DEPR: groupby with as_index=False not including out-of-axis groupings

doc/source/whatsnew/v0.15.1.rst

@@ -70,9 +70,14 @@ API changes
 
   current behavior:
 
-  .. ipython:: python
+  .. code-block:: ipython
 
-    df.groupby(ts, as_index=False).max()
+     In [4]: df.groupby(ts, as_index=False).max()
+     Out[4]:
+        jim  joe
+     0   72   83
+     1   77   84
+     2   96   65
 
 - ``groupby`` will not erroneously exclude columns if the column name conflicts
   with the grouper name (:issue:`8112`):

doc/source/whatsnew/v2.1.0.rst

@@ -221,6 +221,7 @@ Deprecations
 - Deprecated the 'axis' keyword in :meth:`.GroupBy.idxmax`, :meth:`.GroupBy.idxmin`, :meth:`.GroupBy.fillna`, :meth:`.GroupBy.take`, :meth:`.GroupBy.skew`, :meth:`.GroupBy.rank`, :meth:`.GroupBy.cumprod`, :meth:`.GroupBy.cumsum`, :meth:`.GroupBy.cummax`, :meth:`.GroupBy.cummin`, :meth:`.GroupBy.pct_change`, :meth:`GroupBy.diff`, :meth:`.GroupBy.shift`, and :meth:`DataFrameGroupBy.corrwith`; for ``axis=1`` operate on the underlying :class:`DataFrame` instead (:issue:`50405`, :issue:`51046`)
 - Deprecated the "fastpath" keyword in :class:`Categorical` constructor, use :meth:`Categorical.from_codes` instead (:issue:`20110`)
 - Deprecated behavior of :meth:`Series.dt.to_pydatetime`, in a future version this will return a :class:`Series` containing python ``datetime`` objects instead of an ``ndarray`` of datetimes; this matches the behavior of other :meth:`Series.dt` properties (:issue:`20306`)
+- Deprecated :class:`.DataFrameGroupBy` with ``as_index=False`` not including groupings in the result when they are not columns of the DataFrame (:issue:`49519`)
 - Deprecated passing a dictionary to :meth:`.SeriesGroupBy.agg`; pass a list of aggregations instead (:issue:`50684`)
 - Deprecated logical operations (``|``, ``&``, ``^``) between pandas objects and dtype-less sequences (e.g. ``list``, ``tuple``), wrap a sequence in a :class:`Series` or numpy array before operating instead (:issue:`51521`)
 - Deprecated the methods :meth:`Series.bool` and :meth:`DataFrame.bool` (:issue:`51749`)

pandas/core/groupby/groupby.py

@@ -1244,8 +1244,21 @@ def _insert_inaxis_grouper(self, result: Series | DataFrame) -> DataFrame:
         ):
             # GH #28549
             # When using .apply(-), name will be in columns already
-            if in_axis and name not in columns:
-                result.insert(0, name, lev)
+            if name not in columns:
+                if in_axis:
+                    result.insert(0, name, lev)
+                else:
+                    msg = (
+                        "A grouping was used that is not in the columns of the "
+                        "DataFrame and so was excluded from the result. This grouping "
+                        "will be included in a future version of pandas. Add the "
+                        "grouping as a column of the DataFrame to silence this warning."
+                    )
+                    warnings.warn(
+                        message=msg,
+                        category=FutureWarning,
+                        stacklevel=find_stack_level(),
+                    )
 
         return result
 

pandas/tests/groupby/test_categorical.py

@@ -771,7 +771,9 @@ def test_as_index():
 
     # function grouper
     f = lambda r: df.loc[r, "A"]
-    result = df.groupby(["cat", f], as_index=False, observed=True).sum()
+    msg = "A grouping .* was excluded from the result"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        result = df.groupby(["cat", f], as_index=False, observed=True).sum()
     expected = DataFrame(
         {
             "cat": Categorical([1, 2], categories=df.cat.cat.categories),
@@ -784,7 +786,9 @@ def test_as_index():
 
     # another not in-axis grouper (conflicting names in index)
     s = Series(["a", "b", "b"], name="cat")
-    result = df.groupby(["cat", s], as_index=False, observed=True).sum()
+    msg = "A grouping .* was excluded from the result"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        result = df.groupby(["cat", s], as_index=False, observed=True).sum()
     tm.assert_frame_equal(result, expected)
 
     # is original index dropped?

pandas/tests/groupby/test_groupby.py

@@ -244,18 +244,26 @@ def f(x, q=None, axis=0):
     # DataFrame
     for as_index in [True, False]:
         df_grouped = tsframe.groupby(lambda x: x.month, as_index=as_index)
-        agg_result = df_grouped.agg(np.percentile, 80, axis=0)
-        apply_result = df_grouped.apply(DataFrame.quantile, 0.8)
-        expected = df_grouped.quantile(0.8)
+        warn = None if as_index else FutureWarning
+        msg = "A grouping .* was excluded from the result"
+        with tm.assert_produces_warning(warn, match=msg):
+            agg_result = df_grouped.agg(np.percentile, 80, axis=0)
+        with tm.assert_produces_warning(warn, match=msg):
+            apply_result = df_grouped.apply(DataFrame.quantile, 0.8)
+        with tm.assert_produces_warning(warn, match=msg):
+            expected = df_grouped.quantile(0.8)
         tm.assert_frame_equal(apply_result, expected, check_names=False)
         tm.assert_frame_equal(agg_result, expected)
 
         apply_result = df_grouped.apply(DataFrame.quantile, [0.4, 0.8])
-        expected_seq = df_grouped.quantile([0.4, 0.8])
+        with tm.assert_produces_warning(warn, match=msg):
+            expected_seq = df_grouped.quantile([0.4, 0.8])
         tm.assert_frame_equal(apply_result, expected_seq, check_names=False)
 
-        agg_result = df_grouped.agg(f, q=80)
-        apply_result = df_grouped.apply(DataFrame.quantile, q=0.8)
+        with tm.assert_produces_warning(warn, match=msg):
+            agg_result = df_grouped.agg(f, q=80)
+        with tm.assert_produces_warning(warn, match=msg):
+            apply_result = df_grouped.apply(DataFrame.quantile, q=0.8)
         tm.assert_frame_equal(agg_result, expected)
         tm.assert_frame_equal(apply_result, expected, check_names=False)
 
@@ -266,7 +274,10 @@ def test_pass_args_kwargs_duplicate_columns(tsframe, as_index):
     tsframe.columns = ["A", "B", "A", "C"]
     gb = tsframe.groupby(lambda x: x.month, as_index=as_index)
 
-    res = gb.agg(np.percentile, 80, axis=0)
+    warn = None if as_index else FutureWarning
+    msg = "A grouping .* was excluded from the result"
+    with tm.assert_produces_warning(warn, match=msg):
+        res = gb.agg(np.percentile, 80, axis=0)
 
     ex_data = {
         1: tsframe[tsframe.index.month == 1].quantile(0.8),

pandas/tests/groupby/test_groupby_dropna.py

@@ -576,7 +576,13 @@ def test_categorical_reducers(
     gb_keepna = df.groupby(
         keys, dropna=False, observed=observed, sort=sort, as_index=as_index
     )
-    result = getattr(gb_keepna, reduction_func)(*args)
+    if as_index or index_kind == "range" or reduction_func == "size":
+        warn = None
+    else:
+        warn = FutureWarning
+    msg = "A grouping .* was excluded from the result"
+    with tm.assert_produces_warning(warn, match=msg):
+        result = getattr(gb_keepna, reduction_func)(*args)
 
     # size will return a Series, others are DataFrame
     tm.assert_equal(result, expected)

pandas/tests/groupby/test_grouping.py

@@ -1085,7 +1085,9 @@ def test_grouping_by_key_is_in_axis():
 
     # Currently only in-axis groupings are including in the result when as_index=False;
     # This is likely to change in the future.
-    result = gb.sum()
+    msg = "A grouping .* was excluded from the result"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        result = gb.sum()
     expected = DataFrame({"b": [1, 2], "c": [7, 5]})
     tm.assert_frame_equal(result, expected)