated :meth:`Index.__getitem__` with a bool key; use ``index.values[key]`` to get the old behavior (:issue:`44051`)
 - Deprecated downcasting column-by-column in :meth:`DataFrame.where` with integer-dtypes (:issue:`44597`)
+- Deprecated ``numeric_only=None`` in :meth:`DataFrame.rank`; in a future version ``numeric_only`` must be either ``True`` or ``False`` (the default) (:issue:`45036`)
 - Deprecated :meth:`NaT.freq` (:issue:`45071`)
 -
 

pandas/core/generic.py

@@ -8406,7 +8406,7 @@ def rank(
         self: NDFrameT,
         axis=0,
         method: str = "average",
-        numeric_only: bool_t | None = None,
+        numeric_only: bool_t | None | lib.NoDefault = lib.no_default,
         na_option: str = "keep",
         ascending: bool_t = True,
         pct: bool_t = False,
@@ -8492,6 +8492,20 @@ def rank(
         3   spider          8.0           4.0       4.0        4.0     1.000
         4    snake          NaN           NaN       NaN        5.0       NaN
         """
+        warned = False
+        if numeric_only is None:
+            # GH#45036
+            warnings.warn(
+                f"'numeric_only=None' in {type(self).__name__}.rank is deprecated "
+                "and will raise in a future version. Pass either 'True' or "
+                "'False'. 'False' will be the default.",
+                FutureWarning,
+                stacklevel=find_stack_level(),
+            )
+            warned = True
+        elif numeric_only is lib.no_default:
+            numeric_only = None
+
         axis = self._get_axis_number(axis)
 
         if na_option not in {"keep", "top", "bottom"}:
@@ -8521,6 +8535,16 @@ def ranker(data):
                 return ranker(self)
             except TypeError:
                 numeric_only = True
+                if not warned:
+                    # Only warn here if we didn't already issue a warning above
+                    # GH#45036
+                    warnings.warn(
+                        f"Dropping of nuisance columns in {type(self).__name__}.rank "
+                        "is deprecated; in a future version this will raise TypeError. "
+                        "Select only valid columns before calling rank.",
+                        FutureWarning,
+                        stacklevel=find_stack_level(),
+                    )
 
         if numeric_only:
             data = self._get_numeric_data()

pandas/tests/frame/methods/test_rank.py

@@ -136,7 +136,10 @@ def test_rank_mixed_frame(self, float_string_frame):
         float_string_frame["datetime"] = datetime.now()
         float_string_frame["timedelta"] = timedelta(days=1, seconds=1)
 
-        result = float_string_frame.rank(1)
+        with tm.assert_produces_warning(FutureWarning, match="numeric_only=None"):
+            float_string_frame.rank(numeric_only=None)
+        with tm.assert_produces_warning(FutureWarning, match="Dropping of nuisance"):
+            result = float_string_frame.rank(1)
         expected = float_string_frame.rank(1, numeric_only=True)
         tm.assert_frame_equal(result, expected)
 
@@ -489,5 +492,7 @@ def test_rank_object_first(self, frame_or_series, na_option, ascending, expected
     )
     def test_rank_mixed_axis_zero(self, data, expected):
         df = DataFrame(data)
-        result = df.rank()
+        msg = "Dropping of nuisance columns"
+        with tm.assert_produces_warning(FutureWarning, match=msg):
+            result = df.rank()
         tm.assert_frame_equal(result, expected)