venv

venv/lib/python3.12/site-packages/pandas/tests/groupby/__init__.py

@@ -0,0 +1,25 @@
+def get_groupby_method_args(name, obj):
+    """
+    Get required arguments for a groupby method.
+
+    When parametrizing a test over groupby methods (e.g. "sum", "mean", "fillna"),
+    it is often the case that arguments are required for certain methods.
+
+    Parameters
+    ----------
+    name: str
+        Name of the method.
+    obj: Series or DataFrame
+        pandas object that is being grouped.
+
+    Returns
+    -------
+    A tuple of required arguments for the method.
+    """
+    if name in ("nth", "fillna", "take"):
+        return (0,)
+    if name == "quantile":
+        return (0.5,)
+    if name == "corrwith":
+        return (obj,)
+    return ()

venv/lib/python3.12/site-packages/pandas/tests/groupby/aggregate/test_cython.py

@@ -0,0 +1,435 @@
+"""
+test cython .agg behavior
+"""
+
+import numpy as np
+import pytest
+
+from pandas.core.dtypes.common import (
+    is_float_dtype,
+    is_integer_dtype,
+)
+
+import pandas as pd
+from pandas import (
+    DataFrame,
+    Index,
+    NaT,
+    Series,
+    Timedelta,
+    Timestamp,
+    bdate_range,
+)
+import pandas._testing as tm
+import pandas.core.common as com
+
+
+@pytest.mark.parametrize(
+    "op_name",
+    [
+        "count",
+        "sum",
+        "std",
+        "var",
+        "sem",
+        "mean",
+        pytest.param(
+            "median",
+            # ignore mean of empty slice
+            # and all-NaN
+            marks=[pytest.mark.filterwarnings("ignore::RuntimeWarning")],
+        ),
+        "prod",
+        "min",
+        "max",
+    ],
+)
+def test_cythonized_aggers(op_name):
+    data = {
+        "A": [0, 0, 0, 0, 1, 1, 1, 1, 1, 1.0, np.nan, np.nan],
+        "B": ["A", "B"] * 6,
+        "C": np.random.default_rng(2).standard_normal(12),
+    }
+    df = DataFrame(data)
+    df.loc[2:10:2, "C"] = np.nan
+
+    op = lambda x: getattr(x, op_name)()
+
+    # single column
+    grouped = df.drop(["B"], axis=1).groupby("A")
+    exp = {cat: op(group["C"]) for cat, group in grouped}
+    exp = DataFrame({"C": exp})
+    exp.index.name = "A"
+    result = op(grouped)
+    tm.assert_frame_equal(result, exp)
+
+    # multiple columns
+    grouped = df.groupby(["A", "B"])
+    expd = {}
+    for (cat1, cat2), group in grouped:
+        expd.setdefault(cat1, {})[cat2] = op(group["C"])
+    exp = DataFrame(expd).T.stack(future_stack=True)
+    exp.index.names = ["A", "B"]
+    exp.name = "C"
+
+    result = op(grouped)["C"]
+    if op_name in ["sum", "prod"]:
+        tm.assert_series_equal(result, exp)
+
+
+def test_cython_agg_boolean():
+    frame = DataFrame(
+        {
+            "a": np.random.default_rng(2).integers(0, 5, 50),
+            "b": np.random.default_rng(2).integers(0, 2, 50).astype("bool"),
+        }
+    )
+    result = frame.groupby("a")["b"].mean()
+    msg = "using SeriesGroupBy.mean"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        # GH#53425
+        expected = frame.groupby("a")["b"].agg(np.mean)
+
+    tm.assert_series_equal(result, expected)
+
+
+def test_cython_agg_nothing_to_agg():
+    frame = DataFrame(
+        {"a": np.random.default_rng(2).integers(0, 5, 50), "b": ["foo", "bar"] * 25}
+    )
+
+    msg = "Cannot use numeric_only=True with SeriesGroupBy.mean and non-numeric dtypes"
+    with pytest.raises(TypeError, match=msg):
+        frame.groupby("a")["b"].mean(numeric_only=True)
+
+    frame = DataFrame(
+        {"a": np.random.default_rng(2).integers(0, 5, 50), "b": ["foo", "bar"] * 25}
+    )
+
+    result = frame[["b"]].groupby(frame["a"]).mean(numeric_only=True)
+    expected = DataFrame(
+        [], index=frame["a"].sort_values().drop_duplicates(), columns=[]
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+def test_cython_agg_nothing_to_agg_with_dates():
+    frame = DataFrame(
+        {
+            "a": np.random.default_rng(2).integers(0, 5, 50),
+            "b": ["foo", "bar"] * 25,
+            "dates": pd.date_range("now", periods=50, freq="min"),
+        }
+    )
+    msg = "Cannot use numeric_only=True with SeriesGroupBy.mean and non-numeric dtypes"
+    with pytest.raises(TypeError, match=msg):
+        frame.groupby("b").dates.mean(numeric_only=True)
+
+
+def test_cython_agg_frame_columns():
+    # #2113
+    df = DataFrame({"x": [1, 2, 3], "y": [3, 4, 5]})
+
+    msg = "DataFrame.groupby with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        df.groupby(level=0, axis="columns").mean()
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        df.groupby(level=0, axis="columns").mean()
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        df.groupby(level=0, axis="columns").mean()
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        df.groupby(level=0, axis="columns").mean()
+
+
+def test_cython_agg_return_dict():
+    # GH 16741
+    df = DataFrame(
+        {
+            "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"],
+            "B": ["one", "one", "two", "three", "two", "two", "one", "three"],
+            "C": np.random.default_rng(2).standard_normal(8),
+            "D": np.random.default_rng(2).standard_normal(8),
+        }
+    )
+
+    ts = df.groupby("A")["B"].agg(lambda x: x.value_counts().to_dict())
+    expected = Series(
+        [{"two": 1, "one": 1, "three": 1}, {"two": 2, "one": 2, "three": 1}],
+        index=Index(["bar", "foo"], name="A"),
+        name="B",
+    )
+    tm.assert_series_equal(ts, expected)
+
+
+def test_cython_fail_agg():
+    dr = bdate_range("1/1/2000", periods=50)
+    ts = Series(["A", "B", "C", "D", "E"] * 10, index=dr)
+
+    grouped = ts.groupby(lambda x: x.month)
+    summed = grouped.sum()
+    msg = "using SeriesGroupBy.sum"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        # GH#53425
+        expected = grouped.agg(np.sum)
+    tm.assert_series_equal(summed, expected)
+
+
+@pytest.mark.parametrize(
+    "op, targop",
+    [
+        ("mean", np.mean),
+        ("median", np.median),
+        ("var", np.var),
+        ("sum", np.sum),
+        ("prod", np.prod),
+        ("min", np.min),
+        ("max", np.max),
+        ("first", lambda x: x.iloc[0]),
+        ("last", lambda x: x.iloc[-1]),
+    ],
+)
+def test__cython_agg_general(op, targop):
+    df = DataFrame(np.random.default_rng(2).standard_normal(1000))
+    labels = np.random.default_rng(2).integers(0, 50, size=1000).astype(float)
+
+    result = df.groupby(labels)._cython_agg_general(op, alt=None, numeric_only=True)
+    warn = FutureWarning if targop in com._cython_table else None
+    msg = f"using DataFrameGroupBy.{op}"
+    with tm.assert_produces_warning(warn, match=msg):
+        # GH#53425
+        expected = df.groupby(labels).agg(targop)
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "op, targop",
+    [
+        ("mean", np.mean),
+        ("median", lambda x: np.median(x) if len(x) > 0 else np.nan),
+        ("var", lambda x: np.var(x, ddof=1)),
+        ("min", np.min),
+        ("max", np.max),
+    ],
+)
+def test_cython_agg_empty_buckets(op, targop, observed):
+    df = DataFrame([11, 12, 13])
+    grps = range(0, 55, 5)
+
+    # calling _cython_agg_general directly, instead of via the user API
+    # which sets different values for min_count, so do that here.
+    g = df.groupby(pd.cut(df[0], grps), observed=observed)
+    result = g._cython_agg_general(op, alt=None, numeric_only=True)
+
+    g = df.groupby(pd.cut(df[0], grps), observed=observed)
+    expected = g.agg(lambda x: targop(x))
+    tm.assert_frame_equal(result, expected)
+
+
+def test_cython_agg_empty_buckets_nanops(observed):
+    # GH-18869 can't call nanops on empty groups, so hardcode expected
+    # for these
+    df = DataFrame([11, 12, 13], columns=["a"])
+    grps = np.arange(0, 25, 5, dtype=int)
+    # add / sum
+    result = df.groupby(pd.cut(df["a"], grps), observed=observed)._cython_agg_general(
+        "sum", alt=None, numeric_only=True
+    )
+    intervals = pd.interval_range(0, 20, freq=5)
+    expected = DataFrame(
+        {"a": [0, 0, 36, 0]},
+        index=pd.CategoricalIndex(intervals, name="a", ordered=True),
+    )
+    if observed:
+        expected = expected[expected.a != 0]
+
+    tm.assert_frame_equal(result, expected)
+
+    # prod
+    result = df.groupby(pd.cut(df["a"], grps), observed=observed)._cython_agg_general(
+        "prod", alt=None, numeric_only=True
+    )
+    expected = DataFrame(
+        {"a": [1, 1, 1716, 1]},
+        index=pd.CategoricalIndex(intervals, name="a", ordered=True),
+    )
+    if observed:
+        expected = expected[expected.a != 1]
+
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("op", ["first", "last", "max", "min"])
+@pytest.mark.parametrize(
+    "data", [Timestamp("2016-10-14 21:00:44.557"), Timedelta("17088 days 21:00:44.557")]
+)
+def test_cython_with_timestamp_and_nat(op, data):
+    # https://github.com/pandas-dev/pandas/issues/19526
+    df = DataFrame({"a": [0, 1], "b": [data, NaT]})
+    index = Index([0, 1], name="a")
+
+    # We will group by a and test the cython aggregations
+    expected = DataFrame({"b": [data, NaT]}, index=index)
+
+    result = df.groupby("a").aggregate(op)
+    tm.assert_frame_equal(expected, result)
+
+
+@pytest.mark.parametrize(
+    "agg",
+    [
+        "min",
+        "max",
+        "count",
+        "sum",
+        "prod",
+        "var",
+        "mean",
+        "median",
+        "ohlc",
+        "cumprod",
+        "cumsum",
+        "shift",
+        "any",
+        "all",
+        "quantile",
+        "first",
+        "last",
+        "rank",
+        "cummin",
+        "cummax",
+    ],
+)
+def test_read_only_buffer_source_agg(agg):
+    # https://github.com/pandas-dev/pandas/issues/36014
+    df = DataFrame(
+        {
+            "sepal_length": [5.1, 4.9, 4.7, 4.6, 5.0],
+            "species": ["setosa", "setosa", "setosa", "setosa", "setosa"],
+        }
+    )
+    df._mgr.arrays[0].flags.writeable = False
+
+    result = df.groupby(["species"]).agg({"sepal_length": agg})
+    expected = df.copy().groupby(["species"]).agg({"sepal_length": agg})
+
+    tm.assert_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "op_name",
+    [
+        "count",
+        "sum",
+        "std",
+        "var",
+        "sem",
+        "mean",
+        "median",
+        "prod",
+        "min",
+        "max",
+    ],
+)
+def test_cython_agg_nullable_int(op_name):
+    # ensure that the cython-based aggregations don't fail for nullable dtype
+    # (eg https://github.com/pandas-dev/pandas/issues/37415)
+    df = DataFrame(
+        {
+            "A": ["A", "B"] * 5,
+            "B": pd.array([1, 2, 3, 4, 5, 6, 7, 8, 9, pd.NA], dtype="Int64"),
+        }
+    )
+    result = getattr(df.groupby("A")["B"], op_name)()
+    df2 = df.assign(B=df["B"].astype("float64"))
+    expected = getattr(df2.groupby("A")["B"], op_name)()
+    if op_name in ("mean", "median"):
+        convert_integer = False
+    else:
+        convert_integer = True
+    expected = expected.convert_dtypes(convert_integer=convert_integer)
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize("dtype", ["Int64", "Float64", "boolean"])
+def test_count_masked_returns_masked_dtype(dtype):
+    df = DataFrame(
+        {
+            "A": [1, 1],
+            "B": pd.array([1, pd.NA], dtype=dtype),
+            "C": pd.array([1, 1], dtype=dtype),
+        }
+    )
+    result = df.groupby("A").count()
+    expected = DataFrame(
+        [[1, 2]], index=Index([1], name="A"), columns=["B", "C"], dtype="Int64"
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("with_na", [True, False])
+@pytest.mark.parametrize(
+    "op_name, action",
+    [
+        # ("count", "always_int"),
+        ("sum", "large_int"),
+        # ("std", "always_float"),
+        ("var", "always_float"),
+        # ("sem", "always_float"),
+        ("mean", "always_float"),
+        ("median", "always_float"),
+        ("prod", "large_int"),
+        ("min", "preserve"),
+        ("max", "preserve"),
+        ("first", "preserve"),
+        ("last", "preserve"),
+    ],
+)
+@pytest.mark.parametrize(
+    "data",
+    [
+        pd.array([1, 2, 3, 4], dtype="Int64"),
+        pd.array([1, 2, 3, 4], dtype="Int8"),
+        pd.array([0.1, 0.2, 0.3, 0.4], dtype="Float32"),
+        pd.array([0.1, 0.2, 0.3, 0.4], dtype="Float64"),
+        pd.array([True, True, False, False], dtype="boolean"),
+    ],
+)
+def test_cython_agg_EA_known_dtypes(data, op_name, action, with_na):
+    if with_na:
+        data[3] = pd.NA
+
+    df = DataFrame({"key": ["a", "a", "b", "b"], "col": data})
+    grouped = df.groupby("key")
+
+    if action == "always_int":
+        # always Int64
+        expected_dtype = pd.Int64Dtype()
+    elif action == "large_int":
+        # for any int/bool use Int64, for float preserve dtype
+        if is_float_dtype(data.dtype):
+            expected_dtype = data.dtype
+        elif is_integer_dtype(data.dtype):
+            # match the numpy dtype we'd get with the non-nullable analogue
+            expected_dtype = data.dtype
+        else:
+            expected_dtype = pd.Int64Dtype()
+    elif action == "always_float":
+        # for any int/bool use Float64, for float preserve dtype
+        if is_float_dtype(data.dtype):
+            expected_dtype = data.dtype
+        else:
+            expected_dtype = pd.Float64Dtype()
+    elif action == "preserve":
+        expected_dtype = data.dtype
+
+    result = getattr(grouped, op_name)()
+    assert result["col"].dtype == expected_dtype
+
+    result = grouped.aggregate(op_name)
+    assert result["col"].dtype == expected_dtype
+
+    result = getattr(grouped["col"], op_name)()
+    assert result.dtype == expected_dtype
+
+    result = grouped["col"].aggregate(op_name)
+    assert result.dtype == expected_dtype

venv/lib/python3.12/site-packages/pandas/tests/groupby/aggregate/test_numba.py

@@ -0,0 +1,392 @@
+import numpy as np
+import pytest
+
+from pandas.errors import NumbaUtilError
+
+from pandas import (
+    DataFrame,
+    Index,
+    NamedAgg,
+    Series,
+    option_context,
+)
+import pandas._testing as tm
+
+pytestmark = pytest.mark.single_cpu
+
+
+def test_correct_function_signature():
+    pytest.importorskip("numba")
+
+    def incorrect_function(x):
+        return sum(x) * 2.7
+
+    data = DataFrame(
+        {"key": ["a", "a", "b", "b", "a"], "data": [1.0, 2.0, 3.0, 4.0, 5.0]},
+        columns=["key", "data"],
+    )
+    with pytest.raises(NumbaUtilError, match="The first 2"):
+        data.groupby("key").agg(incorrect_function, engine="numba")
+
+    with pytest.raises(NumbaUtilError, match="The first 2"):
+        data.groupby("key")["data"].agg(incorrect_function, engine="numba")
+
+
+def test_check_nopython_kwargs():
+    pytest.importorskip("numba")
+
+    def incorrect_function(values, index):
+        return sum(values) * 2.7
+
+    data = DataFrame(
+        {"key": ["a", "a", "b", "b", "a"], "data": [1.0, 2.0, 3.0, 4.0, 5.0]},
+        columns=["key", "data"],
+    )
+    with pytest.raises(NumbaUtilError, match="numba does not support"):
+        data.groupby("key").agg(incorrect_function, engine="numba", a=1)
+
+    with pytest.raises(NumbaUtilError, match="numba does not support"):
+        data.groupby("key")["data"].agg(incorrect_function, engine="numba", a=1)
+
+
+@pytest.mark.filterwarnings("ignore")
+# Filter warnings when parallel=True and the function can't be parallelized by Numba
+@pytest.mark.parametrize("jit", [True, False])
+@pytest.mark.parametrize("pandas_obj", ["Series", "DataFrame"])
+@pytest.mark.parametrize("as_index", [True, False])
+def test_numba_vs_cython(jit, pandas_obj, nogil, parallel, nopython, as_index):
+    pytest.importorskip("numba")
+
+    def func_numba(values, index):
+        return np.mean(values) * 2.7
+
+    if jit:
+        # Test accepted jitted functions
+        import numba
+
+        func_numba = numba.jit(func_numba)
+
+    data = DataFrame(
+        {0: ["a", "a", "b", "b", "a"], 1: [1.0, 2.0, 3.0, 4.0, 5.0]}, columns=[0, 1]
+    )
+    engine_kwargs = {"nogil": nogil, "parallel": parallel, "nopython": nopython}
+    grouped = data.groupby(0, as_index=as_index)
+    if pandas_obj == "Series":
+        grouped = grouped[1]
+
+    result = grouped.agg(func_numba, engine="numba", engine_kwargs=engine_kwargs)
+    expected = grouped.agg(lambda x: np.mean(x) * 2.7, engine="cython")
+
+    tm.assert_equal(result, expected)
+
+
+@pytest.mark.filterwarnings("ignore")
+# Filter warnings when parallel=True and the function can't be parallelized by Numba
+@pytest.mark.parametrize("jit", [True, False])
+@pytest.mark.parametrize("pandas_obj", ["Series", "DataFrame"])
+def test_cache(jit, pandas_obj, nogil, parallel, nopython):
+    # Test that the functions are cached correctly if we switch functions
+    pytest.importorskip("numba")
+
+    def func_1(values, index):
+        return np.mean(values) - 3.4
+
+    def func_2(values, index):
+        return np.mean(values) * 2.7
+
+    if jit:
+        import numba
+
+        func_1 = numba.jit(func_1)
+        func_2 = numba.jit(func_2)
+
+    data = DataFrame(
+        {0: ["a", "a", "b", "b", "a"], 1: [1.0, 2.0, 3.0, 4.0, 5.0]}, columns=[0, 1]
+    )
+    engine_kwargs = {"nogil": nogil, "parallel": parallel, "nopython": nopython}
+    grouped = data.groupby(0)
+    if pandas_obj == "Series":
+        grouped = grouped[1]
+
+    result = grouped.agg(func_1, engine="numba", engine_kwargs=engine_kwargs)
+    expected = grouped.agg(lambda x: np.mean(x) - 3.4, engine="cython")
+    tm.assert_equal(result, expected)
+
+    # Add func_2 to the cache
+    result = grouped.agg(func_2, engine="numba", engine_kwargs=engine_kwargs)
+    expected = grouped.agg(lambda x: np.mean(x) * 2.7, engine="cython")
+    tm.assert_equal(result, expected)
+
+    # Retest func_1 which should use the cache
+    result = grouped.agg(func_1, engine="numba", engine_kwargs=engine_kwargs)
+    expected = grouped.agg(lambda x: np.mean(x) - 3.4, engine="cython")
+    tm.assert_equal(result, expected)
+
+
+def test_use_global_config():
+    pytest.importorskip("numba")
+
+    def func_1(values, index):
+        return np.mean(values) - 3.4
+
+    data = DataFrame(
+        {0: ["a", "a", "b", "b", "a"], 1: [1.0, 2.0, 3.0, 4.0, 5.0]}, columns=[0, 1]
+    )
+    grouped = data.groupby(0)
+    expected = grouped.agg(func_1, engine="numba")
+    with option_context("compute.use_numba", True):
+        result = grouped.agg(func_1, engine=None)
+    tm.assert_frame_equal(expected, result)
+
+
+@pytest.mark.parametrize(
+    "agg_kwargs",
+    [
+        {"func": ["min", "max"]},
+        {"func": "min"},
+        {"func": {1: ["min", "max"], 2: "sum"}},
+        {"bmin": NamedAgg(column=1, aggfunc="min")},
+    ],
+)
+def test_multifunc_numba_vs_cython_frame(agg_kwargs):
+    pytest.importorskip("numba")
+    data = DataFrame(
+        {
+            0: ["a", "a", "b", "b", "a"],
+            1: [1.0, 2.0, 3.0, 4.0, 5.0],
+            2: [1, 2, 3, 4, 5],
+        },
+        columns=[0, 1, 2],
+    )
+    grouped = data.groupby(0)
+    result = grouped.agg(**agg_kwargs, engine="numba")
+    expected = grouped.agg(**agg_kwargs, engine="cython")
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "agg_kwargs,expected_func",
+    [
+        ({"func": lambda values, index: values.sum()}, "sum"),
+        # FIXME
+        pytest.param(
+            {
+                "func": [
+                    lambda values, index: values.sum(),
+                    lambda values, index: values.min(),
+                ]
+            },
+            ["sum", "min"],
+            marks=pytest.mark.xfail(
+                reason="This doesn't work yet! Fails in nopython pipeline!"
+            ),
+        ),
+    ],
+)
+def test_multifunc_numba_udf_frame(agg_kwargs, expected_func):
+    pytest.importorskip("numba")
+    data = DataFrame(
+        {
+            0: ["a", "a", "b", "b", "a"],
+            1: [1.0, 2.0, 3.0, 4.0, 5.0],
+            2: [1, 2, 3, 4, 5],
+        },
+        columns=[0, 1, 2],
+    )
+    grouped = data.groupby(0)
+    result = grouped.agg(**agg_kwargs, engine="numba")
+    expected = grouped.agg(expected_func, engine="cython")
+    # check_dtype can be removed if GH 44952 is addressed
+    # Currently, UDFs still always return float64 while reductions can preserve dtype
+    tm.assert_frame_equal(result, expected, check_dtype=False)
+
+
+@pytest.mark.parametrize(
+    "agg_kwargs",
+    [{"func": ["min", "max"]}, {"func": "min"}, {"min_val": "min", "max_val": "max"}],
+)
+def test_multifunc_numba_vs_cython_series(agg_kwargs):
+    pytest.importorskip("numba")
+    labels = ["a", "a", "b", "b", "a"]
+    data = Series([1.0, 2.0, 3.0, 4.0, 5.0])
+    grouped = data.groupby(labels)
+    agg_kwargs["engine"] = "numba"
+    result = grouped.agg(**agg_kwargs)
+    agg_kwargs["engine"] = "cython"
+    expected = grouped.agg(**agg_kwargs)
+    if isinstance(expected, DataFrame):
+        tm.assert_frame_equal(result, expected)
+    else:
+        tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.single_cpu
+@pytest.mark.parametrize(
+    "data,agg_kwargs",
+    [
+        (Series([1.0, 2.0, 3.0, 4.0, 5.0]), {"func": ["min", "max"]}),
+        (Series([1.0, 2.0, 3.0, 4.0, 5.0]), {"func": "min"}),
+        (
+            DataFrame(
+                {1: [1.0, 2.0, 3.0, 4.0, 5.0], 2: [1, 2, 3, 4, 5]}, columns=[1, 2]
+            ),
+            {"func": ["min", "max"]},
+        ),
+        (
+            DataFrame(
+                {1: [1.0, 2.0, 3.0, 4.0, 5.0], 2: [1, 2, 3, 4, 5]}, columns=[1, 2]
+            ),
+            {"func": "min"},
+        ),
+        (
+            DataFrame(
+                {1: [1.0, 2.0, 3.0, 4.0, 5.0], 2: [1, 2, 3, 4, 5]}, columns=[1, 2]
+            ),
+            {"func": {1: ["min", "max"], 2: "sum"}},
+        ),
+        (
+            DataFrame(
+                {1: [1.0, 2.0, 3.0, 4.0, 5.0], 2: [1, 2, 3, 4, 5]}, columns=[1, 2]
+            ),
+            {"min_col": NamedAgg(column=1, aggfunc="min")},
+        ),
+    ],
+)
+def test_multifunc_numba_kwarg_propagation(data, agg_kwargs):
+    pytest.importorskip("numba")
+    labels = ["a", "a", "b", "b", "a"]
+    grouped = data.groupby(labels)
+    result = grouped.agg(**agg_kwargs, engine="numba", engine_kwargs={"parallel": True})
+    expected = grouped.agg(**agg_kwargs, engine="numba")
+    if isinstance(expected, DataFrame):
+        tm.assert_frame_equal(result, expected)
+    else:
+        tm.assert_series_equal(result, expected)
+
+
+def test_args_not_cached():
+    # GH 41647
+    pytest.importorskip("numba")
+
+    def sum_last(values, index, n):
+        return values[-n:].sum()
+
+    df = DataFrame({"id": [0, 0, 1, 1], "x": [1, 1, 1, 1]})
+    grouped_x = df.groupby("id")["x"]
+    result = grouped_x.agg(sum_last, 1, engine="numba")
+    expected = Series([1.0] * 2, name="x", index=Index([0, 1], name="id"))
+    tm.assert_series_equal(result, expected)
+
+    result = grouped_x.agg(sum_last, 2, engine="numba")
+    expected = Series([2.0] * 2, name="x", index=Index([0, 1], name="id"))
+    tm.assert_series_equal(result, expected)
+
+
+def test_index_data_correctly_passed():
+    # GH 43133
+    pytest.importorskip("numba")
+
+    def f(values, index):
+        return np.mean(index)
+
+    df = DataFrame({"group": ["A", "A", "B"], "v": [4, 5, 6]}, index=[-1, -2, -3])
+    result = df.groupby("group").aggregate(f, engine="numba")
+    expected = DataFrame(
+        [-1.5, -3.0], columns=["v"], index=Index(["A", "B"], name="group")
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+def test_engine_kwargs_not_cached():
+    # If the user passes a different set of engine_kwargs don't return the same
+    # jitted function
+    pytest.importorskip("numba")
+    nogil = True
+    parallel = False
+    nopython = True
+
+    def func_kwargs(values, index):
+        return nogil + parallel + nopython
+
+    engine_kwargs = {"nopython": nopython, "nogil": nogil, "parallel": parallel}
+    df = DataFrame({"value": [0, 0, 0]})
+    result = df.groupby(level=0).aggregate(
+        func_kwargs, engine="numba", engine_kwargs=engine_kwargs
+    )
+    expected = DataFrame({"value": [2.0, 2.0, 2.0]})
+    tm.assert_frame_equal(result, expected)
+
+    nogil = False
+    engine_kwargs = {"nopython": nopython, "nogil": nogil, "parallel": parallel}
+    result = df.groupby(level=0).aggregate(
+        func_kwargs, engine="numba", engine_kwargs=engine_kwargs
+    )
+    expected = DataFrame({"value": [1.0, 1.0, 1.0]})
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.filterwarnings("ignore")
+def test_multiindex_one_key(nogil, parallel, nopython):
+    pytest.importorskip("numba")
+
+    def numba_func(values, index):
+        return 1
+
+    df = DataFrame([{"A": 1, "B": 2, "C": 3}]).set_index(["A", "B"])
+    engine_kwargs = {"nopython": nopython, "nogil": nogil, "parallel": parallel}
+    result = df.groupby("A").agg(
+        numba_func, engine="numba", engine_kwargs=engine_kwargs
+    )
+    expected = DataFrame([1.0], index=Index([1], name="A"), columns=["C"])
+    tm.assert_frame_equal(result, expected)
+
+
+def test_multiindex_multi_key_not_supported(nogil, parallel, nopython):
+    pytest.importorskip("numba")
+
+    def numba_func(values, index):
+        return 1
+
+    df = DataFrame([{"A": 1, "B": 2, "C": 3}]).set_index(["A", "B"])
+    engine_kwargs = {"nopython": nopython, "nogil": nogil, "parallel": parallel}
+    with pytest.raises(NotImplementedError, match="more than 1 grouping labels"):
+        df.groupby(["A", "B"]).agg(
+            numba_func, engine="numba", engine_kwargs=engine_kwargs
+        )
+
+
+def test_multilabel_numba_vs_cython(numba_supported_reductions):
+    pytest.importorskip("numba")
+    reduction, kwargs = numba_supported_reductions
+    df = DataFrame(
+        {
+            "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"],
+            "B": ["one", "one", "two", "three", "two", "two", "one", "three"],
+            "C": np.random.default_rng(2).standard_normal(8),
+            "D": np.random.default_rng(2).standard_normal(8),
+        }
+    )
+    gb = df.groupby(["A", "B"])
+    res_agg = gb.agg(reduction, engine="numba", **kwargs)
+    expected_agg = gb.agg(reduction, engine="cython", **kwargs)
+    tm.assert_frame_equal(res_agg, expected_agg)
+    # Test that calling the aggregation directly also works
+    direct_res = getattr(gb, reduction)(engine="numba", **kwargs)
+    direct_expected = getattr(gb, reduction)(engine="cython", **kwargs)
+    tm.assert_frame_equal(direct_res, direct_expected)
+
+
+def test_multilabel_udf_numba_vs_cython():
+    pytest.importorskip("numba")
+    df = DataFrame(
+        {
+            "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"],
+            "B": ["one", "one", "two", "three", "two", "two", "one", "three"],
+            "C": np.random.default_rng(2).standard_normal(8),
+            "D": np.random.default_rng(2).standard_normal(8),
+        }
+    )
+    gb = df.groupby(["A", "B"])
+    result = gb.agg(lambda values, index: values.min(), engine="numba")
+    expected = gb.agg(lambda x: x.min(), engine="cython")
+    tm.assert_frame_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/aggregate/test_other.py

@@ -0,0 +1,675 @@
+"""
+test all other .agg behavior
+"""
+
+import datetime as dt
+from functools import partial
+
+import numpy as np
+import pytest
+
+from pandas.errors import SpecificationError
+
+import pandas as pd
+from pandas import (
+    DataFrame,
+    Index,
+    MultiIndex,
+    PeriodIndex,
+    Series,
+    date_range,
+    period_range,
+)
+import pandas._testing as tm
+
+from pandas.io.formats.printing import pprint_thing
+
+
+def test_agg_partial_failure_raises():
+    # GH#43741
+
+    df = DataFrame(
+        {
+            "data1": np.random.default_rng(2).standard_normal(5),
+            "data2": np.random.default_rng(2).standard_normal(5),
+            "key1": ["a", "a", "b", "b", "a"],
+            "key2": ["one", "two", "one", "two", "one"],
+        }
+    )
+    grouped = df.groupby("key1")
+
+    def peak_to_peak(arr):
+        return arr.max() - arr.min()
+
+    with pytest.raises(TypeError, match="unsupported operand type"):
+        grouped.agg([peak_to_peak])
+
+    with pytest.raises(TypeError, match="unsupported operand type"):
+        grouped.agg(peak_to_peak)
+
+
+def test_agg_datetimes_mixed():
+    data = [[1, "2012-01-01", 1.0], [2, "2012-01-02", 2.0], [3, None, 3.0]]
+
+    df1 = DataFrame(
+        {
+            "key": [x[0] for x in data],
+            "date": [x[1] for x in data],
+            "value": [x[2] for x in data],
+        }
+    )
+
+    data = [
+        [
+            row[0],
+            (dt.datetime.strptime(row[1], "%Y-%m-%d").date() if row[1] else None),
+            row[2],
+        ]
+        for row in data
+    ]
+
+    df2 = DataFrame(
+        {
+            "key": [x[0] for x in data],
+            "date": [x[1] for x in data],
+            "value": [x[2] for x in data],
+        }
+    )
+
+    df1["weights"] = df1["value"] / df1["value"].sum()
+    gb1 = df1.groupby("date").aggregate("sum")
+
+    df2["weights"] = df1["value"] / df1["value"].sum()
+    gb2 = df2.groupby("date").aggregate("sum")
+
+    assert len(gb1) == len(gb2)
+
+
+def test_agg_period_index():
+    prng = period_range("2012-1-1", freq="M", periods=3)
+    df = DataFrame(np.random.default_rng(2).standard_normal((3, 2)), index=prng)
+    rs = df.groupby(level=0).sum()
+    assert isinstance(rs.index, PeriodIndex)
+
+    # GH 3579
+    index = period_range(start="1999-01", periods=5, freq="M")
+    s1 = Series(np.random.default_rng(2).random(len(index)), index=index)
+    s2 = Series(np.random.default_rng(2).random(len(index)), index=index)
+    df = DataFrame.from_dict({"s1": s1, "s2": s2})
+    grouped = df.groupby(df.index.month)
+    list(grouped)
+
+
+def test_agg_dict_parameter_cast_result_dtypes():
+    # GH 12821
+
+    df = DataFrame(
+        {
+            "class": ["A", "A", "B", "B", "C", "C", "D", "D"],
+            "time": date_range("1/1/2011", periods=8, freq="h"),
+        }
+    )
+    df.loc[[0, 1, 2, 5], "time"] = None
+
+    # test for `first` function
+    exp = df.loc[[0, 3, 4, 6]].set_index("class")
+    grouped = df.groupby("class")
+    tm.assert_frame_equal(grouped.first(), exp)
+    tm.assert_frame_equal(grouped.agg("first"), exp)
+    tm.assert_frame_equal(grouped.agg({"time": "first"}), exp)
+    tm.assert_series_equal(grouped.time.first(), exp["time"])
+    tm.assert_series_equal(grouped.time.agg("first"), exp["time"])
+
+    # test for `last` function
+    exp = df.loc[[0, 3, 4, 7]].set_index("class")
+    grouped = df.groupby("class")
+    tm.assert_frame_equal(grouped.last(), exp)
+    tm.assert_frame_equal(grouped.agg("last"), exp)
+    tm.assert_frame_equal(grouped.agg({"time": "last"}), exp)
+    tm.assert_series_equal(grouped.time.last(), exp["time"])
+    tm.assert_series_equal(grouped.time.agg("last"), exp["time"])
+
+    # count
+    exp = Series([2, 2, 2, 2], index=Index(list("ABCD"), name="class"), name="time")
+    tm.assert_series_equal(grouped.time.agg(len), exp)
+    tm.assert_series_equal(grouped.time.size(), exp)
+
+    exp = Series([0, 1, 1, 2], index=Index(list("ABCD"), name="class"), name="time")
+    tm.assert_series_equal(grouped.time.count(), exp)
+
+
+def test_agg_cast_results_dtypes():
+    # similar to GH12821
+    # xref #11444
+    u = [dt.datetime(2015, x + 1, 1) for x in range(12)]
+    v = list("aaabbbbbbccd")
+    df = DataFrame({"X": v, "Y": u})
+
+    result = df.groupby("X")["Y"].agg(len)
+    expected = df.groupby("X")["Y"].count()
+    tm.assert_series_equal(result, expected)
+
+
+def test_aggregate_float64_no_int64():
+    # see gh-11199
+    df = DataFrame({"a": [1, 2, 3, 4, 5], "b": [1, 2, 2, 4, 5], "c": [1, 2, 3, 4, 5]})
+
+    expected = DataFrame({"a": [1, 2.5, 4, 5]}, index=[1, 2, 4, 5])
+    expected.index.name = "b"
+
+    result = df.groupby("b")[["a"]].mean()
+    tm.assert_frame_equal(result, expected)
+
+    expected = DataFrame({"a": [1, 2.5, 4, 5], "c": [1, 2.5, 4, 5]}, index=[1, 2, 4, 5])
+    expected.index.name = "b"
+
+    result = df.groupby("b")[["a", "c"]].mean()
+    tm.assert_frame_equal(result, expected)
+
+
+def test_aggregate_api_consistency():
+    # GH 9052
+    # make sure that the aggregates via dict
+    # are consistent
+    df = DataFrame(
+        {
+            "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"],
+            "B": ["one", "one", "two", "two", "two", "two", "one", "two"],
+            "C": np.random.default_rng(2).standard_normal(8) + 1.0,
+            "D": np.arange(8),
+        }
+    )
+
+    grouped = df.groupby(["A", "B"])
+    c_mean = grouped["C"].mean()
+    c_sum = grouped["C"].sum()
+    d_mean = grouped["D"].mean()
+    d_sum = grouped["D"].sum()
+
+    result = grouped["D"].agg(["sum", "mean"])
+    expected = pd.concat([d_sum, d_mean], axis=1)
+    expected.columns = ["sum", "mean"]
+    tm.assert_frame_equal(result, expected, check_like=True)
+
+    result = grouped.agg(["sum", "mean"])
+    expected = pd.concat([c_sum, c_mean, d_sum, d_mean], axis=1)
+    expected.columns = MultiIndex.from_product([["C", "D"], ["sum", "mean"]])
+    tm.assert_frame_equal(result, expected, check_like=True)
+
+    result = grouped[["D", "C"]].agg(["sum", "mean"])
+    expected = pd.concat([d_sum, d_mean, c_sum, c_mean], axis=1)
+    expected.columns = MultiIndex.from_product([["D", "C"], ["sum", "mean"]])
+    tm.assert_frame_equal(result, expected, check_like=True)
+
+    result = grouped.agg({"C": "mean", "D": "sum"})
+    expected = pd.concat([d_sum, c_mean], axis=1)
+    tm.assert_frame_equal(result, expected, check_like=True)
+
+    result = grouped.agg({"C": ["mean", "sum"], "D": ["mean", "sum"]})
+    expected = pd.concat([c_mean, c_sum, d_mean, d_sum], axis=1)
+    expected.columns = MultiIndex.from_product([["C", "D"], ["mean", "sum"]])
+
+    msg = r"Column\(s\) \['r', 'r2'\] do not exist"
+    with pytest.raises(KeyError, match=msg):
+        grouped[["D", "C"]].agg({"r": "sum", "r2": "mean"})
+
+
+def test_agg_dict_renaming_deprecation():
+    # 15931
+    df = DataFrame({"A": [1, 1, 1, 2, 2], "B": range(5), "C": range(5)})
+
+    msg = r"nested renamer is not supported"
+    with pytest.raises(SpecificationError, match=msg):
+        df.groupby("A").agg(
+            {"B": {"foo": ["sum", "max"]}, "C": {"bar": ["count", "min"]}}
+        )
+
+    msg = r"Column\(s\) \['ma'\] do not exist"
+    with pytest.raises(KeyError, match=msg):
+        df.groupby("A")[["B", "C"]].agg({"ma": "max"})
+
+    msg = r"nested renamer is not supported"
+    with pytest.raises(SpecificationError, match=msg):
+        df.groupby("A").B.agg({"foo": "count"})
+
+
+def test_agg_compat():
+    # GH 12334
+    df = DataFrame(
+        {
+            "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"],
+            "B": ["one", "one", "two", "two", "two", "two", "one", "two"],
+            "C": np.random.default_rng(2).standard_normal(8) + 1.0,
+            "D": np.arange(8),
+        }
+    )
+
+    g = df.groupby(["A", "B"])
+
+    msg = r"nested renamer is not supported"
+    with pytest.raises(SpecificationError, match=msg):
+        g["D"].agg({"C": ["sum", "std"]})
+
+    with pytest.raises(SpecificationError, match=msg):
+        g["D"].agg({"C": "sum", "D": "std"})
+
+
+def test_agg_nested_dicts():
+    # API change for disallowing these types of nested dicts
+    df = DataFrame(
+        {
+            "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"],
+            "B": ["one", "one", "two", "two", "two", "two", "one", "two"],
+            "C": np.random.default_rng(2).standard_normal(8) + 1.0,
+            "D": np.arange(8),
+        }
+    )
+
+    g = df.groupby(["A", "B"])
+
+    msg = r"nested renamer is not supported"
+    with pytest.raises(SpecificationError, match=msg):
+        g.aggregate({"r1": {"C": ["mean", "sum"]}, "r2": {"D": ["mean", "sum"]}})
+
+    with pytest.raises(SpecificationError, match=msg):
+        g.agg({"C": {"ra": ["mean", "std"]}, "D": {"rb": ["mean", "std"]}})
+
+    # same name as the original column
+    # GH9052
+    with pytest.raises(SpecificationError, match=msg):
+        g["D"].agg({"result1": np.sum, "result2": np.mean})
+
+    with pytest.raises(SpecificationError, match=msg):
+        g["D"].agg({"D": np.sum, "result2": np.mean})
+
+
+def test_agg_item_by_item_raise_typeerror():
+    df = DataFrame(np.random.default_rng(2).integers(10, size=(20, 10)))
+
+    def raiseException(df):
+        pprint_thing("----------------------------------------")
+        pprint_thing(df.to_string())
+        raise TypeError("test")
+
+    with pytest.raises(TypeError, match="test"):
+        df.groupby(0).agg(raiseException)
+
+
+def test_series_agg_multikey():
+    ts = Series(
+        np.arange(10, dtype=np.float64), index=date_range("2020-01-01", periods=10)
+    )
+    grouped = ts.groupby([lambda x: x.year, lambda x: x.month])
+
+    result = grouped.agg("sum")
+    expected = grouped.sum()
+    tm.assert_series_equal(result, expected)
+
+
+def test_series_agg_multi_pure_python():
+    data = DataFrame(
+        {
+            "A": [
+                "foo",
+                "foo",
+                "foo",
+                "foo",
+                "bar",
+                "bar",
+                "bar",
+                "bar",
+                "foo",
+                "foo",
+                "foo",
+            ],
+            "B": [
+                "one",
+                "one",
+                "one",
+                "two",
+                "one",
+                "one",
+                "one",
+                "two",
+                "two",
+                "two",
+                "one",
+            ],
+            "C": [
+                "dull",
+                "dull",
+                "shiny",
+                "dull",
+                "dull",
+                "shiny",
+                "shiny",
+                "dull",
+                "shiny",
+                "shiny",
+                "shiny",
+            ],
+            "D": np.random.default_rng(2).standard_normal(11),
+            "E": np.random.default_rng(2).standard_normal(11),
+            "F": np.random.default_rng(2).standard_normal(11),
+        }
+    )
+
+    def bad(x):
+        assert len(x.values.base) > 0
+        return "foo"
+
+    result = data.groupby(["A", "B"]).agg(bad)
+    expected = data.groupby(["A", "B"]).agg(lambda x: "foo")
+    tm.assert_frame_equal(result, expected)
+
+
+def test_agg_consistency():
+    # agg with ([]) and () not consistent
+    # GH 6715
+    def P1(a):
+        return np.percentile(a.dropna(), q=1)
+
+    df = DataFrame(
+        {
+            "col1": [1, 2, 3, 4],
+            "col2": [10, 25, 26, 31],
+            "date": [
+                dt.date(2013, 2, 10),
+                dt.date(2013, 2, 10),
+                dt.date(2013, 2, 11),
+                dt.date(2013, 2, 11),
+            ],
+        }
+    )
+
+    g = df.groupby("date")
+
+    expected = g.agg([P1])
+    expected.columns = expected.columns.levels[0]
+
+    result = g.agg(P1)
+    tm.assert_frame_equal(result, expected)
+
+
+def test_agg_callables():
+    # GH 7929
+    df = DataFrame({"foo": [1, 2], "bar": [3, 4]}).astype(np.int64)
+
+    class fn_class:
+        def __call__(self, x):
+            return sum(x)
+
+    equiv_callables = [
+        sum,
+        np.sum,
+        lambda x: sum(x),
+        lambda x: x.sum(),
+        partial(sum),
+        fn_class(),
+    ]
+
+    expected = df.groupby("foo").agg("sum")
+    for ecall in equiv_callables:
+        warn = FutureWarning if ecall is sum or ecall is np.sum else None
+        msg = "using DataFrameGroupBy.sum"
+        with tm.assert_produces_warning(warn, match=msg):
+            result = df.groupby("foo").agg(ecall)
+        tm.assert_frame_equal(result, expected)
+
+
+def test_agg_over_numpy_arrays():
+    # GH 3788
+    df = DataFrame(
+        [
+            [1, np.array([10, 20, 30])],
+            [1, np.array([40, 50, 60])],
+            [2, np.array([20, 30, 40])],
+        ],
+        columns=["category", "arraydata"],
+    )
+    gb = df.groupby("category")
+
+    expected_data = [[np.array([50, 70, 90])], [np.array([20, 30, 40])]]
+    expected_index = Index([1, 2], name="category")
+    expected_column = ["arraydata"]
+    expected = DataFrame(expected_data, index=expected_index, columns=expected_column)
+
+    alt = gb.sum(numeric_only=False)
+    tm.assert_frame_equal(alt, expected)
+
+    result = gb.agg("sum", numeric_only=False)
+    tm.assert_frame_equal(result, expected)
+
+    # FIXME: the original version of this test called `gb.agg(sum)`
+    #  and that raises TypeError if `numeric_only=False` is passed
+
+
+@pytest.mark.parametrize("as_period", [True, False])
+def test_agg_tzaware_non_datetime_result(as_period):
+    # discussed in GH#29589, fixed in GH#29641, operating on tzaware values
+    #  with function that is not dtype-preserving
+    dti = date_range("2012-01-01", periods=4, tz="UTC")
+    if as_period:
+        dti = dti.tz_localize(None).to_period("D")
+
+    df = DataFrame({"a": [0, 0, 1, 1], "b": dti})
+    gb = df.groupby("a")
+
+    # Case that _does_ preserve the dtype
+    result = gb["b"].agg(lambda x: x.iloc[0])
+    expected = Series(dti[::2], name="b")
+    expected.index.name = "a"
+    tm.assert_series_equal(result, expected)
+
+    # Cases that do _not_ preserve the dtype
+    result = gb["b"].agg(lambda x: x.iloc[0].year)
+    expected = Series([2012, 2012], name="b")
+    expected.index.name = "a"
+    tm.assert_series_equal(result, expected)
+
+    result = gb["b"].agg(lambda x: x.iloc[-1] - x.iloc[0])
+    expected = Series([pd.Timedelta(days=1), pd.Timedelta(days=1)], name="b")
+    expected.index.name = "a"
+    if as_period:
+        expected = Series([pd.offsets.Day(1), pd.offsets.Day(1)], name="b")
+        expected.index.name = "a"
+    tm.assert_series_equal(result, expected)
+
+
+def test_agg_timezone_round_trip():
+    # GH 15426
+    ts = pd.Timestamp("2016-01-01 12:00:00", tz="US/Pacific")
+    df = DataFrame({"a": 1, "b": [ts + dt.timedelta(minutes=nn) for nn in range(10)]})
+
+    result1 = df.groupby("a")["b"].agg("min").iloc[0]
+    result2 = df.groupby("a")["b"].agg(lambda x: np.min(x)).iloc[0]
+    result3 = df.groupby("a")["b"].min().iloc[0]
+
+    assert result1 == ts
+    assert result2 == ts
+    assert result3 == ts
+
+    dates = [
+        pd.Timestamp(f"2016-01-0{i:d} 12:00:00", tz="US/Pacific") for i in range(1, 5)
+    ]
+    df = DataFrame({"A": ["a", "b"] * 2, "B": dates})
+    grouped = df.groupby("A")
+
+    ts = df["B"].iloc[0]
+    assert ts == grouped.nth(0)["B"].iloc[0]
+    assert ts == grouped.head(1)["B"].iloc[0]
+    assert ts == grouped.first()["B"].iloc[0]
+
+    # GH#27110 applying iloc should return a DataFrame
+    msg = "DataFrameGroupBy.apply operated on the grouping columns"
+    with tm.assert_produces_warning(DeprecationWarning, match=msg):
+        assert ts == grouped.apply(lambda x: x.iloc[0]).iloc[0, 1]
+
+    ts = df["B"].iloc[2]
+    assert ts == grouped.last()["B"].iloc[0]
+
+    # GH#27110 applying iloc should return a DataFrame
+    msg = "DataFrameGroupBy.apply operated on the grouping columns"
+    with tm.assert_produces_warning(DeprecationWarning, match=msg):
+        assert ts == grouped.apply(lambda x: x.iloc[-1]).iloc[0, 1]
+
+
+def test_sum_uint64_overflow():
+    # see gh-14758
+    # Convert to uint64 and don't overflow
+    df = DataFrame([[1, 2], [3, 4], [5, 6]], dtype=object)
+    df = df + 9223372036854775807
+
+    index = Index(
+        [9223372036854775808, 9223372036854775810, 9223372036854775812], dtype=np.uint64
+    )
+    expected = DataFrame(
+        {1: [9223372036854775809, 9223372036854775811, 9223372036854775813]},
+        index=index,
+        dtype=object,
+    )
+
+    expected.index.name = 0
+    result = df.groupby(0).sum(numeric_only=False)
+    tm.assert_frame_equal(result, expected)
+
+    # out column is non-numeric, so with numeric_only=True it is dropped
+    result2 = df.groupby(0).sum(numeric_only=True)
+    expected2 = expected[[]]
+    tm.assert_frame_equal(result2, expected2)
+
+
+@pytest.mark.parametrize(
+    "structure, expected",
+    [
+        (tuple, DataFrame({"C": {(1, 1): (1, 1, 1), (3, 4): (3, 4, 4)}})),
+        (list, DataFrame({"C": {(1, 1): [1, 1, 1], (3, 4): [3, 4, 4]}})),
+        (
+            lambda x: tuple(x),
+            DataFrame({"C": {(1, 1): (1, 1, 1), (3, 4): (3, 4, 4)}}),
+        ),
+        (
+            lambda x: list(x),
+            DataFrame({"C": {(1, 1): [1, 1, 1], (3, 4): [3, 4, 4]}}),
+        ),
+    ],
+)
+def test_agg_structs_dataframe(structure, expected):
+    df = DataFrame(
+        {"A": [1, 1, 1, 3, 3, 3], "B": [1, 1, 1, 4, 4, 4], "C": [1, 1, 1, 3, 4, 4]}
+    )
+
+    result = df.groupby(["A", "B"]).aggregate(structure)
+    expected.index.names = ["A", "B"]
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "structure, expected",
+    [
+        (tuple, Series([(1, 1, 1), (3, 4, 4)], index=[1, 3], name="C")),
+        (list, Series([[1, 1, 1], [3, 4, 4]], index=[1, 3], name="C")),
+        (lambda x: tuple(x), Series([(1, 1, 1), (3, 4, 4)], index=[1, 3], name="C")),
+        (lambda x: list(x), Series([[1, 1, 1], [3, 4, 4]], index=[1, 3], name="C")),
+    ],
+)
+def test_agg_structs_series(structure, expected):
+    # Issue #18079
+    df = DataFrame(
+        {"A": [1, 1, 1, 3, 3, 3], "B": [1, 1, 1, 4, 4, 4], "C": [1, 1, 1, 3, 4, 4]}
+    )
+
+    result = df.groupby("A")["C"].aggregate(structure)
+    expected.index.name = "A"
+    tm.assert_series_equal(result, expected)
+
+
+def test_agg_category_nansum(observed):
+    categories = ["a", "b", "c"]
+    df = DataFrame(
+        {"A": pd.Categorical(["a", "a", "b"], categories=categories), "B": [1, 2, 3]}
+    )
+    msg = "using SeriesGroupBy.sum"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        result = df.groupby("A", observed=observed).B.agg(np.nansum)
+    expected = Series(
+        [3, 3, 0],
+        index=pd.CategoricalIndex(["a", "b", "c"], categories=categories, name="A"),
+        name="B",
+    )
+    if observed:
+        expected = expected[expected != 0]
+    tm.assert_series_equal(result, expected)
+
+
+def test_agg_list_like_func():
+    # GH 18473
+    df = DataFrame({"A": [str(x) for x in range(3)], "B": [str(x) for x in range(3)]})
+    grouped = df.groupby("A", as_index=False, sort=False)
+    result = grouped.agg({"B": lambda x: list(x)})
+    expected = DataFrame(
+        {"A": [str(x) for x in range(3)], "B": [[str(x)] for x in range(3)]}
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+def test_agg_lambda_with_timezone():
+    # GH 23683
+    df = DataFrame(
+        {
+            "tag": [1, 1],
+            "date": [
+                pd.Timestamp("2018-01-01", tz="UTC"),
+                pd.Timestamp("2018-01-02", tz="UTC"),
+            ],
+        }
+    )
+    result = df.groupby("tag").agg({"date": lambda e: e.head(1)})
+    expected = DataFrame(
+        [pd.Timestamp("2018-01-01", tz="UTC")],
+        index=Index([1], name="tag"),
+        columns=["date"],
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "err_cls",
+    [
+        NotImplementedError,
+        RuntimeError,
+        KeyError,
+        IndexError,
+        OSError,
+        ValueError,
+        ArithmeticError,
+        AttributeError,
+    ],
+)
+def test_groupby_agg_err_catching(err_cls):
+    # make sure we suppress anything other than TypeError or AssertionError
+    #  in _python_agg_general
+
+    # Use a non-standard EA to make sure we don't go down ndarray paths
+    from pandas.tests.extension.decimal.array import (
+        DecimalArray,
+        make_data,
+        to_decimal,
+    )
+
+    data = make_data()[:5]
+    df = DataFrame(
+        {"id1": [0, 0, 0, 1, 1], "id2": [0, 1, 0, 1, 1], "decimals": DecimalArray(data)}
+    )
+
+    expected = Series(to_decimal([data[0], data[3]]))
+
+    def weird_func(x):
+        # weird function that raise something other than TypeError or IndexError
+        #  in _python_agg_general
+        if len(x) == 0:
+            raise err_cls
+        return x.iloc[0]
+
+    result = df["decimals"].groupby(df["id1"]).agg(weird_func)
+    tm.assert_series_equal(result, expected, check_names=False)

venv/lib/python3.12/site-packages/pandas/tests/groupby/conftest.py

@@ -0,0 +1,208 @@
+import numpy as np
+import pytest
+
+from pandas import (
+    DataFrame,
+    Index,
+    Series,
+    date_range,
+)
+from pandas.core.groupby.base import (
+    reduction_kernels,
+    transformation_kernels,
+)
+
+
+@pytest.fixture(params=[True, False])
+def sort(request):
+    return request.param
+
+
+@pytest.fixture(params=[True, False])
+def as_index(request):
+    return request.param
+
+
+@pytest.fixture(params=[True, False])
+def dropna(request):
+    return request.param
+
+
+@pytest.fixture(params=[True, False])
+def observed(request):
+    return request.param
+
+
+@pytest.fixture
+def df():
+    return DataFrame(
+        {
+            "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"],
+            "B": ["one", "one", "two", "three", "two", "two", "one", "three"],
+            "C": np.random.default_rng(2).standard_normal(8),
+            "D": np.random.default_rng(2).standard_normal(8),
+        }
+    )
+
+
+@pytest.fixture
+def ts():
+    return Series(
+        np.random.default_rng(2).standard_normal(30),
+        index=date_range("2000-01-01", periods=30, freq="B"),
+    )
+
+
+@pytest.fixture
+def tsframe():
+    return DataFrame(
+        np.random.default_rng(2).standard_normal((30, 4)),
+        columns=Index(list("ABCD"), dtype=object),
+        index=date_range("2000-01-01", periods=30, freq="B"),
+    )
+
+
+@pytest.fixture
+def three_group():
+    return DataFrame(
+        {
+            "A": [
+                "foo",
+                "foo",
+                "foo",
+                "foo",
+                "bar",
+                "bar",
+                "bar",
+                "bar",
+                "foo",
+                "foo",
+                "foo",
+            ],
+            "B": [
+                "one",
+                "one",
+                "one",
+                "two",
+                "one",
+                "one",
+                "one",
+                "two",
+                "two",
+                "two",
+                "one",
+            ],
+            "C": [
+                "dull",
+                "dull",
+                "shiny",
+                "dull",
+                "dull",
+                "shiny",
+                "shiny",
+                "dull",
+                "shiny",
+                "shiny",
+                "shiny",
+            ],
+            "D": np.random.default_rng(2).standard_normal(11),
+            "E": np.random.default_rng(2).standard_normal(11),
+            "F": np.random.default_rng(2).standard_normal(11),
+        }
+    )
+
+
+@pytest.fixture()
+def slice_test_df():
+    data = [
+        [0, "a", "a0_at_0"],
+        [1, "b", "b0_at_1"],
+        [2, "a", "a1_at_2"],
+        [3, "b", "b1_at_3"],
+        [4, "c", "c0_at_4"],
+        [5, "a", "a2_at_5"],
+        [6, "a", "a3_at_6"],
+        [7, "a", "a4_at_7"],
+    ]
+    df = DataFrame(data, columns=["Index", "Group", "Value"])
+    return df.set_index("Index")
+
+
+@pytest.fixture()
+def slice_test_grouped(slice_test_df):
+    return slice_test_df.groupby("Group", as_index=False)
+
+
+@pytest.fixture(params=sorted(reduction_kernels))
+def reduction_func(request):
+    """
+    yields the string names of all groupby reduction functions, one at a time.
+    """
+    return request.param
+
+
+@pytest.fixture(params=sorted(transformation_kernels))
+def transformation_func(request):
+    """yields the string names of all groupby transformation functions."""
+    return request.param
+
+
+@pytest.fixture(params=sorted(reduction_kernels) + sorted(transformation_kernels))
+def groupby_func(request):
+    """yields both aggregation and transformation functions."""
+    return request.param
+
+
+@pytest.fixture(params=[True, False])
+def parallel(request):
+    """parallel keyword argument for numba.jit"""
+    return request.param
+
+
+# Can parameterize nogil & nopython over True | False, but limiting per
+# https://github.com/pandas-dev/pandas/pull/41971#issuecomment-860607472
+
+
+@pytest.fixture(params=[False])
+def nogil(request):
+    """nogil keyword argument for numba.jit"""
+    return request.param
+
+
+@pytest.fixture(params=[True])
+def nopython(request):
+    """nopython keyword argument for numba.jit"""
+    return request.param
+
+
+@pytest.fixture(
+    params=[
+        ("mean", {}),
+        ("var", {"ddof": 1}),
+        ("var", {"ddof": 0}),
+        ("std", {"ddof": 1}),
+        ("std", {"ddof": 0}),
+        ("sum", {}),
+        ("min", {}),
+        ("max", {}),
+        ("sum", {"min_count": 2}),
+        ("min", {"min_count": 2}),
+        ("max", {"min_count": 2}),
+    ],
+    ids=[
+        "mean",
+        "var_1",
+        "var_0",
+        "std_1",
+        "std_0",
+        "sum",
+        "min",
+        "max",
+        "sum-min_count",
+        "min-min_count",
+        "max-min_count",
+    ],
+)
+def numba_supported_reductions(request):
+    """reductions supported with engine='numba'"""
+    return request.param

venv/lib/python3.12/site-packages/pandas/tests/groupby/methods/test_corrwith.py

@@ -0,0 +1,24 @@
+import numpy as np
+
+from pandas import (
+    DataFrame,
+    Index,
+    Series,
+)
+import pandas._testing as tm
+
+
+def test_corrwith_with_1_axis():
+    # GH 47723
+    df = DataFrame({"a": [1, 1, 2], "b": [3, 7, 4]})
+    gb = df.groupby("a")
+
+    msg = "DataFrameGroupBy.corrwith with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        result = gb.corrwith(df, axis=1)
+    index = Index(
+        data=[(1, 0), (1, 1), (1, 2), (2, 2), (2, 0), (2, 1)],
+        name=("a", None),
+    )
+    expected = Series([np.nan] * 6, index=index)
+    tm.assert_series_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/methods/test_describe.py

@@ -0,0 +1,297 @@
+import numpy as np
+import pytest
+
+import pandas as pd
+from pandas import (
+    DataFrame,
+    Index,
+    MultiIndex,
+    Series,
+    Timestamp,
+    date_range,
+)
+import pandas._testing as tm
+
+
+def test_apply_describe_bug(multiindex_dataframe_random_data):
+    grouped = multiindex_dataframe_random_data.groupby(level="first")
+    grouped.describe()  # it works!
+
+
+def test_series_describe_multikey():
+    ts = Series(
+        np.arange(10, dtype=np.float64), index=date_range("2020-01-01", periods=10)
+    )
+    grouped = ts.groupby([lambda x: x.year, lambda x: x.month])
+    result = grouped.describe()
+    tm.assert_series_equal(result["mean"], grouped.mean(), check_names=False)
+    tm.assert_series_equal(result["std"], grouped.std(), check_names=False)
+    tm.assert_series_equal(result["min"], grouped.min(), check_names=False)
+
+
+def test_series_describe_single():
+    ts = Series(
+        np.arange(10, dtype=np.float64), index=date_range("2020-01-01", periods=10)
+    )
+    grouped = ts.groupby(lambda x: x.month)
+    result = grouped.apply(lambda x: x.describe())
+    expected = grouped.describe().stack(future_stack=True)
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize("keys", ["key1", ["key1", "key2"]])
+def test_series_describe_as_index(as_index, keys):
+    # GH#49256
+    df = DataFrame(
+        {
+            "key1": ["one", "two", "two", "three", "two"],
+            "key2": ["one", "two", "two", "three", "two"],
+            "foo2": [1, 2, 4, 4, 6],
+        }
+    )
+    gb = df.groupby(keys, as_index=as_index)["foo2"]
+    result = gb.describe()
+    expected = DataFrame(
+        {
+            "key1": ["one", "three", "two"],
+            "count": [1.0, 1.0, 3.0],
+            "mean": [1.0, 4.0, 4.0],
+            "std": [np.nan, np.nan, 2.0],
+            "min": [1.0, 4.0, 2.0],
+            "25%": [1.0, 4.0, 3.0],
+            "50%": [1.0, 4.0, 4.0],
+            "75%": [1.0, 4.0, 5.0],
+            "max": [1.0, 4.0, 6.0],
+        }
+    )
+    if len(keys) == 2:
+        expected.insert(1, "key2", expected["key1"])
+    if as_index:
+        expected = expected.set_index(keys)
+    tm.assert_frame_equal(result, expected)
+
+
+def test_frame_describe_multikey(tsframe):
+    grouped = tsframe.groupby([lambda x: x.year, lambda x: x.month])
+    result = grouped.describe()
+    desc_groups = []
+    for col in tsframe:
+        group = grouped[col].describe()
+        # GH 17464 - Remove duplicate MultiIndex levels
+        group_col = MultiIndex(
+            levels=[[col], group.columns],
+            codes=[[0] * len(group.columns), range(len(group.columns))],
+        )
+        group = DataFrame(group.values, columns=group_col, index=group.index)
+        desc_groups.append(group)
+    expected = pd.concat(desc_groups, axis=1)
+    tm.assert_frame_equal(result, expected)
+
+    msg = "DataFrame.groupby with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        groupedT = tsframe.groupby({"A": 0, "B": 0, "C": 1, "D": 1}, axis=1)
+    result = groupedT.describe()
+    expected = tsframe.describe().T
+    # reverting the change from https://github.com/pandas-dev/pandas/pull/35441/
+    expected.index = MultiIndex(
+        levels=[[0, 1], expected.index],
+        codes=[[0, 0, 1, 1], range(len(expected.index))],
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+def test_frame_describe_tupleindex():
+    # GH 14848 - regression from 0.19.0 to 0.19.1
+    df1 = DataFrame(
+        {
+            "x": [1, 2, 3, 4, 5] * 3,
+            "y": [10, 20, 30, 40, 50] * 3,
+            "z": [100, 200, 300, 400, 500] * 3,
+        }
+    )
+    df1["k"] = [(0, 0, 1), (0, 1, 0), (1, 0, 0)] * 5
+    df2 = df1.rename(columns={"k": "key"})
+    msg = "Names should be list-like for a MultiIndex"
+    with pytest.raises(ValueError, match=msg):
+        df1.groupby("k").describe()
+    with pytest.raises(ValueError, match=msg):
+        df2.groupby("key").describe()
+
+
+def test_frame_describe_unstacked_format():
+    # GH 4792
+    prices = {
+        Timestamp("2011-01-06 10:59:05", tz=None): 24990,
+        Timestamp("2011-01-06 12:43:33", tz=None): 25499,
+        Timestamp("2011-01-06 12:54:09", tz=None): 25499,
+    }
+    volumes = {
+        Timestamp("2011-01-06 10:59:05", tz=None): 1500000000,
+        Timestamp("2011-01-06 12:43:33", tz=None): 5000000000,
+        Timestamp("2011-01-06 12:54:09", tz=None): 100000000,
+    }
+    df = DataFrame({"PRICE": prices, "VOLUME": volumes})
+    result = df.groupby("PRICE").VOLUME.describe()
+    data = [
+        df[df.PRICE == 24990].VOLUME.describe().values.tolist(),
+        df[df.PRICE == 25499].VOLUME.describe().values.tolist(),
+    ]
+    expected = DataFrame(
+        data,
+        index=Index([24990, 25499], name="PRICE"),
+        columns=["count", "mean", "std", "min", "25%", "50%", "75%", "max"],
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.filterwarnings(
+    "ignore:"
+    "indexing past lexsort depth may impact performance:"
+    "pandas.errors.PerformanceWarning"
+)
+@pytest.mark.parametrize("as_index", [True, False])
+@pytest.mark.parametrize("keys", [["a1"], ["a1", "a2"]])
+def test_describe_with_duplicate_output_column_names(as_index, keys):
+    # GH 35314
+    df = DataFrame(
+        {
+            "a1": [99, 99, 99, 88, 88, 88],
+            "a2": [99, 99, 99, 88, 88, 88],
+            "b": [1, 2, 3, 4, 5, 6],
+            "c": [10, 20, 30, 40, 50, 60],
+        },
+        columns=["a1", "a2", "b", "b"],
+        copy=False,
+    )
+    if keys == ["a1"]:
+        df = df.drop(columns="a2")
+
+    expected = (
+        DataFrame.from_records(
+            [
+                ("b", "count", 3.0, 3.0),
+                ("b", "mean", 5.0, 2.0),
+                ("b", "std", 1.0, 1.0),
+                ("b", "min", 4.0, 1.0),
+                ("b", "25%", 4.5, 1.5),
+                ("b", "50%", 5.0, 2.0),
+                ("b", "75%", 5.5, 2.5),
+                ("b", "max", 6.0, 3.0),
+                ("b", "count", 3.0, 3.0),
+                ("b", "mean", 5.0, 2.0),
+                ("b", "std", 1.0, 1.0),
+                ("b", "min", 4.0, 1.0),
+                ("b", "25%", 4.5, 1.5),
+                ("b", "50%", 5.0, 2.0),
+                ("b", "75%", 5.5, 2.5),
+                ("b", "max", 6.0, 3.0),
+            ],
+        )
+        .set_index([0, 1])
+        .T
+    )
+    expected.columns.names = [None, None]
+    if len(keys) == 2:
+        expected.index = MultiIndex(
+            levels=[[88, 99], [88, 99]], codes=[[0, 1], [0, 1]], names=["a1", "a2"]
+        )
+    else:
+        expected.index = Index([88, 99], name="a1")
+
+    if not as_index:
+        expected = expected.reset_index()
+
+    result = df.groupby(keys, as_index=as_index).describe()
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_describe_duplicate_columns():
+    # GH#50806
+    df = DataFrame([[0, 1, 2, 3]])
+    df.columns = [0, 1, 2, 0]
+    gb = df.groupby(df[1])
+    result = gb.describe(percentiles=[])
+
+    columns = ["count", "mean", "std", "min", "50%", "max"]
+    frames = [
+        DataFrame([[1.0, val, np.nan, val, val, val]], index=[1], columns=columns)
+        for val in (0.0, 2.0, 3.0)
+    ]
+    expected = pd.concat(frames, axis=1)
+    expected.columns = MultiIndex(
+        levels=[[0, 2], columns],
+        codes=[6 * [0] + 6 * [1] + 6 * [0], 3 * list(range(6))],
+    )
+    expected.index.names = [1]
+    tm.assert_frame_equal(result, expected)
+
+
+class TestGroupByNonCythonPaths:
+    # GH#5610 non-cython calls should not include the grouper
+    # Tests for code not expected to go through cython paths.
+
+    @pytest.fixture
+    def df(self):
+        df = DataFrame(
+            [[1, 2, "foo"], [1, np.nan, "bar"], [3, np.nan, "baz"]],
+            columns=["A", "B", "C"],
+        )
+        return df
+
+    @pytest.fixture
+    def gb(self, df):
+        gb = df.groupby("A")
+        return gb
+
+    @pytest.fixture
+    def gni(self, df):
+        gni = df.groupby("A", as_index=False)
+        return gni
+
+    def test_describe(self, df, gb, gni):
+        # describe
+        expected_index = Index([1, 3], name="A")
+        expected_col = MultiIndex(
+            levels=[["B"], ["count", "mean", "std", "min", "25%", "50%", "75%", "max"]],
+            codes=[[0] * 8, list(range(8))],
+        )
+        expected = DataFrame(
+            [
+                [1.0, 2.0, np.nan, 2.0, 2.0, 2.0, 2.0, 2.0],
+                [0.0, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan, np.nan],
+            ],
+            index=expected_index,
+            columns=expected_col,
+        )
+        result = gb.describe()
+        tm.assert_frame_equal(result, expected)
+
+        expected = expected.reset_index()
+        result = gni.describe()
+        tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("dtype", [int, float, object])
+@pytest.mark.parametrize(
+    "kwargs",
+    [
+        {"percentiles": [0.10, 0.20, 0.30], "include": "all", "exclude": None},
+        {"percentiles": [0.10, 0.20, 0.30], "include": None, "exclude": ["int"]},
+        {"percentiles": [0.10, 0.20, 0.30], "include": ["int"], "exclude": None},
+    ],
+)
+def test_groupby_empty_dataset(dtype, kwargs):
+    # GH#41575
+    df = DataFrame([[1, 2, 3]], columns=["A", "B", "C"], dtype=dtype)
+    df["B"] = df["B"].astype(int)
+    df["C"] = df["C"].astype(float)
+
+    result = df.iloc[:0].groupby("A").describe(**kwargs)
+    expected = df.groupby("A").describe(**kwargs).reset_index(drop=True).iloc[:0]
+    tm.assert_frame_equal(result, expected)
+
+    result = df.iloc[:0].groupby("A").B.describe(**kwargs)
+    expected = df.groupby("A").B.describe(**kwargs).reset_index(drop=True).iloc[:0]
+    expected.index = Index([])
+    tm.assert_frame_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/methods/test_groupby_shift_diff.py

@@ -0,0 +1,255 @@
+import numpy as np
+import pytest
+
+from pandas import (
+    DataFrame,
+    NaT,
+    Series,
+    Timedelta,
+    Timestamp,
+    date_range,
+)
+import pandas._testing as tm
+
+
+def test_group_shift_with_null_key():
+    # This test is designed to replicate the segfault in issue #13813.
+    n_rows = 1200
+
+    # Generate a moderately large dataframe with occasional missing
+    # values in column `B`, and then group by [`A`, `B`]. This should
+    # force `-1` in `labels` array of `g._grouper.group_info` exactly
+    # at those places, where the group-by key is partially missing.
+    df = DataFrame(
+        [(i % 12, i % 3 if i % 3 else np.nan, i) for i in range(n_rows)],
+        dtype=float,
+        columns=["A", "B", "Z"],
+        index=None,
+    )
+    g = df.groupby(["A", "B"])
+
+    expected = DataFrame(
+        [(i + 12 if i % 3 and i < n_rows - 12 else np.nan) for i in range(n_rows)],
+        dtype=float,
+        columns=["Z"],
+        index=None,
+    )
+    result = g.shift(-1)
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_group_shift_with_fill_value():
+    # GH #24128
+    n_rows = 24
+    df = DataFrame(
+        [(i % 12, i % 3, i) for i in range(n_rows)],
+        dtype=float,
+        columns=["A", "B", "Z"],
+        index=None,
+    )
+    g = df.groupby(["A", "B"])
+
+    expected = DataFrame(
+        [(i + 12 if i < n_rows - 12 else 0) for i in range(n_rows)],
+        dtype=float,
+        columns=["Z"],
+        index=None,
+    )
+    result = g.shift(-1, fill_value=0)
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_group_shift_lose_timezone():
+    # GH 30134
+    now_dt = Timestamp.utcnow().as_unit("ns")
+    df = DataFrame({"a": [1, 1], "date": now_dt})
+    result = df.groupby("a").shift(0).iloc[0]
+    expected = Series({"date": now_dt}, name=result.name)
+    tm.assert_series_equal(result, expected)
+
+
+def test_group_diff_real_series(any_real_numpy_dtype):
+    df = DataFrame(
+        {"a": [1, 2, 3, 3, 2], "b": [1, 2, 3, 4, 5]},
+        dtype=any_real_numpy_dtype,
+    )
+    result = df.groupby("a")["b"].diff()
+    exp_dtype = "float"
+    if any_real_numpy_dtype in ["int8", "int16", "float32"]:
+        exp_dtype = "float32"
+    expected = Series([np.nan, np.nan, np.nan, 1.0, 3.0], dtype=exp_dtype, name="b")
+    tm.assert_series_equal(result, expected)
+
+
+def test_group_diff_real_frame(any_real_numpy_dtype):
+    df = DataFrame(
+        {
+            "a": [1, 2, 3, 3, 2],
+            "b": [1, 2, 3, 4, 5],
+            "c": [1, 2, 3, 4, 6],
+        },
+        dtype=any_real_numpy_dtype,
+    )
+    result = df.groupby("a").diff()
+    exp_dtype = "float"
+    if any_real_numpy_dtype in ["int8", "int16", "float32"]:
+        exp_dtype = "float32"
+    expected = DataFrame(
+        {
+            "b": [np.nan, np.nan, np.nan, 1.0, 3.0],
+            "c": [np.nan, np.nan, np.nan, 1.0, 4.0],
+        },
+        dtype=exp_dtype,
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "data",
+    [
+        [
+            Timestamp("2013-01-01"),
+            Timestamp("2013-01-02"),
+            Timestamp("2013-01-03"),
+        ],
+        [Timedelta("5 days"), Timedelta("6 days"), Timedelta("7 days")],
+    ],
+)
+def test_group_diff_datetimelike(data, unit):
+    df = DataFrame({"a": [1, 2, 2], "b": data})
+    df["b"] = df["b"].dt.as_unit(unit)
+    result = df.groupby("a")["b"].diff()
+    expected = Series([NaT, NaT, Timedelta("1 days")], name="b").dt.as_unit(unit)
+    tm.assert_series_equal(result, expected)
+
+
+def test_group_diff_bool():
+    df = DataFrame({"a": [1, 2, 3, 3, 2], "b": [True, True, False, False, True]})
+    result = df.groupby("a")["b"].diff()
+    expected = Series([np.nan, np.nan, np.nan, False, False], name="b")
+    tm.assert_series_equal(result, expected)
+
+
+def test_group_diff_object_raises(object_dtype):
+    df = DataFrame(
+        {"a": ["foo", "bar", "bar"], "b": ["baz", "foo", "foo"]}, dtype=object_dtype
+    )
+    with pytest.raises(TypeError, match=r"unsupported operand type\(s\) for -"):
+        df.groupby("a")["b"].diff()
+
+
+def test_empty_shift_with_fill():
+    # GH 41264, single-index check
+    df = DataFrame(columns=["a", "b", "c"])
+    shifted = df.groupby(["a"]).shift(1)
+    shifted_with_fill = df.groupby(["a"]).shift(1, fill_value=0)
+    tm.assert_frame_equal(shifted, shifted_with_fill)
+    tm.assert_index_equal(shifted.index, shifted_with_fill.index)
+
+
+def test_multindex_empty_shift_with_fill():
+    # GH 41264, multi-index check
+    df = DataFrame(columns=["a", "b", "c"])
+    shifted = df.groupby(["a", "b"]).shift(1)
+    shifted_with_fill = df.groupby(["a", "b"]).shift(1, fill_value=0)
+    tm.assert_frame_equal(shifted, shifted_with_fill)
+    tm.assert_index_equal(shifted.index, shifted_with_fill.index)
+
+
+def test_shift_periods_freq():
+    # GH 54093
+    data = {"a": [1, 2, 3, 4, 5, 6], "b": [0, 0, 0, 1, 1, 1]}
+    df = DataFrame(data, index=date_range(start="20100101", periods=6))
+    result = df.groupby(df.index).shift(periods=-2, freq="D")
+    expected = DataFrame(data, index=date_range(start="2009-12-30", periods=6))
+    tm.assert_frame_equal(result, expected)
+
+
+def test_shift_deprecate_freq_and_fill_value():
+    # GH 53832
+    data = {"a": [1, 2, 3, 4, 5, 6], "b": [0, 0, 0, 1, 1, 1]}
+    df = DataFrame(data, index=date_range(start="20100101", periods=6))
+    msg = (
+        "Passing a 'freq' together with a 'fill_value' silently ignores the fill_value"
+    )
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        df.groupby(df.index).shift(periods=-2, freq="D", fill_value="1")
+
+
+def test_shift_disallow_suffix_if_periods_is_int():
+    # GH#44424
+    data = {"a": [1, 2, 3, 4, 5, 6], "b": [0, 0, 0, 1, 1, 1]}
+    df = DataFrame(data)
+    msg = "Cannot specify `suffix` if `periods` is an int."
+    with pytest.raises(ValueError, match=msg):
+        df.groupby("b").shift(1, suffix="fails")
+
+
+def test_group_shift_with_multiple_periods():
+    # GH#44424
+    df = DataFrame({"a": [1, 2, 3, 3, 2], "b": [True, True, False, False, True]})
+
+    shifted_df = df.groupby("b")[["a"]].shift([0, 1])
+    expected_df = DataFrame(
+        {"a_0": [1, 2, 3, 3, 2], "a_1": [np.nan, 1.0, np.nan, 3.0, 2.0]}
+    )
+    tm.assert_frame_equal(shifted_df, expected_df)
+
+    # series
+    shifted_series = df.groupby("b")["a"].shift([0, 1])
+    tm.assert_frame_equal(shifted_series, expected_df)
+
+
+def test_group_shift_with_multiple_periods_and_freq():
+    # GH#44424
+    df = DataFrame(
+        {"a": [1, 2, 3, 4, 5], "b": [True, True, False, False, True]},
+        index=date_range("1/1/2000", periods=5, freq="h"),
+    )
+    shifted_df = df.groupby("b")[["a"]].shift(
+        [0, 1],
+        freq="h",
+    )
+    expected_df = DataFrame(
+        {
+            "a_0": [1.0, 2.0, 3.0, 4.0, 5.0, np.nan],
+            "a_1": [
+                np.nan,
+                1.0,
+                2.0,
+                3.0,
+                4.0,
+                5.0,
+            ],
+        },
+        index=date_range("1/1/2000", periods=6, freq="h"),
+    )
+    tm.assert_frame_equal(shifted_df, expected_df)
+
+
+def test_group_shift_with_multiple_periods_and_fill_value():
+    # GH#44424
+    df = DataFrame(
+        {"a": [1, 2, 3, 4, 5], "b": [True, True, False, False, True]},
+    )
+    shifted_df = df.groupby("b")[["a"]].shift([0, 1], fill_value=-1)
+    expected_df = DataFrame(
+        {"a_0": [1, 2, 3, 4, 5], "a_1": [-1, 1, -1, 3, 2]},
+    )
+    tm.assert_frame_equal(shifted_df, expected_df)
+
+
+def test_group_shift_with_multiple_periods_and_both_fill_and_freq_deprecated():
+    # GH#44424
+    df = DataFrame(
+        {"a": [1, 2, 3, 4, 5], "b": [True, True, False, False, True]},
+        index=date_range("1/1/2000", periods=5, freq="h"),
+    )
+    msg = (
+        "Passing a 'freq' together with a 'fill_value' silently ignores the "
+        "fill_value"
+    )
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        df.groupby("b")[["a"]].shift([1, 2], fill_value=1, freq="h")

venv/lib/python3.12/site-packages/pandas/tests/groupby/methods/test_is_monotonic.py

@@ -0,0 +1,78 @@
+import numpy as np
+import pytest
+
+from pandas import (
+    DataFrame,
+    Index,
+    Series,
+)
+import pandas._testing as tm
+
+
+@pytest.mark.parametrize(
+    "in_vals, out_vals",
+    [
+        # Basics: strictly increasing (T), strictly decreasing (F),
+        # abs val increasing (F), non-strictly increasing (T)
+        ([1, 2, 5, 3, 2, 0, 4, 5, -6, 1, 1], [True, False, False, True]),
+        # Test with inf vals
+        (
+            [1, 2.1, np.inf, 3, 2, np.inf, -np.inf, 5, 11, 1, -np.inf],
+            [True, False, True, False],
+        ),
+        # Test with nan vals; should always be False
+        (
+            [1, 2, np.nan, 3, 2, np.nan, np.nan, 5, -np.inf, 1, np.nan],
+            [False, False, False, False],
+        ),
+    ],
+)
+def test_is_monotonic_increasing(in_vals, out_vals):
+    # GH 17015
+    source_dict = {
+        "A": ["1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11"],
+        "B": ["a", "a", "a", "b", "b", "b", "c", "c", "c", "d", "d"],
+        "C": in_vals,
+    }
+    df = DataFrame(source_dict)
+    result = df.groupby("B").C.is_monotonic_increasing
+    index = Index(list("abcd"), name="B")
+    expected = Series(index=index, data=out_vals, name="C")
+    tm.assert_series_equal(result, expected)
+
+    # Also check result equal to manually taking x.is_monotonic_increasing.
+    expected = df.groupby(["B"]).C.apply(lambda x: x.is_monotonic_increasing)
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "in_vals, out_vals",
+    [
+        # Basics: strictly decreasing (T), strictly increasing (F),
+        # abs val decreasing (F), non-strictly increasing (T)
+        ([10, 9, 7, 3, 4, 5, -3, 2, 0, 1, 1], [True, False, False, True]),
+        # Test with inf vals
+        (
+            [np.inf, 1, -np.inf, np.inf, 2, -3, -np.inf, 5, -3, -np.inf, -np.inf],
+            [True, True, False, True],
+        ),
+        # Test with nan vals; should always be False
+        (
+            [1, 2, np.nan, 3, 2, np.nan, np.nan, 5, -np.inf, 1, np.nan],
+            [False, False, False, False],
+        ),
+    ],
+)
+def test_is_monotonic_decreasing(in_vals, out_vals):
+    # GH 17015
+    source_dict = {
+        "A": ["1", "2", "3", "4", "5", "6", "7", "8", "9", "10", "11"],
+        "B": ["a", "a", "a", "b", "b", "b", "c", "c", "c", "d", "d"],
+        "C": in_vals,
+    }
+
+    df = DataFrame(source_dict)
+    result = df.groupby("B").C.is_monotonic_decreasing
+    index = Index(list("abcd"), name="B")
+    expected = Series(index=index, data=out_vals, name="C")
+    tm.assert_series_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/methods/test_nlargest_nsmallest.py

@@ -0,0 +1,115 @@
+import numpy as np
+import pytest
+
+from pandas import (
+    MultiIndex,
+    Series,
+    date_range,
+)
+import pandas._testing as tm
+
+
+def test_nlargest():
+    a = Series([1, 3, 5, 7, 2, 9, 0, 4, 6, 10])
+    b = Series(list("a" * 5 + "b" * 5))
+    gb = a.groupby(b)
+    r = gb.nlargest(3)
+    e = Series(
+        [7, 5, 3, 10, 9, 6],
+        index=MultiIndex.from_arrays([list("aaabbb"), [3, 2, 1, 9, 5, 8]]),
+    )
+    tm.assert_series_equal(r, e)
+
+    a = Series([1, 1, 3, 2, 0, 3, 3, 2, 1, 0])
+    gb = a.groupby(b)
+    e = Series(
+        [3, 2, 1, 3, 3, 2],
+        index=MultiIndex.from_arrays([list("aaabbb"), [2, 3, 1, 6, 5, 7]]),
+    )
+    tm.assert_series_equal(gb.nlargest(3, keep="last"), e)
+
+
+def test_nlargest_mi_grouper():
+    # see gh-21411
+    npr = np.random.default_rng(2)
+
+    dts = date_range("20180101", periods=10)
+    iterables = [dts, ["one", "two"]]
+
+    idx = MultiIndex.from_product(iterables, names=["first", "second"])
+    s = Series(npr.standard_normal(20), index=idx)
+
+    result = s.groupby("first").nlargest(1)
+
+    exp_idx = MultiIndex.from_tuples(
+        [
+            (dts[0], dts[0], "one"),
+            (dts[1], dts[1], "one"),
+            (dts[2], dts[2], "one"),
+            (dts[3], dts[3], "two"),
+            (dts[4], dts[4], "one"),
+            (dts[5], dts[5], "one"),
+            (dts[6], dts[6], "one"),
+            (dts[7], dts[7], "one"),
+            (dts[8], dts[8], "one"),
+            (dts[9], dts[9], "one"),
+        ],
+        names=["first", "first", "second"],
+    )
+
+    exp_values = [
+        0.18905338179353307,
+        -0.41306354339189344,
+        1.799707382720902,
+        0.7738065867276614,
+        0.28121066979764925,
+        0.9775674511260357,
+        -0.3288239040579627,
+        0.45495807124085547,
+        0.5452887139646817,
+        0.12682784711186987,
+    ]
+
+    expected = Series(exp_values, index=exp_idx)
+    tm.assert_series_equal(result, expected, check_exact=False, rtol=1e-3)
+
+
+def test_nsmallest():
+    a = Series([1, 3, 5, 7, 2, 9, 0, 4, 6, 10])
+    b = Series(list("a" * 5 + "b" * 5))
+    gb = a.groupby(b)
+    r = gb.nsmallest(3)
+    e = Series(
+        [1, 2, 3, 0, 4, 6],
+        index=MultiIndex.from_arrays([list("aaabbb"), [0, 4, 1, 6, 7, 8]]),
+    )
+    tm.assert_series_equal(r, e)
+
+    a = Series([1, 1, 3, 2, 0, 3, 3, 2, 1, 0])
+    gb = a.groupby(b)
+    e = Series(
+        [0, 1, 1, 0, 1, 2],
+        index=MultiIndex.from_arrays([list("aaabbb"), [4, 1, 0, 9, 8, 7]]),
+    )
+    tm.assert_series_equal(gb.nsmallest(3, keep="last"), e)
+
+
+@pytest.mark.parametrize(
+    "data, groups",
+    [([0, 1, 2, 3], [0, 0, 1, 1]), ([0], [0])],
+)
+@pytest.mark.parametrize("dtype", [None, *tm.ALL_INT_NUMPY_DTYPES])
+@pytest.mark.parametrize("method", ["nlargest", "nsmallest"])
+def test_nlargest_and_smallest_noop(data, groups, dtype, method):
+    # GH 15272, GH 16345, GH 29129
+    # Test nlargest/smallest when it results in a noop,
+    # i.e. input is sorted and group size <= n
+    if dtype is not None:
+        data = np.array(data, dtype=dtype)
+    if method == "nlargest":
+        data = list(reversed(data))
+    ser = Series(data, name="a")
+    result = getattr(ser.groupby(groups), method)(n=2)
+    expidx = np.array(groups, dtype=int) if isinstance(groups, list) else groups
+    expected = Series(data, index=MultiIndex.from_arrays([expidx, ser.index]), name="a")
+    tm.assert_series_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/methods/test_nth.py

@@ -0,0 +1,921 @@
+import numpy as np
+import pytest
+
+import pandas as pd
+from pandas import (
+    DataFrame,
+    Index,
+    MultiIndex,
+    Series,
+    Timestamp,
+    isna,
+)
+import pandas._testing as tm
+
+
+def test_first_last_nth(df):
+    # tests for first / last / nth
+    grouped = df.groupby("A")
+    first = grouped.first()
+    expected = df.loc[[1, 0], ["B", "C", "D"]]
+    expected.index = Index(["bar", "foo"], name="A")
+    expected = expected.sort_index()
+    tm.assert_frame_equal(first, expected)
+
+    nth = grouped.nth(0)
+    expected = df.loc[[0, 1]]
+    tm.assert_frame_equal(nth, expected)
+
+    last = grouped.last()
+    expected = df.loc[[5, 7], ["B", "C", "D"]]
+    expected.index = Index(["bar", "foo"], name="A")
+    tm.assert_frame_equal(last, expected)
+
+    nth = grouped.nth(-1)
+    expected = df.iloc[[5, 7]]
+    tm.assert_frame_equal(nth, expected)
+
+    nth = grouped.nth(1)
+    expected = df.iloc[[2, 3]]
+    tm.assert_frame_equal(nth, expected)
+
+    # it works!
+    grouped["B"].first()
+    grouped["B"].last()
+    grouped["B"].nth(0)
+
+    df = df.copy()
+    df.loc[df["A"] == "foo", "B"] = np.nan
+    grouped = df.groupby("A")
+    assert isna(grouped["B"].first()["foo"])
+    assert isna(grouped["B"].last()["foo"])
+    assert isna(grouped["B"].nth(0).iloc[0])
+
+    # v0.14.0 whatsnew
+    df = DataFrame([[1, np.nan], [1, 4], [5, 6]], columns=["A", "B"])
+    g = df.groupby("A")
+    result = g.first()
+    expected = df.iloc[[1, 2]].set_index("A")
+    tm.assert_frame_equal(result, expected)
+
+    expected = df.iloc[[1, 2]]
+    result = g.nth(0, dropna="any")
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("method", ["first", "last"])
+def test_first_last_with_na_object(method, nulls_fixture):
+    # https://github.com/pandas-dev/pandas/issues/32123
+    groups = DataFrame({"a": [1, 1, 2, 2], "b": [1, 2, 3, nulls_fixture]}).groupby("a")
+    result = getattr(groups, method)()
+
+    if method == "first":
+        values = [1, 3]
+    else:
+        values = [2, 3]
+
+    values = np.array(values, dtype=result["b"].dtype)
+    idx = Index([1, 2], name="a")
+    expected = DataFrame({"b": values}, index=idx)
+
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("index", [0, -1])
+def test_nth_with_na_object(index, nulls_fixture):
+    # https://github.com/pandas-dev/pandas/issues/32123
+    df = DataFrame({"a": [1, 1, 2, 2], "b": [1, 2, 3, nulls_fixture]})
+    groups = df.groupby("a")
+    result = groups.nth(index)
+    expected = df.iloc[[0, 2]] if index == 0 else df.iloc[[1, 3]]
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("method", ["first", "last"])
+def test_first_last_with_None(method):
+    # https://github.com/pandas-dev/pandas/issues/32800
+    # None should be preserved as object dtype
+    df = DataFrame.from_dict({"id": ["a"], "value": [None]})
+    groups = df.groupby("id", as_index=False)
+    result = getattr(groups, method)()
+
+    tm.assert_frame_equal(result, df)
+
+
+@pytest.mark.parametrize("method", ["first", "last"])
+@pytest.mark.parametrize(
+    "df, expected",
+    [
+        (
+            DataFrame({"id": "a", "value": [None, "foo", np.nan]}),
+            DataFrame({"value": ["foo"]}, index=Index(["a"], name="id")),
+        ),
+        (
+            DataFrame({"id": "a", "value": [np.nan]}, dtype=object),
+            DataFrame({"value": [None]}, index=Index(["a"], name="id")),
+        ),
+    ],
+)
+def test_first_last_with_None_expanded(method, df, expected):
+    # GH 32800, 38286
+    result = getattr(df.groupby("id"), method)()
+    tm.assert_frame_equal(result, expected)
+
+
+def test_first_last_nth_dtypes():
+    df = DataFrame(
+        {
+            "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"],
+            "B": ["one", "one", "two", "three", "two", "two", "one", "three"],
+            "C": np.random.default_rng(2).standard_normal(8),
+            "D": np.array(np.random.default_rng(2).standard_normal(8), dtype="float32"),
+        }
+    )
+    df["E"] = True
+    df["F"] = 1
+
+    # tests for first / last / nth
+    grouped = df.groupby("A")
+    first = grouped.first()
+    expected = df.loc[[1, 0], ["B", "C", "D", "E", "F"]]
+    expected.index = Index(["bar", "foo"], name="A")
+    expected = expected.sort_index()
+    tm.assert_frame_equal(first, expected)
+
+    last = grouped.last()
+    expected = df.loc[[5, 7], ["B", "C", "D", "E", "F"]]
+    expected.index = Index(["bar", "foo"], name="A")
+    expected = expected.sort_index()
+    tm.assert_frame_equal(last, expected)
+
+    nth = grouped.nth(1)
+    expected = df.iloc[[2, 3]]
+    tm.assert_frame_equal(nth, expected)
+
+
+def test_first_last_nth_dtypes2():
+    # GH 2763, first/last shifting dtypes
+    idx = list(range(10))
+    idx.append(9)
+    ser = Series(data=range(11), index=idx, name="IntCol")
+    assert ser.dtype == "int64"
+    f = ser.groupby(level=0).first()
+    assert f.dtype == "int64"
+
+
+def test_first_last_nth_nan_dtype():
+    # GH 33591
+    df = DataFrame({"data": ["A"], "nans": Series([None], dtype=object)})
+    grouped = df.groupby("data")
+
+    expected = df.set_index("data").nans
+    tm.assert_series_equal(grouped.nans.first(), expected)
+    tm.assert_series_equal(grouped.nans.last(), expected)
+
+    expected = df.nans
+    tm.assert_series_equal(grouped.nans.nth(-1), expected)
+    tm.assert_series_equal(grouped.nans.nth(0), expected)
+
+
+def test_first_strings_timestamps():
+    # GH 11244
+    test = DataFrame(
+        {
+            Timestamp("2012-01-01 00:00:00"): ["a", "b"],
+            Timestamp("2012-01-02 00:00:00"): ["c", "d"],
+            "name": ["e", "e"],
+            "aaaa": ["f", "g"],
+        }
+    )
+    result = test.groupby("name").first()
+    expected = DataFrame(
+        [["a", "c", "f"]],
+        columns=Index([Timestamp("2012-01-01"), Timestamp("2012-01-02"), "aaaa"]),
+        index=Index(["e"], name="name"),
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+def test_nth():
+    df = DataFrame([[1, np.nan], [1, 4], [5, 6]], columns=["A", "B"])
+    gb = df.groupby("A")
+
+    tm.assert_frame_equal(gb.nth(0), df.iloc[[0, 2]])
+    tm.assert_frame_equal(gb.nth(1), df.iloc[[1]])
+    tm.assert_frame_equal(gb.nth(2), df.loc[[]])
+    tm.assert_frame_equal(gb.nth(-1), df.iloc[[1, 2]])
+    tm.assert_frame_equal(gb.nth(-2), df.iloc[[0]])
+    tm.assert_frame_equal(gb.nth(-3), df.loc[[]])
+    tm.assert_series_equal(gb.B.nth(0), df.B.iloc[[0, 2]])
+    tm.assert_series_equal(gb.B.nth(1), df.B.iloc[[1]])
+    tm.assert_frame_equal(gb[["B"]].nth(0), df[["B"]].iloc[[0, 2]])
+
+    tm.assert_frame_equal(gb.nth(0, dropna="any"), df.iloc[[1, 2]])
+    tm.assert_frame_equal(gb.nth(-1, dropna="any"), df.iloc[[1, 2]])
+
+    tm.assert_frame_equal(gb.nth(7, dropna="any"), df.iloc[:0])
+    tm.assert_frame_equal(gb.nth(2, dropna="any"), df.iloc[:0])
+
+
+def test_nth2():
+    # out of bounds, regression from 0.13.1
+    # GH 6621
+    df = DataFrame(
+        {
+            "color": {0: "green", 1: "green", 2: "red", 3: "red", 4: "red"},
+            "food": {0: "ham", 1: "eggs", 2: "eggs", 3: "ham", 4: "pork"},
+            "two": {
+                0: 1.5456590000000001,
+                1: -0.070345000000000005,
+                2: -2.4004539999999999,
+                3: 0.46206000000000003,
+                4: 0.52350799999999997,
+            },
+            "one": {
+                0: 0.56573799999999996,
+                1: -0.9742360000000001,
+                2: 1.033801,
+                3: -0.78543499999999999,
+                4: 0.70422799999999997,
+            },
+        }
+    ).set_index(["color", "food"])
+
+    result = df.groupby(level=0, as_index=False).nth(2)
+    expected = df.iloc[[-1]]
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby(level=0, as_index=False).nth(3)
+    expected = df.loc[[]]
+    tm.assert_frame_equal(result, expected)
+
+
+def test_nth3():
+    # GH 7559
+    # from the vbench
+    df = DataFrame(np.random.default_rng(2).integers(1, 10, (100, 2)), dtype="int64")
+    ser = df[1]
+    gb = df[0]
+    expected = ser.groupby(gb).first()
+    expected2 = ser.groupby(gb).apply(lambda x: x.iloc[0])
+    tm.assert_series_equal(expected2, expected, check_names=False)
+    assert expected.name == 1
+    assert expected2.name == 1
+
+    # validate first
+    v = ser[gb == 1].iloc[0]
+    assert expected.iloc[0] == v
+    assert expected2.iloc[0] == v
+
+    with pytest.raises(ValueError, match="For a DataFrame"):
+        ser.groupby(gb, sort=False).nth(0, dropna=True)
+
+
+def test_nth4():
+    # doc example
+    df = DataFrame([[1, np.nan], [1, 4], [5, 6]], columns=["A", "B"])
+    gb = df.groupby("A")
+    result = gb.B.nth(0, dropna="all")
+    expected = df.B.iloc[[1, 2]]
+    tm.assert_series_equal(result, expected)
+
+
+def test_nth5():
+    # test multiple nth values
+    df = DataFrame([[1, np.nan], [1, 3], [1, 4], [5, 6], [5, 7]], columns=["A", "B"])
+    gb = df.groupby("A")
+
+    tm.assert_frame_equal(gb.nth(0), df.iloc[[0, 3]])
+    tm.assert_frame_equal(gb.nth([0]), df.iloc[[0, 3]])
+    tm.assert_frame_equal(gb.nth([0, 1]), df.iloc[[0, 1, 3, 4]])
+    tm.assert_frame_equal(gb.nth([0, -1]), df.iloc[[0, 2, 3, 4]])
+    tm.assert_frame_equal(gb.nth([0, 1, 2]), df.iloc[[0, 1, 2, 3, 4]])
+    tm.assert_frame_equal(gb.nth([0, 1, -1]), df.iloc[[0, 1, 2, 3, 4]])
+    tm.assert_frame_equal(gb.nth([2]), df.iloc[[2]])
+    tm.assert_frame_equal(gb.nth([3, 4]), df.loc[[]])
+
+
+def test_nth_bdays(unit):
+    business_dates = pd.date_range(
+        start="4/1/2014", end="6/30/2014", freq="B", unit=unit
+    )
+    df = DataFrame(1, index=business_dates, columns=["a", "b"])
+    # get the first, fourth and last two business days for each month
+    key = [df.index.year, df.index.month]
+    result = df.groupby(key, as_index=False).nth([0, 3, -2, -1])
+    expected_dates = pd.to_datetime(
+        [
+            "2014/4/1",
+            "2014/4/4",
+            "2014/4/29",
+            "2014/4/30",
+            "2014/5/1",
+            "2014/5/6",
+            "2014/5/29",
+            "2014/5/30",
+            "2014/6/2",
+            "2014/6/5",
+            "2014/6/27",
+            "2014/6/30",
+        ]
+    ).as_unit(unit)
+    expected = DataFrame(1, columns=["a", "b"], index=expected_dates)
+    tm.assert_frame_equal(result, expected)
+
+
+def test_nth_multi_grouper(three_group):
+    # PR 9090, related to issue 8979
+    # test nth on multiple groupers
+    grouped = three_group.groupby(["A", "B"])
+    result = grouped.nth(0)
+    expected = three_group.iloc[[0, 3, 4, 7]]
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "data, expected_first, expected_last",
+    [
+        (
+            {
+                "id": ["A"],
+                "time": Timestamp("2012-02-01 14:00:00", tz="US/Central"),
+                "foo": [1],
+            },
+            {
+                "id": ["A"],
+                "time": Timestamp("2012-02-01 14:00:00", tz="US/Central"),
+                "foo": [1],
+            },
+            {
+                "id": ["A"],
+                "time": Timestamp("2012-02-01 14:00:00", tz="US/Central"),
+                "foo": [1],
+            },
+        ),
+        (
+            {
+                "id": ["A", "B", "A"],
+                "time": [
+                    Timestamp("2012-01-01 13:00:00", tz="America/New_York"),
+                    Timestamp("2012-02-01 14:00:00", tz="US/Central"),
+                    Timestamp("2012-03-01 12:00:00", tz="Europe/London"),
+                ],
+                "foo": [1, 2, 3],
+            },
+            {
+                "id": ["A", "B"],
+                "time": [
+                    Timestamp("2012-01-01 13:00:00", tz="America/New_York"),
+                    Timestamp("2012-02-01 14:00:00", tz="US/Central"),
+                ],
+                "foo": [1, 2],
+            },
+            {
+                "id": ["A", "B"],
+                "time": [
+                    Timestamp("2012-03-01 12:00:00", tz="Europe/London"),
+                    Timestamp("2012-02-01 14:00:00", tz="US/Central"),
+                ],
+                "foo": [3, 2],
+            },
+        ),
+    ],
+)
+def test_first_last_tz(data, expected_first, expected_last):
+    # GH15884
+    # Test that the timezone is retained when calling first
+    # or last on groupby with as_index=False
+
+    df = DataFrame(data)
+
+    result = df.groupby("id", as_index=False).first()
+    expected = DataFrame(expected_first)
+    cols = ["id", "time", "foo"]
+    tm.assert_frame_equal(result[cols], expected[cols])
+
+    result = df.groupby("id", as_index=False)["time"].first()
+    tm.assert_frame_equal(result, expected[["id", "time"]])
+
+    result = df.groupby("id", as_index=False).last()
+    expected = DataFrame(expected_last)
+    cols = ["id", "time", "foo"]
+    tm.assert_frame_equal(result[cols], expected[cols])
+
+    result = df.groupby("id", as_index=False)["time"].last()
+    tm.assert_frame_equal(result, expected[["id", "time"]])
+
+
+@pytest.mark.parametrize(
+    "method, ts, alpha",
+    [
+        ["first", Timestamp("2013-01-01", tz="US/Eastern"), "a"],
+        ["last", Timestamp("2013-01-02", tz="US/Eastern"), "b"],
+    ],
+)
+def test_first_last_tz_multi_column(method, ts, alpha, unit):
+    # GH 21603
+    category_string = Series(list("abc")).astype("category")
+    dti = pd.date_range("20130101", periods=3, tz="US/Eastern", unit=unit)
+    df = DataFrame(
+        {
+            "group": [1, 1, 2],
+            "category_string": category_string,
+            "datetimetz": dti,
+        }
+    )
+    result = getattr(df.groupby("group"), method)()
+    expected = DataFrame(
+        {
+            "category_string": pd.Categorical(
+                [alpha, "c"], dtype=category_string.dtype
+            ),
+            "datetimetz": [ts, Timestamp("2013-01-03", tz="US/Eastern")],
+        },
+        index=Index([1, 2], name="group"),
+    )
+    expected["datetimetz"] = expected["datetimetz"].dt.as_unit(unit)
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "values",
+    [
+        pd.array([True, False], dtype="boolean"),
+        pd.array([1, 2], dtype="Int64"),
+        pd.to_datetime(["2020-01-01", "2020-02-01"]),
+        pd.to_timedelta([1, 2], unit="D"),
+    ],
+)
+@pytest.mark.parametrize("function", ["first", "last", "min", "max"])
+def test_first_last_extension_array_keeps_dtype(values, function):
+    # https://github.com/pandas-dev/pandas/issues/33071
+    # https://github.com/pandas-dev/pandas/issues/32194
+    df = DataFrame({"a": [1, 2], "b": values})
+    grouped = df.groupby("a")
+    idx = Index([1, 2], name="a")
+    expected_series = Series(values, name="b", index=idx)
+    expected_frame = DataFrame({"b": values}, index=idx)
+
+    result_series = getattr(grouped["b"], function)()
+    tm.assert_series_equal(result_series, expected_series)
+
+    result_frame = grouped.agg({"b": function})
+    tm.assert_frame_equal(result_frame, expected_frame)
+
+
+def test_nth_multi_index_as_expected():
+    # PR 9090, related to issue 8979
+    # test nth on MultiIndex
+    three_group = DataFrame(
+        {
+            "A": [
+                "foo",
+                "foo",
+                "foo",
+                "foo",
+                "bar",
+                "bar",
+                "bar",
+                "bar",
+                "foo",
+                "foo",
+                "foo",
+            ],
+            "B": [
+                "one",
+                "one",
+                "one",
+                "two",
+                "one",
+                "one",
+                "one",
+                "two",
+                "two",
+                "two",
+                "one",
+            ],
+            "C": [
+                "dull",
+                "dull",
+                "shiny",
+                "dull",
+                "dull",
+                "shiny",
+                "shiny",
+                "dull",
+                "shiny",
+                "shiny",
+                "shiny",
+            ],
+        }
+    )
+    grouped = three_group.groupby(["A", "B"])
+    result = grouped.nth(0)
+    expected = three_group.iloc[[0, 3, 4, 7]]
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "op, n, expected_rows",
+    [
+        ("head", -1, [0]),
+        ("head", 0, []),
+        ("head", 1, [0, 2]),
+        ("head", 7, [0, 1, 2]),
+        ("tail", -1, [1]),
+        ("tail", 0, []),
+        ("tail", 1, [1, 2]),
+        ("tail", 7, [0, 1, 2]),
+    ],
+)
+@pytest.mark.parametrize("columns", [None, [], ["A"], ["B"], ["A", "B"]])
+@pytest.mark.parametrize("as_index", [True, False])
+def test_groupby_head_tail(op, n, expected_rows, columns, as_index):
+    df = DataFrame([[1, 2], [1, 4], [5, 6]], columns=["A", "B"])
+    g = df.groupby("A", as_index=as_index)
+    expected = df.iloc[expected_rows]
+    if columns is not None:
+        g = g[columns]
+        expected = expected[columns]
+    result = getattr(g, op)(n)
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "op, n, expected_cols",
+    [
+        ("head", -1, [0]),
+        ("head", 0, []),
+        ("head", 1, [0, 2]),
+        ("head", 7, [0, 1, 2]),
+        ("tail", -1, [1]),
+        ("tail", 0, []),
+        ("tail", 1, [1, 2]),
+        ("tail", 7, [0, 1, 2]),
+    ],
+)
+def test_groupby_head_tail_axis_1(op, n, expected_cols):
+    # GH 9772
+    df = DataFrame(
+        [[1, 2, 3], [1, 4, 5], [2, 6, 7], [3, 8, 9]], columns=["A", "B", "C"]
+    )
+    msg = "DataFrame.groupby with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        g = df.groupby([0, 0, 1], axis=1)
+    expected = df.iloc[:, expected_cols]
+    result = getattr(g, op)(n)
+    tm.assert_frame_equal(result, expected)
+
+
+def test_group_selection_cache():
+    # GH 12839 nth, head, and tail should return same result consistently
+    df = DataFrame([[1, 2], [1, 4], [5, 6]], columns=["A", "B"])
+    expected = df.iloc[[0, 2]]
+
+    g = df.groupby("A")
+    result1 = g.head(n=2)
+    result2 = g.nth(0)
+    tm.assert_frame_equal(result1, df)
+    tm.assert_frame_equal(result2, expected)
+
+    g = df.groupby("A")
+    result1 = g.tail(n=2)
+    result2 = g.nth(0)
+    tm.assert_frame_equal(result1, df)
+    tm.assert_frame_equal(result2, expected)
+
+    g = df.groupby("A")
+    result1 = g.nth(0)
+    result2 = g.head(n=2)
+    tm.assert_frame_equal(result1, expected)
+    tm.assert_frame_equal(result2, df)
+
+    g = df.groupby("A")
+    result1 = g.nth(0)
+    result2 = g.tail(n=2)
+    tm.assert_frame_equal(result1, expected)
+    tm.assert_frame_equal(result2, df)
+
+
+def test_nth_empty():
+    # GH 16064
+    df = DataFrame(index=[0], columns=["a", "b", "c"])
+    result = df.groupby("a").nth(10)
+    expected = df.iloc[:0]
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby(["a", "b"]).nth(10)
+    expected = df.iloc[:0]
+    tm.assert_frame_equal(result, expected)
+
+
+def test_nth_column_order():
+    # GH 20760
+    # Check that nth preserves column order
+    df = DataFrame(
+        [[1, "b", 100], [1, "a", 50], [1, "a", np.nan], [2, "c", 200], [2, "d", 150]],
+        columns=["A", "C", "B"],
+    )
+    result = df.groupby("A").nth(0)
+    expected = df.iloc[[0, 3]]
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby("A").nth(-1, dropna="any")
+    expected = df.iloc[[1, 4]]
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("dropna", [None, "any", "all"])
+def test_nth_nan_in_grouper(dropna):
+    # GH 26011
+    df = DataFrame(
+        {
+            "a": [np.nan, "a", np.nan, "b", np.nan],
+            "b": [0, 2, 4, 6, 8],
+            "c": [1, 3, 5, 7, 9],
+        }
+    )
+    result = df.groupby("a").nth(0, dropna=dropna)
+    expected = df.iloc[[1, 3]]
+
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("dropna", [None, "any", "all"])
+def test_nth_nan_in_grouper_series(dropna):
+    # GH 26454
+    df = DataFrame(
+        {
+            "a": [np.nan, "a", np.nan, "b", np.nan],
+            "b": [0, 2, 4, 6, 8],
+        }
+    )
+    result = df.groupby("a")["b"].nth(0, dropna=dropna)
+    expected = df["b"].iloc[[1, 3]]
+
+    tm.assert_series_equal(result, expected)
+
+
+def test_first_categorical_and_datetime_data_nat():
+    # GH 20520
+    df = DataFrame(
+        {
+            "group": ["first", "first", "second", "third", "third"],
+            "time": 5 * [np.datetime64("NaT")],
+            "categories": Series(["a", "b", "c", "a", "b"], dtype="category"),
+        }
+    )
+    result = df.groupby("group").first()
+    expected = DataFrame(
+        {
+            "time": 3 * [np.datetime64("NaT")],
+            "categories": Series(["a", "c", "a"]).astype(
+                pd.CategoricalDtype(["a", "b", "c"])
+            ),
+        }
+    )
+    expected.index = Index(["first", "second", "third"], name="group")
+    tm.assert_frame_equal(result, expected)
+
+
+def test_first_multi_key_groupby_categorical():
+    # GH 22512
+    df = DataFrame(
+        {
+            "A": [1, 1, 1, 2, 2],
+            "B": [100, 100, 200, 100, 100],
+            "C": ["apple", "orange", "mango", "mango", "orange"],
+            "D": ["jupiter", "mercury", "mars", "venus", "venus"],
+        }
+    )
+    df = df.astype({"D": "category"})
+    result = df.groupby(by=["A", "B"]).first()
+    expected = DataFrame(
+        {
+            "C": ["apple", "mango", "mango"],
+            "D": Series(["jupiter", "mars", "venus"]).astype(
+                pd.CategoricalDtype(["jupiter", "mars", "mercury", "venus"])
+            ),
+        }
+    )
+    expected.index = MultiIndex.from_tuples(
+        [(1, 100), (1, 200), (2, 100)], names=["A", "B"]
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("method", ["first", "last", "nth"])
+def test_groupby_last_first_nth_with_none(method, nulls_fixture):
+    # GH29645
+    expected = Series(["y"])
+    data = Series(
+        [nulls_fixture, nulls_fixture, nulls_fixture, "y", nulls_fixture],
+        index=[0, 0, 0, 0, 0],
+    ).groupby(level=0)
+
+    if method == "nth":
+        result = getattr(data, method)(3)
+    else:
+        result = getattr(data, method)()
+
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "arg, expected_rows",
+    [
+        [slice(None, 3, 2), [0, 1, 4, 5]],
+        [slice(None, -2), [0, 2, 5]],
+        [[slice(None, 2), slice(-2, None)], [0, 1, 2, 3, 4, 6, 7]],
+        [[0, 1, slice(-2, None)], [0, 1, 2, 3, 4, 6, 7]],
+    ],
+)
+def test_slice(slice_test_df, slice_test_grouped, arg, expected_rows):
+    # Test slices     GH #42947
+
+    result = slice_test_grouped.nth[arg]
+    equivalent = slice_test_grouped.nth(arg)
+    expected = slice_test_df.iloc[expected_rows]
+
+    tm.assert_frame_equal(result, expected)
+    tm.assert_frame_equal(equivalent, expected)
+
+
+def test_nth_indexed(slice_test_df, slice_test_grouped):
+    # Test index notation     GH #44688
+
+    result = slice_test_grouped.nth[0, 1, -2:]
+    equivalent = slice_test_grouped.nth([0, 1, slice(-2, None)])
+    expected = slice_test_df.iloc[[0, 1, 2, 3, 4, 6, 7]]
+
+    tm.assert_frame_equal(result, expected)
+    tm.assert_frame_equal(equivalent, expected)
+
+
+def test_invalid_argument(slice_test_grouped):
+    # Test for error on invalid argument
+
+    with pytest.raises(TypeError, match="Invalid index"):
+        slice_test_grouped.nth(3.14)
+
+
+def test_negative_step(slice_test_grouped):
+    # Test for error on negative slice step
+
+    with pytest.raises(ValueError, match="Invalid step"):
+        slice_test_grouped.nth(slice(None, None, -1))
+
+
+def test_np_ints(slice_test_df, slice_test_grouped):
+    # Test np ints work
+
+    result = slice_test_grouped.nth(np.array([0, 1]))
+    expected = slice_test_df.iloc[[0, 1, 2, 3, 4]]
+    tm.assert_frame_equal(result, expected)
+
+
+def test_groupby_nth_with_column_axis():
+    # GH43926
+    df = DataFrame(
+        [
+            [4, 5, 6],
+            [8, 8, 7],
+        ],
+        index=["z", "y"],
+        columns=["C", "B", "A"],
+    )
+    msg = "DataFrame.groupby with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        gb = df.groupby(df.iloc[1], axis=1)
+    result = gb.nth(0)
+    expected = df.iloc[:, [0, 2]]
+    tm.assert_frame_equal(result, expected)
+
+
+def test_groupby_nth_interval():
+    # GH#24205
+    idx_result = MultiIndex(
+        [
+            pd.CategoricalIndex([pd.Interval(0, 1), pd.Interval(1, 2)]),
+            pd.CategoricalIndex([pd.Interval(0, 10), pd.Interval(10, 20)]),
+        ],
+        [[0, 0, 0, 1, 1], [0, 1, 1, 0, -1]],
+    )
+    df_result = DataFrame({"col": range(len(idx_result))}, index=idx_result)
+    result = df_result.groupby(level=[0, 1], observed=False).nth(0)
+    val_expected = [0, 1, 3]
+    idx_expected = MultiIndex(
+        [
+            pd.CategoricalIndex([pd.Interval(0, 1), pd.Interval(1, 2)]),
+            pd.CategoricalIndex([pd.Interval(0, 10), pd.Interval(10, 20)]),
+        ],
+        [[0, 0, 1], [0, 1, 0]],
+    )
+    expected = DataFrame(val_expected, index=idx_expected, columns=["col"])
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "start, stop, expected_values, expected_columns",
+    [
+        (None, None, [0, 1, 2, 3, 4], list("ABCDE")),
+        (None, 1, [0, 3], list("AD")),
+        (None, 9, [0, 1, 2, 3, 4], list("ABCDE")),
+        (None, -1, [0, 1, 3], list("ABD")),
+        (1, None, [1, 2, 4], list("BCE")),
+        (1, -1, [1], list("B")),
+        (-1, None, [2, 4], list("CE")),
+        (-1, 2, [4], list("E")),
+    ],
+)
+@pytest.mark.parametrize("method", ["call", "index"])
+def test_nth_slices_with_column_axis(
+    start, stop, expected_values, expected_columns, method
+):
+    df = DataFrame([range(5)], columns=[list("ABCDE")])
+    msg = "DataFrame.groupby with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        gb = df.groupby([5, 5, 5, 6, 6], axis=1)
+    result = {
+        "call": lambda start, stop: gb.nth(slice(start, stop)),
+        "index": lambda start, stop: gb.nth[start:stop],
+    }[method](start, stop)
+    expected = DataFrame([expected_values], columns=[expected_columns])
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.filterwarnings(
+    "ignore:invalid value encountered in remainder:RuntimeWarning"
+)
+def test_head_tail_dropna_true():
+    # GH#45089
+    df = DataFrame(
+        [["a", "z"], ["b", np.nan], ["c", np.nan], ["c", np.nan]], columns=["X", "Y"]
+    )
+    expected = DataFrame([["a", "z"]], columns=["X", "Y"])
+
+    result = df.groupby(["X", "Y"]).head(n=1)
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby(["X", "Y"]).tail(n=1)
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby(["X", "Y"]).nth(n=0)
+    tm.assert_frame_equal(result, expected)
+
+
+def test_head_tail_dropna_false():
+    # GH#45089
+    df = DataFrame([["a", "z"], ["b", np.nan], ["c", np.nan]], columns=["X", "Y"])
+    expected = DataFrame([["a", "z"], ["b", np.nan], ["c", np.nan]], columns=["X", "Y"])
+
+    result = df.groupby(["X", "Y"], dropna=False).head(n=1)
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby(["X", "Y"], dropna=False).tail(n=1)
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby(["X", "Y"], dropna=False).nth(n=0)
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("selection", ("b", ["b"], ["b", "c"]))
+@pytest.mark.parametrize("dropna", ["any", "all", None])
+def test_nth_after_selection(selection, dropna):
+    # GH#11038, GH#53518
+    df = DataFrame(
+        {
+            "a": [1, 1, 2],
+            "b": [np.nan, 3, 4],
+            "c": [5, 6, 7],
+        }
+    )
+    gb = df.groupby("a")[selection]
+    result = gb.nth(0, dropna=dropna)
+    if dropna == "any" or (dropna == "all" and selection != ["b", "c"]):
+        locs = [1, 2]
+    else:
+        locs = [0, 2]
+    expected = df.loc[locs, selection]
+    tm.assert_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "data",
+    [
+        (
+            Timestamp("2011-01-15 12:50:28.502376"),
+            Timestamp("2011-01-20 12:50:28.593448"),
+        ),
+        (24650000000000001, 24650000000000002),
+    ],
+)
+def test_groupby_nth_int_like_precision(data):
+    # GH#6620, GH#9311
+    df = DataFrame({"a": [1, 1], "b": data})
+
+    grouped = df.groupby("a")
+    result = grouped.nth(0)
+    expected = DataFrame({"a": 1, "b": [data[0]]})
+
+    tm.assert_frame_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/methods/test_quantile.py

@@ -0,0 +1,496 @@
+import numpy as np
+import pytest
+
+import pandas as pd
+from pandas import (
+    DataFrame,
+    Index,
+)
+import pandas._testing as tm
+
+
+@pytest.mark.parametrize(
+    "interpolation", ["linear", "lower", "higher", "nearest", "midpoint"]
+)
+@pytest.mark.parametrize(
+    "a_vals,b_vals",
+    [
+        # Ints
+        ([1, 2, 3, 4, 5], [5, 4, 3, 2, 1]),
+        ([1, 2, 3, 4], [4, 3, 2, 1]),
+        ([1, 2, 3, 4, 5], [4, 3, 2, 1]),
+        # Floats
+        ([1.0, 2.0, 3.0, 4.0, 5.0], [5.0, 4.0, 3.0, 2.0, 1.0]),
+        # Missing data
+        ([1.0, np.nan, 3.0, np.nan, 5.0], [5.0, np.nan, 3.0, np.nan, 1.0]),
+        ([np.nan, 4.0, np.nan, 2.0, np.nan], [np.nan, 4.0, np.nan, 2.0, np.nan]),
+        # Timestamps
+        (
+            pd.date_range("1/1/18", freq="D", periods=5),
+            pd.date_range("1/1/18", freq="D", periods=5)[::-1],
+        ),
+        (
+            pd.date_range("1/1/18", freq="D", periods=5).as_unit("s"),
+            pd.date_range("1/1/18", freq="D", periods=5)[::-1].as_unit("s"),
+        ),
+        # All NA
+        ([np.nan] * 5, [np.nan] * 5),
+    ],
+)
+@pytest.mark.parametrize("q", [0, 0.25, 0.5, 0.75, 1])
+def test_quantile(interpolation, a_vals, b_vals, q, request):
+    if (
+        interpolation == "nearest"
+        and q == 0.5
+        and isinstance(b_vals, list)
+        and b_vals == [4, 3, 2, 1]
+    ):
+        request.applymarker(
+            pytest.mark.xfail(
+                reason="Unclear numpy expectation for nearest "
+                "result with equidistant data"
+            )
+        )
+    all_vals = pd.concat([pd.Series(a_vals), pd.Series(b_vals)])
+
+    a_expected = pd.Series(a_vals).quantile(q, interpolation=interpolation)
+    b_expected = pd.Series(b_vals).quantile(q, interpolation=interpolation)
+
+    df = DataFrame({"key": ["a"] * len(a_vals) + ["b"] * len(b_vals), "val": all_vals})
+
+    expected = DataFrame(
+        [a_expected, b_expected], columns=["val"], index=Index(["a", "b"], name="key")
+    )
+    if all_vals.dtype.kind == "M" and expected.dtypes.values[0].kind == "M":
+        # TODO(non-nano): this should be unnecessary once array_to_datetime
+        #  correctly infers non-nano from Timestamp.unit
+        expected = expected.astype(all_vals.dtype)
+    result = df.groupby("key").quantile(q, interpolation=interpolation)
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_quantile_array():
+    # https://github.com/pandas-dev/pandas/issues/27526
+    df = DataFrame({"A": [0, 1, 2, 3, 4]})
+    key = np.array([0, 0, 1, 1, 1], dtype=np.int64)
+    result = df.groupby(key).quantile([0.25])
+
+    index = pd.MultiIndex.from_product([[0, 1], [0.25]])
+    expected = DataFrame({"A": [0.25, 2.50]}, index=index)
+    tm.assert_frame_equal(result, expected)
+
+    df = DataFrame({"A": [0, 1, 2, 3], "B": [4, 5, 6, 7]})
+    index = pd.MultiIndex.from_product([[0, 1], [0.25, 0.75]])
+
+    key = np.array([0, 0, 1, 1], dtype=np.int64)
+    result = df.groupby(key).quantile([0.25, 0.75])
+    expected = DataFrame(
+        {"A": [0.25, 0.75, 2.25, 2.75], "B": [4.25, 4.75, 6.25, 6.75]}, index=index
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+def test_quantile_array2():
+    # https://github.com/pandas-dev/pandas/pull/28085#issuecomment-524066959
+    arr = np.random.default_rng(2).integers(0, 5, size=(10, 3), dtype=np.int64)
+    df = DataFrame(arr, columns=list("ABC"))
+    result = df.groupby("A").quantile([0.3, 0.7])
+    expected = DataFrame(
+        {
+            "B": [2.0, 2.0, 2.3, 2.7, 0.3, 0.7, 3.2, 4.0, 0.3, 0.7],
+            "C": [1.0, 1.0, 1.9, 3.0999999999999996, 0.3, 0.7, 2.6, 3.0, 1.2, 2.8],
+        },
+        index=pd.MultiIndex.from_product(
+            [[0, 1, 2, 3, 4], [0.3, 0.7]], names=["A", None]
+        ),
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+def test_quantile_array_no_sort():
+    df = DataFrame({"A": [0, 1, 2], "B": [3, 4, 5]})
+    key = np.array([1, 0, 1], dtype=np.int64)
+    result = df.groupby(key, sort=False).quantile([0.25, 0.5, 0.75])
+    expected = DataFrame(
+        {"A": [0.5, 1.0, 1.5, 1.0, 1.0, 1.0], "B": [3.5, 4.0, 4.5, 4.0, 4.0, 4.0]},
+        index=pd.MultiIndex.from_product([[1, 0], [0.25, 0.5, 0.75]]),
+    )
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby(key, sort=False).quantile([0.75, 0.25])
+    expected = DataFrame(
+        {"A": [1.5, 0.5, 1.0, 1.0], "B": [4.5, 3.5, 4.0, 4.0]},
+        index=pd.MultiIndex.from_product([[1, 0], [0.75, 0.25]]),
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+def test_quantile_array_multiple_levels():
+    df = DataFrame(
+        {"A": [0, 1, 2], "B": [3, 4, 5], "c": ["a", "a", "a"], "d": ["a", "a", "b"]}
+    )
+    result = df.groupby(["c", "d"]).quantile([0.25, 0.75])
+    index = pd.MultiIndex.from_tuples(
+        [("a", "a", 0.25), ("a", "a", 0.75), ("a", "b", 0.25), ("a", "b", 0.75)],
+        names=["c", "d", None],
+    )
+    expected = DataFrame(
+        {"A": [0.25, 0.75, 2.0, 2.0], "B": [3.25, 3.75, 5.0, 5.0]}, index=index
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("frame_size", [(2, 3), (100, 10)])
+@pytest.mark.parametrize("groupby", [[0], [0, 1]])
+@pytest.mark.parametrize("q", [[0.5, 0.6]])
+def test_groupby_quantile_with_arraylike_q_and_int_columns(frame_size, groupby, q):
+    # GH30289
+    nrow, ncol = frame_size
+    df = DataFrame(np.array([ncol * [_ % 4] for _ in range(nrow)]), columns=range(ncol))
+
+    idx_levels = [np.arange(min(nrow, 4))] * len(groupby) + [q]
+    idx_codes = [[x for x in range(min(nrow, 4)) for _ in q]] * len(groupby) + [
+        list(range(len(q))) * min(nrow, 4)
+    ]
+    expected_index = pd.MultiIndex(
+        levels=idx_levels, codes=idx_codes, names=groupby + [None]
+    )
+    expected_values = [
+        [float(x)] * (ncol - len(groupby)) for x in range(min(nrow, 4)) for _ in q
+    ]
+    expected_columns = [x for x in range(ncol) if x not in groupby]
+    expected = DataFrame(
+        expected_values, index=expected_index, columns=expected_columns
+    )
+    result = df.groupby(groupby).quantile(q)
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_quantile_raises():
+    df = DataFrame([["foo", "a"], ["foo", "b"], ["foo", "c"]], columns=["key", "val"])
+
+    with pytest.raises(TypeError, match="cannot be performed against 'object' dtypes"):
+        df.groupby("key").quantile()
+
+
+def test_quantile_out_of_bounds_q_raises():
+    # https://github.com/pandas-dev/pandas/issues/27470
+    df = DataFrame({"a": [0, 0, 0, 1, 1, 1], "b": range(6)})
+    g = df.groupby([0, 0, 0, 1, 1, 1])
+    with pytest.raises(ValueError, match="Got '50.0' instead"):
+        g.quantile(50)
+
+    with pytest.raises(ValueError, match="Got '-1.0' instead"):
+        g.quantile(-1)
+
+
+def test_quantile_missing_group_values_no_segfaults():
+    # GH 28662
+    data = np.array([1.0, np.nan, 1.0])
+    df = DataFrame({"key": data, "val": range(3)})
+
+    # Random segfaults; would have been guaranteed in loop
+    grp = df.groupby("key")
+    for _ in range(100):
+        grp.quantile()
+
+
+@pytest.mark.parametrize(
+    "key, val, expected_key, expected_val",
+    [
+        ([1.0, np.nan, 3.0, np.nan], range(4), [1.0, 3.0], [0.0, 2.0]),
+        ([1.0, np.nan, 2.0, 2.0], range(4), [1.0, 2.0], [0.0, 2.5]),
+        (["a", "b", "b", np.nan], range(4), ["a", "b"], [0, 1.5]),
+        ([0], [42], [0], [42.0]),
+        ([], [], np.array([], dtype="float64"), np.array([], dtype="float64")),
+    ],
+)
+def test_quantile_missing_group_values_correct_results(
+    key, val, expected_key, expected_val
+):
+    # GH 28662, GH 33200, GH 33569
+    df = DataFrame({"key": key, "val": val})
+
+    expected = DataFrame(
+        expected_val, index=Index(expected_key, name="key"), columns=["val"]
+    )
+
+    grp = df.groupby("key")
+
+    result = grp.quantile(0.5)
+    tm.assert_frame_equal(result, expected)
+
+    result = grp.quantile()
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "values",
+    [
+        pd.array([1, 0, None] * 2, dtype="Int64"),
+        pd.array([True, False, None] * 2, dtype="boolean"),
+    ],
+)
+@pytest.mark.parametrize("q", [0.5, [0.0, 0.5, 1.0]])
+def test_groupby_quantile_nullable_array(values, q):
+    # https://github.com/pandas-dev/pandas/issues/33136
+    df = DataFrame({"a": ["x"] * 3 + ["y"] * 3, "b": values})
+    result = df.groupby("a")["b"].quantile(q)
+
+    if isinstance(q, list):
+        idx = pd.MultiIndex.from_product((["x", "y"], q), names=["a", None])
+        true_quantiles = [0.0, 0.5, 1.0]
+    else:
+        idx = Index(["x", "y"], name="a")
+        true_quantiles = [0.5]
+
+    expected = pd.Series(true_quantiles * 2, index=idx, name="b", dtype="Float64")
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize("q", [0.5, [0.0, 0.5, 1.0]])
+@pytest.mark.parametrize("numeric_only", [True, False])
+def test_groupby_quantile_raises_on_invalid_dtype(q, numeric_only):
+    df = DataFrame({"a": [1], "b": [2.0], "c": ["x"]})
+    if numeric_only:
+        result = df.groupby("a").quantile(q, numeric_only=numeric_only)
+        expected = df.groupby("a")[["b"]].quantile(q)
+        tm.assert_frame_equal(result, expected)
+    else:
+        with pytest.raises(
+            TypeError, match="'quantile' cannot be performed against 'object' dtypes!"
+        ):
+            df.groupby("a").quantile(q, numeric_only=numeric_only)
+
+
+def test_groupby_quantile_NA_float(any_float_dtype):
+    # GH#42849
+    df = DataFrame({"x": [1, 1], "y": [0.2, np.nan]}, dtype=any_float_dtype)
+    result = df.groupby("x")["y"].quantile(0.5)
+    exp_index = Index([1.0], dtype=any_float_dtype, name="x")
+
+    if any_float_dtype in ["Float32", "Float64"]:
+        expected_dtype = any_float_dtype
+    else:
+        expected_dtype = None
+
+    expected = pd.Series([0.2], dtype=expected_dtype, index=exp_index, name="y")
+    tm.assert_series_equal(result, expected)
+
+    result = df.groupby("x")["y"].quantile([0.5, 0.75])
+    expected = pd.Series(
+        [0.2] * 2,
+        index=pd.MultiIndex.from_product((exp_index, [0.5, 0.75]), names=["x", None]),
+        name="y",
+        dtype=expected_dtype,
+    )
+    tm.assert_series_equal(result, expected)
+
+
+def test_groupby_quantile_NA_int(any_int_ea_dtype):
+    # GH#42849
+    df = DataFrame({"x": [1, 1], "y": [2, 5]}, dtype=any_int_ea_dtype)
+    result = df.groupby("x")["y"].quantile(0.5)
+    expected = pd.Series(
+        [3.5],
+        dtype="Float64",
+        index=Index([1], name="x", dtype=any_int_ea_dtype),
+        name="y",
+    )
+    tm.assert_series_equal(expected, result)
+
+    result = df.groupby("x").quantile(0.5)
+    expected = DataFrame(
+        {"y": 3.5}, dtype="Float64", index=Index([1], name="x", dtype=any_int_ea_dtype)
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "interpolation, val1, val2", [("lower", 2, 2), ("higher", 2, 3), ("nearest", 2, 2)]
+)
+def test_groupby_quantile_all_na_group_masked(
+    interpolation, val1, val2, any_numeric_ea_dtype
+):
+    # GH#37493
+    df = DataFrame(
+        {"a": [1, 1, 1, 2], "b": [1, 2, 3, pd.NA]}, dtype=any_numeric_ea_dtype
+    )
+    result = df.groupby("a").quantile(q=[0.5, 0.7], interpolation=interpolation)
+    expected = DataFrame(
+        {"b": [val1, val2, pd.NA, pd.NA]},
+        dtype=any_numeric_ea_dtype,
+        index=pd.MultiIndex.from_arrays(
+            [pd.Series([1, 1, 2, 2], dtype=any_numeric_ea_dtype), [0.5, 0.7, 0.5, 0.7]],
+            names=["a", None],
+        ),
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("interpolation", ["midpoint", "linear"])
+def test_groupby_quantile_all_na_group_masked_interp(
+    interpolation, any_numeric_ea_dtype
+):
+    # GH#37493
+    df = DataFrame(
+        {"a": [1, 1, 1, 2], "b": [1, 2, 3, pd.NA]}, dtype=any_numeric_ea_dtype
+    )
+    result = df.groupby("a").quantile(q=[0.5, 0.75], interpolation=interpolation)
+
+    if any_numeric_ea_dtype == "Float32":
+        expected_dtype = any_numeric_ea_dtype
+    else:
+        expected_dtype = "Float64"
+
+    expected = DataFrame(
+        {"b": [2.0, 2.5, pd.NA, pd.NA]},
+        dtype=expected_dtype,
+        index=pd.MultiIndex.from_arrays(
+            [
+                pd.Series([1, 1, 2, 2], dtype=any_numeric_ea_dtype),
+                [0.5, 0.75, 0.5, 0.75],
+            ],
+            names=["a", None],
+        ),
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("dtype", ["Float64", "Float32"])
+def test_groupby_quantile_allNA_column(dtype):
+    # GH#42849
+    df = DataFrame({"x": [1, 1], "y": [pd.NA] * 2}, dtype=dtype)
+    result = df.groupby("x")["y"].quantile(0.5)
+    expected = pd.Series(
+        [np.nan], dtype=dtype, index=Index([1.0], dtype=dtype), name="y"
+    )
+    expected.index.name = "x"
+    tm.assert_series_equal(expected, result)
+
+
+def test_groupby_timedelta_quantile():
+    # GH: 29485
+    df = DataFrame(
+        {"value": pd.to_timedelta(np.arange(4), unit="s"), "group": [1, 1, 2, 2]}
+    )
+    result = df.groupby("group").quantile(0.99)
+    expected = DataFrame(
+        {
+            "value": [
+                pd.Timedelta("0 days 00:00:00.990000"),
+                pd.Timedelta("0 days 00:00:02.990000"),
+            ]
+        },
+        index=Index([1, 2], name="group"),
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+def test_columns_groupby_quantile():
+    # GH 33795
+    df = DataFrame(
+        np.arange(12).reshape(3, -1),
+        index=list("XYZ"),
+        columns=pd.Series(list("ABAB"), name="col"),
+    )
+    msg = "DataFrame.groupby with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        gb = df.groupby("col", axis=1)
+    result = gb.quantile(q=[0.8, 0.2])
+    expected = DataFrame(
+        [
+            [1.6, 0.4, 2.6, 1.4],
+            [5.6, 4.4, 6.6, 5.4],
+            [9.6, 8.4, 10.6, 9.4],
+        ],
+        index=list("XYZ"),
+        columns=pd.MultiIndex.from_tuples(
+            [("A", 0.8), ("A", 0.2), ("B", 0.8), ("B", 0.2)], names=["col", None]
+        ),
+    )
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_timestamp_groupby_quantile(unit):
+    # GH 33168
+    dti = pd.date_range(
+        start="2020-04-19 00:00:00", freq="1min", periods=100, tz="UTC", unit=unit
+    ).floor("1h")
+    df = DataFrame(
+        {
+            "timestamp": dti,
+            "category": list(range(1, 101)),
+            "value": list(range(101, 201)),
+        }
+    )
+
+    result = df.groupby("timestamp").quantile([0.2, 0.8])
+
+    mi = pd.MultiIndex.from_product([dti[::99], [0.2, 0.8]], names=("timestamp", None))
+    expected = DataFrame(
+        [
+            {"category": 12.8, "value": 112.8},
+            {"category": 48.2, "value": 148.2},
+            {"category": 68.8, "value": 168.8},
+            {"category": 92.2, "value": 192.2},
+        ],
+        index=mi,
+    )
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_groupby_quantile_dt64tz_period():
+    # GH#51373
+    dti = pd.date_range("2016-01-01", periods=1000)
+    df = pd.Series(dti).to_frame().copy()
+    df[1] = dti.tz_localize("US/Pacific")
+    df[2] = dti.to_period("D")
+    df[3] = dti - dti[0]
+    df.iloc[-1] = pd.NaT
+
+    by = np.tile(np.arange(5), 200)
+    gb = df.groupby(by)
+
+    result = gb.quantile(0.5)
+
+    # Check that we match the group-by-group result
+    exp = {i: df.iloc[i::5].quantile(0.5) for i in range(5)}
+    expected = DataFrame(exp).T.infer_objects()
+    expected.index = expected.index.astype(int)
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_groupby_quantile_nonmulti_levels_order():
+    # Non-regression test for GH #53009
+    ind = pd.MultiIndex.from_tuples(
+        [
+            (0, "a", "B"),
+            (0, "a", "A"),
+            (0, "b", "B"),
+            (0, "b", "A"),
+            (1, "a", "B"),
+            (1, "a", "A"),
+            (1, "b", "B"),
+            (1, "b", "A"),
+        ],
+        names=["sample", "cat0", "cat1"],
+    )
+    ser = pd.Series(range(8), index=ind)
+    result = ser.groupby(level="cat1", sort=False).quantile([0.2, 0.8])
+
+    qind = pd.MultiIndex.from_tuples(
+        [("B", 0.2), ("B", 0.8), ("A", 0.2), ("A", 0.8)], names=["cat1", None]
+    )
+    expected = pd.Series([1.2, 4.8, 2.2, 5.8], index=qind)
+
+    tm.assert_series_equal(result, expected)
+
+    # We need to check that index levels are not sorted
+    expected_levels = pd.core.indexes.frozen.FrozenList([["B", "A"], [0.2, 0.8]])
+    tm.assert_equal(result.index.levels, expected_levels)

venv/lib/python3.12/site-packages/pandas/tests/groupby/methods/test_rank.py

@@ -0,0 +1,721 @@
+from datetime import datetime
+
+import numpy as np
+import pytest
+
+import pandas as pd
+from pandas import (
+    DataFrame,
+    NaT,
+    Series,
+    concat,
+)
+import pandas._testing as tm
+
+
+def test_rank_unordered_categorical_typeerror():
+    # GH#51034 should be TypeError, not NotImplementedError
+    cat = pd.Categorical([], ordered=False)
+    ser = Series(cat)
+    df = ser.to_frame()
+
+    msg = "Cannot perform rank with non-ordered Categorical"
+
+    gb = ser.groupby(cat, observed=False)
+    with pytest.raises(TypeError, match=msg):
+        gb.rank()
+
+    gb2 = df.groupby(cat, observed=False)
+    with pytest.raises(TypeError, match=msg):
+        gb2.rank()
+
+
+def test_rank_apply():
+    lev1 = np.array(["a" * 10] * 100, dtype=object)
+    lev2 = np.array(["b" * 10] * 130, dtype=object)
+    lab1 = np.random.default_rng(2).integers(0, 100, size=500, dtype=int)
+    lab2 = np.random.default_rng(2).integers(0, 130, size=500, dtype=int)
+
+    df = DataFrame(
+        {
+            "value": np.random.default_rng(2).standard_normal(500),
+            "key1": lev1.take(lab1),
+            "key2": lev2.take(lab2),
+        }
+    )
+
+    result = df.groupby(["key1", "key2"]).value.rank()
+
+    expected = [piece.value.rank() for key, piece in df.groupby(["key1", "key2"])]
+    expected = concat(expected, axis=0)
+    expected = expected.reindex(result.index)
+    tm.assert_series_equal(result, expected)
+
+    result = df.groupby(["key1", "key2"]).value.rank(pct=True)
+
+    expected = [
+        piece.value.rank(pct=True) for key, piece in df.groupby(["key1", "key2"])
+    ]
+    expected = concat(expected, axis=0)
+    expected = expected.reindex(result.index)
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize("grps", [["qux"], ["qux", "quux"]])
+@pytest.mark.parametrize(
+    "vals",
+    [
+        np.array([2, 2, 8, 2, 6], dtype=dtype)
+        for dtype in ["i8", "i4", "i2", "i1", "u8", "u4", "u2", "u1", "f8", "f4", "f2"]
+    ]
+    + [
+        [
+            pd.Timestamp("2018-01-02"),
+            pd.Timestamp("2018-01-02"),
+            pd.Timestamp("2018-01-08"),
+            pd.Timestamp("2018-01-02"),
+            pd.Timestamp("2018-01-06"),
+        ],
+        [
+            pd.Timestamp("2018-01-02", tz="US/Pacific"),
+            pd.Timestamp("2018-01-02", tz="US/Pacific"),
+            pd.Timestamp("2018-01-08", tz="US/Pacific"),
+            pd.Timestamp("2018-01-02", tz="US/Pacific"),
+            pd.Timestamp("2018-01-06", tz="US/Pacific"),
+        ],
+        [
+            pd.Timestamp("2018-01-02") - pd.Timestamp(0),
+            pd.Timestamp("2018-01-02") - pd.Timestamp(0),
+            pd.Timestamp("2018-01-08") - pd.Timestamp(0),
+            pd.Timestamp("2018-01-02") - pd.Timestamp(0),
+            pd.Timestamp("2018-01-06") - pd.Timestamp(0),
+        ],
+        [
+            pd.Timestamp("2018-01-02").to_period("D"),
+            pd.Timestamp("2018-01-02").to_period("D"),
+            pd.Timestamp("2018-01-08").to_period("D"),
+            pd.Timestamp("2018-01-02").to_period("D"),
+            pd.Timestamp("2018-01-06").to_period("D"),
+        ],
+    ],
+    ids=lambda x: type(x[0]),
+)
+@pytest.mark.parametrize(
+    "ties_method,ascending,pct,exp",
+    [
+        ("average", True, False, [2.0, 2.0, 5.0, 2.0, 4.0]),
+        ("average", True, True, [0.4, 0.4, 1.0, 0.4, 0.8]),
+        ("average", False, False, [4.0, 4.0, 1.0, 4.0, 2.0]),
+        ("average", False, True, [0.8, 0.8, 0.2, 0.8, 0.4]),
+        ("min", True, False, [1.0, 1.0, 5.0, 1.0, 4.0]),
+        ("min", True, True, [0.2, 0.2, 1.0, 0.2, 0.8]),
+        ("min", False, False, [3.0, 3.0, 1.0, 3.0, 2.0]),
+        ("min", False, True, [0.6, 0.6, 0.2, 0.6, 0.4]),
+        ("max", True, False, [3.0, 3.0, 5.0, 3.0, 4.0]),
+        ("max", True, True, [0.6, 0.6, 1.0, 0.6, 0.8]),
+        ("max", False, False, [5.0, 5.0, 1.0, 5.0, 2.0]),
+        ("max", False, True, [1.0, 1.0, 0.2, 1.0, 0.4]),
+        ("first", True, False, [1.0, 2.0, 5.0, 3.0, 4.0]),
+        ("first", True, True, [0.2, 0.4, 1.0, 0.6, 0.8]),
+        ("first", False, False, [3.0, 4.0, 1.0, 5.0, 2.0]),
+        ("first", False, True, [0.6, 0.8, 0.2, 1.0, 0.4]),
+        ("dense", True, False, [1.0, 1.0, 3.0, 1.0, 2.0]),
+        ("dense", True, True, [1.0 / 3.0, 1.0 / 3.0, 3.0 / 3.0, 1.0 / 3.0, 2.0 / 3.0]),
+        ("dense", False, False, [3.0, 3.0, 1.0, 3.0, 2.0]),
+        ("dense", False, True, [3.0 / 3.0, 3.0 / 3.0, 1.0 / 3.0, 3.0 / 3.0, 2.0 / 3.0]),
+    ],
+)
+def test_rank_args(grps, vals, ties_method, ascending, pct, exp):
+    key = np.repeat(grps, len(vals))
+
+    orig_vals = vals
+    vals = list(vals) * len(grps)
+    if isinstance(orig_vals, np.ndarray):
+        vals = np.array(vals, dtype=orig_vals.dtype)
+
+    df = DataFrame({"key": key, "val": vals})
+    result = df.groupby("key").rank(method=ties_method, ascending=ascending, pct=pct)
+
+    exp_df = DataFrame(exp * len(grps), columns=["val"])
+    tm.assert_frame_equal(result, exp_df)
+
+
+@pytest.mark.parametrize("grps", [["qux"], ["qux", "quux"]])
+@pytest.mark.parametrize(
+    "vals", [[-np.inf, -np.inf, np.nan, 1.0, np.nan, np.inf, np.inf]]
+)
+@pytest.mark.parametrize(
+    "ties_method,ascending,na_option,exp",
+    [
+        ("average", True, "keep", [1.5, 1.5, np.nan, 3, np.nan, 4.5, 4.5]),
+        ("average", True, "top", [3.5, 3.5, 1.5, 5.0, 1.5, 6.5, 6.5]),
+        ("average", True, "bottom", [1.5, 1.5, 6.5, 3.0, 6.5, 4.5, 4.5]),
+        ("average", False, "keep", [4.5, 4.5, np.nan, 3, np.nan, 1.5, 1.5]),
+        ("average", False, "top", [6.5, 6.5, 1.5, 5.0, 1.5, 3.5, 3.5]),
+        ("average", False, "bottom", [4.5, 4.5, 6.5, 3.0, 6.5, 1.5, 1.5]),
+        ("min", True, "keep", [1.0, 1.0, np.nan, 3.0, np.nan, 4.0, 4.0]),
+        ("min", True, "top", [3.0, 3.0, 1.0, 5.0, 1.0, 6.0, 6.0]),
+        ("min", True, "bottom", [1.0, 1.0, 6.0, 3.0, 6.0, 4.0, 4.0]),
+        ("min", False, "keep", [4.0, 4.0, np.nan, 3.0, np.nan, 1.0, 1.0]),
+        ("min", False, "top", [6.0, 6.0, 1.0, 5.0, 1.0, 3.0, 3.0]),
+        ("min", False, "bottom", [4.0, 4.0, 6.0, 3.0, 6.0, 1.0, 1.0]),
+        ("max", True, "keep", [2.0, 2.0, np.nan, 3.0, np.nan, 5.0, 5.0]),
+        ("max", True, "top", [4.0, 4.0, 2.0, 5.0, 2.0, 7.0, 7.0]),
+        ("max", True, "bottom", [2.0, 2.0, 7.0, 3.0, 7.0, 5.0, 5.0]),
+        ("max", False, "keep", [5.0, 5.0, np.nan, 3.0, np.nan, 2.0, 2.0]),
+        ("max", False, "top", [7.0, 7.0, 2.0, 5.0, 2.0, 4.0, 4.0]),
+        ("max", False, "bottom", [5.0, 5.0, 7.0, 3.0, 7.0, 2.0, 2.0]),
+        ("first", True, "keep", [1.0, 2.0, np.nan, 3.0, np.nan, 4.0, 5.0]),
+        ("first", True, "top", [3.0, 4.0, 1.0, 5.0, 2.0, 6.0, 7.0]),
+        ("first", True, "bottom", [1.0, 2.0, 6.0, 3.0, 7.0, 4.0, 5.0]),
+        ("first", False, "keep", [4.0, 5.0, np.nan, 3.0, np.nan, 1.0, 2.0]),
+        ("first", False, "top", [6.0, 7.0, 1.0, 5.0, 2.0, 3.0, 4.0]),
+        ("first", False, "bottom", [4.0, 5.0, 6.0, 3.0, 7.0, 1.0, 2.0]),
+        ("dense", True, "keep", [1.0, 1.0, np.nan, 2.0, np.nan, 3.0, 3.0]),
+        ("dense", True, "top", [2.0, 2.0, 1.0, 3.0, 1.0, 4.0, 4.0]),
+        ("dense", True, "bottom", [1.0, 1.0, 4.0, 2.0, 4.0, 3.0, 3.0]),
+        ("dense", False, "keep", [3.0, 3.0, np.nan, 2.0, np.nan, 1.0, 1.0]),
+        ("dense", False, "top", [4.0, 4.0, 1.0, 3.0, 1.0, 2.0, 2.0]),
+        ("dense", False, "bottom", [3.0, 3.0, 4.0, 2.0, 4.0, 1.0, 1.0]),
+    ],
+)
+def test_infs_n_nans(grps, vals, ties_method, ascending, na_option, exp):
+    # GH 20561
+    key = np.repeat(grps, len(vals))
+    vals = vals * len(grps)
+    df = DataFrame({"key": key, "val": vals})
+    result = df.groupby("key").rank(
+        method=ties_method, ascending=ascending, na_option=na_option
+    )
+    exp_df = DataFrame(exp * len(grps), columns=["val"])
+    tm.assert_frame_equal(result, exp_df)
+
+
+@pytest.mark.parametrize("grps", [["qux"], ["qux", "quux"]])
+@pytest.mark.parametrize(
+    "vals",
+    [
+        np.array([2, 2, np.nan, 8, 2, 6, np.nan, np.nan], dtype=dtype)
+        for dtype in ["f8", "f4", "f2"]
+    ]
+    + [
+        [
+            pd.Timestamp("2018-01-02"),
+            pd.Timestamp("2018-01-02"),
+            np.nan,
+            pd.Timestamp("2018-01-08"),
+            pd.Timestamp("2018-01-02"),
+            pd.Timestamp("2018-01-06"),
+            np.nan,
+            np.nan,
+        ],
+        [
+            pd.Timestamp("2018-01-02", tz="US/Pacific"),
+            pd.Timestamp("2018-01-02", tz="US/Pacific"),
+            np.nan,
+            pd.Timestamp("2018-01-08", tz="US/Pacific"),
+            pd.Timestamp("2018-01-02", tz="US/Pacific"),
+            pd.Timestamp("2018-01-06", tz="US/Pacific"),
+            np.nan,
+            np.nan,
+        ],
+        [
+            pd.Timestamp("2018-01-02") - pd.Timestamp(0),
+            pd.Timestamp("2018-01-02") - pd.Timestamp(0),
+            np.nan,
+            pd.Timestamp("2018-01-08") - pd.Timestamp(0),
+            pd.Timestamp("2018-01-02") - pd.Timestamp(0),
+            pd.Timestamp("2018-01-06") - pd.Timestamp(0),
+            np.nan,
+            np.nan,
+        ],
+        [
+            pd.Timestamp("2018-01-02").to_period("D"),
+            pd.Timestamp("2018-01-02").to_period("D"),
+            np.nan,
+            pd.Timestamp("2018-01-08").to_period("D"),
+            pd.Timestamp("2018-01-02").to_period("D"),
+            pd.Timestamp("2018-01-06").to_period("D"),
+            np.nan,
+            np.nan,
+        ],
+    ],
+    ids=lambda x: type(x[0]),
+)
+@pytest.mark.parametrize(
+    "ties_method,ascending,na_option,pct,exp",
+    [
+        (
+            "average",
+            True,
+            "keep",
+            False,
+            [2.0, 2.0, np.nan, 5.0, 2.0, 4.0, np.nan, np.nan],
+        ),
+        (
+            "average",
+            True,
+            "keep",
+            True,
+            [0.4, 0.4, np.nan, 1.0, 0.4, 0.8, np.nan, np.nan],
+        ),
+        (
+            "average",
+            False,
+            "keep",
+            False,
+            [4.0, 4.0, np.nan, 1.0, 4.0, 2.0, np.nan, np.nan],
+        ),
+        (
+            "average",
+            False,
+            "keep",
+            True,
+            [0.8, 0.8, np.nan, 0.2, 0.8, 0.4, np.nan, np.nan],
+        ),
+        ("min", True, "keep", False, [1.0, 1.0, np.nan, 5.0, 1.0, 4.0, np.nan, np.nan]),
+        ("min", True, "keep", True, [0.2, 0.2, np.nan, 1.0, 0.2, 0.8, np.nan, np.nan]),
+        (
+            "min",
+            False,
+            "keep",
+            False,
+            [3.0, 3.0, np.nan, 1.0, 3.0, 2.0, np.nan, np.nan],
+        ),
+        ("min", False, "keep", True, [0.6, 0.6, np.nan, 0.2, 0.6, 0.4, np.nan, np.nan]),
+        ("max", True, "keep", False, [3.0, 3.0, np.nan, 5.0, 3.0, 4.0, np.nan, np.nan]),
+        ("max", True, "keep", True, [0.6, 0.6, np.nan, 1.0, 0.6, 0.8, np.nan, np.nan]),
+        (
+            "max",
+            False,
+            "keep",
+            False,
+            [5.0, 5.0, np.nan, 1.0, 5.0, 2.0, np.nan, np.nan],
+        ),
+        ("max", False, "keep", True, [1.0, 1.0, np.nan, 0.2, 1.0, 0.4, np.nan, np.nan]),
+        (
+            "first",
+            True,
+            "keep",
+            False,
+            [1.0, 2.0, np.nan, 5.0, 3.0, 4.0, np.nan, np.nan],
+        ),
+        (
+            "first",
+            True,
+            "keep",
+            True,
+            [0.2, 0.4, np.nan, 1.0, 0.6, 0.8, np.nan, np.nan],
+        ),
+        (
+            "first",
+            False,
+            "keep",
+            False,
+            [3.0, 4.0, np.nan, 1.0, 5.0, 2.0, np.nan, np.nan],
+        ),
+        (
+            "first",
+            False,
+            "keep",
+            True,
+            [0.6, 0.8, np.nan, 0.2, 1.0, 0.4, np.nan, np.nan],
+        ),
+        (
+            "dense",
+            True,
+            "keep",
+            False,
+            [1.0, 1.0, np.nan, 3.0, 1.0, 2.0, np.nan, np.nan],
+        ),
+        (
+            "dense",
+            True,
+            "keep",
+            True,
+            [
+                1.0 / 3.0,
+                1.0 / 3.0,
+                np.nan,
+                3.0 / 3.0,
+                1.0 / 3.0,
+                2.0 / 3.0,
+                np.nan,
+                np.nan,
+            ],
+        ),
+        (
+            "dense",
+            False,
+            "keep",
+            False,
+            [3.0, 3.0, np.nan, 1.0, 3.0, 2.0, np.nan, np.nan],
+        ),
+        (
+            "dense",
+            False,
+            "keep",
+            True,
+            [
+                3.0 / 3.0,
+                3.0 / 3.0,
+                np.nan,
+                1.0 / 3.0,
+                3.0 / 3.0,
+                2.0 / 3.0,
+                np.nan,
+                np.nan,
+            ],
+        ),
+        ("average", True, "bottom", False, [2.0, 2.0, 7.0, 5.0, 2.0, 4.0, 7.0, 7.0]),
+        (
+            "average",
+            True,
+            "bottom",
+            True,
+            [0.25, 0.25, 0.875, 0.625, 0.25, 0.5, 0.875, 0.875],
+        ),
+        ("average", False, "bottom", False, [4.0, 4.0, 7.0, 1.0, 4.0, 2.0, 7.0, 7.0]),
+        (
+            "average",
+            False,
+            "bottom",
+            True,
+            [0.5, 0.5, 0.875, 0.125, 0.5, 0.25, 0.875, 0.875],
+        ),
+        ("min", True, "bottom", False, [1.0, 1.0, 6.0, 5.0, 1.0, 4.0, 6.0, 6.0]),
+        (
+            "min",
+            True,
+            "bottom",
+            True,
+            [0.125, 0.125, 0.75, 0.625, 0.125, 0.5, 0.75, 0.75],
+        ),
+        ("min", False, "bottom", False, [3.0, 3.0, 6.0, 1.0, 3.0, 2.0, 6.0, 6.0]),
+        (
+            "min",
+            False,
+            "bottom",
+            True,
+            [0.375, 0.375, 0.75, 0.125, 0.375, 0.25, 0.75, 0.75],
+        ),
+        ("max", True, "bottom", False, [3.0, 3.0, 8.0, 5.0, 3.0, 4.0, 8.0, 8.0]),
+        ("max", True, "bottom", True, [0.375, 0.375, 1.0, 0.625, 0.375, 0.5, 1.0, 1.0]),
+        ("max", False, "bottom", False, [5.0, 5.0, 8.0, 1.0, 5.0, 2.0, 8.0, 8.0]),
+        (
+            "max",
+            False,
+            "bottom",
+            True,
+            [0.625, 0.625, 1.0, 0.125, 0.625, 0.25, 1.0, 1.0],
+        ),
+        ("first", True, "bottom", False, [1.0, 2.0, 6.0, 5.0, 3.0, 4.0, 7.0, 8.0]),
+        (
+            "first",
+            True,
+            "bottom",
+            True,
+            [0.125, 0.25, 0.75, 0.625, 0.375, 0.5, 0.875, 1.0],
+        ),
+        ("first", False, "bottom", False, [3.0, 4.0, 6.0, 1.0, 5.0, 2.0, 7.0, 8.0]),
+        (
+            "first",
+            False,
+            "bottom",
+            True,
+            [0.375, 0.5, 0.75, 0.125, 0.625, 0.25, 0.875, 1.0],
+        ),
+        ("dense", True, "bottom", False, [1.0, 1.0, 4.0, 3.0, 1.0, 2.0, 4.0, 4.0]),
+        ("dense", True, "bottom", True, [0.25, 0.25, 1.0, 0.75, 0.25, 0.5, 1.0, 1.0]),
+        ("dense", False, "bottom", False, [3.0, 3.0, 4.0, 1.0, 3.0, 2.0, 4.0, 4.0]),
+        ("dense", False, "bottom", True, [0.75, 0.75, 1.0, 0.25, 0.75, 0.5, 1.0, 1.0]),
+    ],
+)
+def test_rank_args_missing(grps, vals, ties_method, ascending, na_option, pct, exp):
+    key = np.repeat(grps, len(vals))
+
+    orig_vals = vals
+    vals = list(vals) * len(grps)
+    if isinstance(orig_vals, np.ndarray):
+        vals = np.array(vals, dtype=orig_vals.dtype)
+
+    df = DataFrame({"key": key, "val": vals})
+    result = df.groupby("key").rank(
+        method=ties_method, ascending=ascending, na_option=na_option, pct=pct
+    )
+
+    exp_df = DataFrame(exp * len(grps), columns=["val"])
+    tm.assert_frame_equal(result, exp_df)
+
+
+@pytest.mark.parametrize(
+    "pct,exp", [(False, [3.0, 3.0, 3.0, 3.0, 3.0]), (True, [0.6, 0.6, 0.6, 0.6, 0.6])]
+)
+def test_rank_resets_each_group(pct, exp):
+    df = DataFrame(
+        {"key": ["a", "a", "a", "a", "a", "b", "b", "b", "b", "b"], "val": [1] * 10}
+    )
+    result = df.groupby("key").rank(pct=pct)
+    exp_df = DataFrame(exp * 2, columns=["val"])
+    tm.assert_frame_equal(result, exp_df)
+
+
+@pytest.mark.parametrize(
+    "dtype", ["int64", "int32", "uint64", "uint32", "float64", "float32"]
+)
+@pytest.mark.parametrize("upper", [True, False])
+def test_rank_avg_even_vals(dtype, upper):
+    if upper:
+        # use IntegerDtype/FloatingDtype
+        dtype = dtype[0].upper() + dtype[1:]
+        dtype = dtype.replace("Ui", "UI")
+    df = DataFrame({"key": ["a"] * 4, "val": [1] * 4})
+    df["val"] = df["val"].astype(dtype)
+    assert df["val"].dtype == dtype
+
+    result = df.groupby("key").rank()
+    exp_df = DataFrame([2.5, 2.5, 2.5, 2.5], columns=["val"])
+    if upper:
+        exp_df = exp_df.astype("Float64")
+    tm.assert_frame_equal(result, exp_df)
+
+
+@pytest.mark.parametrize("ties_method", ["average", "min", "max", "first", "dense"])
+@pytest.mark.parametrize("ascending", [True, False])
+@pytest.mark.parametrize("na_option", ["keep", "top", "bottom"])
+@pytest.mark.parametrize("pct", [True, False])
+@pytest.mark.parametrize(
+    "vals", [["bar", "bar", "foo", "bar", "baz"], ["bar", np.nan, "foo", np.nan, "baz"]]
+)
+def test_rank_object_dtype(ties_method, ascending, na_option, pct, vals):
+    df = DataFrame({"key": ["foo"] * 5, "val": vals})
+    mask = df["val"].isna()
+
+    gb = df.groupby("key")
+    res = gb.rank(method=ties_method, ascending=ascending, na_option=na_option, pct=pct)
+
+    # construct our expected by using numeric values with the same ordering
+    if mask.any():
+        df2 = DataFrame({"key": ["foo"] * 5, "val": [0, np.nan, 2, np.nan, 1]})
+    else:
+        df2 = DataFrame({"key": ["foo"] * 5, "val": [0, 0, 2, 0, 1]})
+
+    gb2 = df2.groupby("key")
+    alt = gb2.rank(
+        method=ties_method, ascending=ascending, na_option=na_option, pct=pct
+    )
+
+    tm.assert_frame_equal(res, alt)
+
+
+@pytest.mark.parametrize("na_option", [True, "bad", 1])
+@pytest.mark.parametrize("ties_method", ["average", "min", "max", "first", "dense"])
+@pytest.mark.parametrize("ascending", [True, False])
+@pytest.mark.parametrize("pct", [True, False])
+@pytest.mark.parametrize(
+    "vals",
+    [
+        ["bar", "bar", "foo", "bar", "baz"],
+        ["bar", np.nan, "foo", np.nan, "baz"],
+        [1, np.nan, 2, np.nan, 3],
+    ],
+)
+def test_rank_naoption_raises(ties_method, ascending, na_option, pct, vals):
+    df = DataFrame({"key": ["foo"] * 5, "val": vals})
+    msg = "na_option must be one of 'keep', 'top', or 'bottom'"
+
+    with pytest.raises(ValueError, match=msg):
+        df.groupby("key").rank(
+            method=ties_method, ascending=ascending, na_option=na_option, pct=pct
+        )
+
+
+def test_rank_empty_group():
+    # see gh-22519
+    column = "A"
+    df = DataFrame({"A": [0, 1, 0], "B": [1.0, np.nan, 2.0]})
+
+    result = df.groupby(column).B.rank(pct=True)
+    expected = Series([0.5, np.nan, 1.0], name="B")
+    tm.assert_series_equal(result, expected)
+
+    result = df.groupby(column).rank(pct=True)
+    expected = DataFrame({"B": [0.5, np.nan, 1.0]})
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "input_key,input_value,output_value",
+    [
+        ([1, 2], [1, 1], [1.0, 1.0]),
+        ([1, 1, 2, 2], [1, 2, 1, 2], [0.5, 1.0, 0.5, 1.0]),
+        ([1, 1, 2, 2], [1, 2, 1, np.nan], [0.5, 1.0, 1.0, np.nan]),
+        ([1, 1, 2], [1, 2, np.nan], [0.5, 1.0, np.nan]),
+    ],
+)
+def test_rank_zero_div(input_key, input_value, output_value):
+    # GH 23666
+    df = DataFrame({"A": input_key, "B": input_value})
+
+    result = df.groupby("A").rank(method="dense", pct=True)
+    expected = DataFrame({"B": output_value})
+    tm.assert_frame_equal(result, expected)
+
+
+def test_rank_min_int():
+    # GH-32859
+    df = DataFrame(
+        {
+            "grp": [1, 1, 2],
+            "int_col": [
+                np.iinfo(np.int64).min,
+                np.iinfo(np.int64).max,
+                np.iinfo(np.int64).min,
+            ],
+            "datetimelike": [NaT, datetime(2001, 1, 1), NaT],
+        }
+    )
+
+    result = df.groupby("grp").rank()
+    expected = DataFrame(
+        {"int_col": [1.0, 2.0, 1.0], "datetimelike": [np.nan, 1.0, np.nan]}
+    )
+
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("use_nan", [True, False])
+def test_rank_pct_equal_values_on_group_transition(use_nan):
+    # GH#40518
+    fill_value = np.nan if use_nan else 3
+    df = DataFrame(
+        [
+            [-1, 1],
+            [-1, 2],
+            [1, fill_value],
+            [-1, fill_value],
+        ],
+        columns=["group", "val"],
+    )
+    result = df.groupby(["group"])["val"].rank(
+        method="dense",
+        pct=True,
+    )
+    if use_nan:
+        expected = Series([0.5, 1, np.nan, np.nan], name="val")
+    else:
+        expected = Series([1 / 3, 2 / 3, 1, 1], name="val")
+
+    tm.assert_series_equal(result, expected)
+
+
+def test_rank_multiindex():
+    # GH27721
+    df = concat(
+        {
+            "a": DataFrame({"col1": [3, 4], "col2": [1, 2]}),
+            "b": DataFrame({"col3": [5, 6], "col4": [7, 8]}),
+        },
+        axis=1,
+    )
+
+    msg = "DataFrame.groupby with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        gb = df.groupby(level=0, axis=1)
+    msg = "DataFrameGroupBy.rank with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        result = gb.rank(axis=1)
+
+    expected = concat(
+        [
+            df["a"].rank(axis=1),
+            df["b"].rank(axis=1),
+        ],
+        axis=1,
+        keys=["a", "b"],
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+def test_groupby_axis0_rank_axis1():
+    # GH#41320
+    df = DataFrame(
+        {0: [1, 3, 5, 7], 1: [2, 4, 6, 8], 2: [1.5, 3.5, 5.5, 7.5]},
+        index=["a", "a", "b", "b"],
+    )
+    msg = "The 'axis' keyword in DataFrame.groupby is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        gb = df.groupby(level=0, axis=0)
+
+    msg = "DataFrameGroupBy.rank with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        res = gb.rank(axis=1)
+
+    # This should match what we get when "manually" operating group-by-group
+    expected = concat([df.loc["a"].rank(axis=1), df.loc["b"].rank(axis=1)], axis=0)
+    tm.assert_frame_equal(res, expected)
+
+    # check that we haven't accidentally written a case that coincidentally
+    # matches rank(axis=0)
+    msg = "The 'axis' keyword in DataFrameGroupBy.rank"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        alt = gb.rank(axis=0)
+    assert not alt.equals(expected)
+
+
+def test_groupby_axis0_cummax_axis1():
+    # case where groupby axis is 0 and axis keyword in transform is 1
+
+    # df has mixed dtype -> multiple blocks
+    df = DataFrame(
+        {0: [1, 3, 5, 7], 1: [2, 4, 6, 8], 2: [1.5, 3.5, 5.5, 7.5]},
+        index=["a", "a", "b", "b"],
+    )
+    msg = "The 'axis' keyword in DataFrame.groupby is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        gb = df.groupby(level=0, axis=0)
+
+    msg = "DataFrameGroupBy.cummax with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        cmax = gb.cummax(axis=1)
+    expected = df[[0, 1]].astype(np.float64)
+    expected[2] = expected[1]
+    tm.assert_frame_equal(cmax, expected)
+
+
+def test_non_unique_index():
+    # GH 16577
+    df = DataFrame(
+        {"A": [1.0, 2.0, 3.0, np.nan], "value": 1.0},
+        index=[pd.Timestamp("20170101", tz="US/Eastern")] * 4,
+    )
+    result = df.groupby([df.index, "A"]).value.rank(ascending=True, pct=True)
+    expected = Series(
+        [1.0, 1.0, 1.0, np.nan],
+        index=[pd.Timestamp("20170101", tz="US/Eastern")] * 4,
+        name="value",
+    )
+    tm.assert_series_equal(result, expected)
+
+
+def test_rank_categorical():
+    cat = pd.Categorical(["a", "a", "b", np.nan, "c", "b"], ordered=True)
+    cat2 = pd.Categorical([1, 2, 3, np.nan, 4, 5], ordered=True)
+
+    df = DataFrame({"col1": [0, 1, 0, 1, 0, 1], "col2": cat, "col3": cat2})
+
+    gb = df.groupby("col1")
+
+    res = gb.rank()
+
+    expected = df.astype(object).groupby("col1").rank()
+    tm.assert_frame_equal(res, expected)
+
+
+@pytest.mark.parametrize("na_option", ["top", "bottom"])
+def test_groupby_op_with_nullables(na_option):
+    # GH 54206
+    df = DataFrame({"x": [None]}, dtype="Float64")
+    result = df.groupby("x", dropna=False)["x"].rank(method="min", na_option=na_option)
+    expected = Series([1.0], dtype="Float64", name=result.name)
+    tm.assert_series_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/methods/test_sample.py

@@ -0,0 +1,154 @@
+import pytest
+
+from pandas import (
+    DataFrame,
+    Index,
+    Series,
+)
+import pandas._testing as tm
+
+
+@pytest.mark.parametrize("n, frac", [(2, None), (None, 0.2)])
+def test_groupby_sample_balanced_groups_shape(n, frac):
+    values = [1] * 10 + [2] * 10
+    df = DataFrame({"a": values, "b": values})
+
+    result = df.groupby("a").sample(n=n, frac=frac)
+    values = [1] * 2 + [2] * 2
+    expected = DataFrame({"a": values, "b": values}, index=result.index)
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby("a")["b"].sample(n=n, frac=frac)
+    expected = Series(values, name="b", index=result.index)
+    tm.assert_series_equal(result, expected)
+
+
+def test_groupby_sample_unbalanced_groups_shape():
+    values = [1] * 10 + [2] * 20
+    df = DataFrame({"a": values, "b": values})
+
+    result = df.groupby("a").sample(n=5)
+    values = [1] * 5 + [2] * 5
+    expected = DataFrame({"a": values, "b": values}, index=result.index)
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby("a")["b"].sample(n=5)
+    expected = Series(values, name="b", index=result.index)
+    tm.assert_series_equal(result, expected)
+
+
+def test_groupby_sample_index_value_spans_groups():
+    values = [1] * 3 + [2] * 3
+    df = DataFrame({"a": values, "b": values}, index=[1, 2, 2, 2, 2, 2])
+
+    result = df.groupby("a").sample(n=2)
+    values = [1] * 2 + [2] * 2
+    expected = DataFrame({"a": values, "b": values}, index=result.index)
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby("a")["b"].sample(n=2)
+    expected = Series(values, name="b", index=result.index)
+    tm.assert_series_equal(result, expected)
+
+
+def test_groupby_sample_n_and_frac_raises():
+    df = DataFrame({"a": [1, 2], "b": [1, 2]})
+    msg = "Please enter a value for `frac` OR `n`, not both"
+
+    with pytest.raises(ValueError, match=msg):
+        df.groupby("a").sample(n=1, frac=1.0)
+
+    with pytest.raises(ValueError, match=msg):
+        df.groupby("a")["b"].sample(n=1, frac=1.0)
+
+
+def test_groupby_sample_frac_gt_one_without_replacement_raises():
+    df = DataFrame({"a": [1, 2], "b": [1, 2]})
+    msg = "Replace has to be set to `True` when upsampling the population `frac` > 1."
+
+    with pytest.raises(ValueError, match=msg):
+        df.groupby("a").sample(frac=1.5, replace=False)
+
+    with pytest.raises(ValueError, match=msg):
+        df.groupby("a")["b"].sample(frac=1.5, replace=False)
+
+
+@pytest.mark.parametrize("n", [-1, 1.5])
+def test_groupby_sample_invalid_n_raises(n):
+    df = DataFrame({"a": [1, 2], "b": [1, 2]})
+
+    if n < 0:
+        msg = "A negative number of rows requested. Please provide `n` >= 0."
+    else:
+        msg = "Only integers accepted as `n` values"
+
+    with pytest.raises(ValueError, match=msg):
+        df.groupby("a").sample(n=n)
+
+    with pytest.raises(ValueError, match=msg):
+        df.groupby("a")["b"].sample(n=n)
+
+
+def test_groupby_sample_oversample():
+    values = [1] * 10 + [2] * 10
+    df = DataFrame({"a": values, "b": values})
+
+    result = df.groupby("a").sample(frac=2.0, replace=True)
+    values = [1] * 20 + [2] * 20
+    expected = DataFrame({"a": values, "b": values}, index=result.index)
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby("a")["b"].sample(frac=2.0, replace=True)
+    expected = Series(values, name="b", index=result.index)
+    tm.assert_series_equal(result, expected)
+
+
+def test_groupby_sample_without_n_or_frac():
+    values = [1] * 10 + [2] * 10
+    df = DataFrame({"a": values, "b": values})
+
+    result = df.groupby("a").sample(n=None, frac=None)
+    expected = DataFrame({"a": [1, 2], "b": [1, 2]}, index=result.index)
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby("a")["b"].sample(n=None, frac=None)
+    expected = Series([1, 2], name="b", index=result.index)
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "index, expected_index",
+    [(["w", "x", "y", "z"], ["w", "w", "y", "y"]), ([3, 4, 5, 6], [3, 3, 5, 5])],
+)
+def test_groupby_sample_with_weights(index, expected_index):
+    # GH 39927 - tests for integer index needed
+    values = [1] * 2 + [2] * 2
+    df = DataFrame({"a": values, "b": values}, index=Index(index))
+
+    result = df.groupby("a").sample(n=2, replace=True, weights=[1, 0, 1, 0])
+    expected = DataFrame({"a": values, "b": values}, index=Index(expected_index))
+    tm.assert_frame_equal(result, expected)
+
+    result = df.groupby("a")["b"].sample(n=2, replace=True, weights=[1, 0, 1, 0])
+    expected = Series(values, name="b", index=Index(expected_index))
+    tm.assert_series_equal(result, expected)
+
+
+def test_groupby_sample_with_selections():
+    # GH 39928
+    values = [1] * 10 + [2] * 10
+    df = DataFrame({"a": values, "b": values, "c": values})
+
+    result = df.groupby("a")[["b", "c"]].sample(n=None, frac=None)
+    expected = DataFrame({"b": [1, 2], "c": [1, 2]}, index=result.index)
+    tm.assert_frame_equal(result, expected)
+
+
+def test_groupby_sample_with_empty_inputs():
+    # GH48459
+    df = DataFrame({"a": [], "b": []})
+    groupby_df = df.groupby("a")
+
+    result = groupby_df.sample()
+    expected = df
+    tm.assert_frame_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/methods/test_size.py

@@ -0,0 +1,130 @@
+import numpy as np
+import pytest
+
+import pandas.util._test_decorators as td
+
+from pandas.core.dtypes.common import is_integer_dtype
+
+from pandas import (
+    DataFrame,
+    Index,
+    PeriodIndex,
+    Series,
+)
+import pandas._testing as tm
+
+
+@pytest.mark.parametrize("by", ["A", "B", ["A", "B"]])
+def test_size(df, by):
+    grouped = df.groupby(by=by)
+    result = grouped.size()
+    for key, group in grouped:
+        assert result[key] == len(group)
+
+
+@pytest.mark.parametrize(
+    "by",
+    [
+        [0, 0, 0, 0],
+        [0, 1, 1, 1],
+        [1, 0, 1, 1],
+        [0, None, None, None],
+        pytest.param([None, None, None, None], marks=pytest.mark.xfail),
+    ],
+)
+def test_size_axis_1(df, axis_1, by, sort, dropna):
+    # GH#45715
+    counts = {key: sum(value == key for value in by) for key in dict.fromkeys(by)}
+    if dropna:
+        counts = {key: value for key, value in counts.items() if key is not None}
+    expected = Series(counts, dtype="int64")
+    if sort:
+        expected = expected.sort_index()
+    if is_integer_dtype(expected.index.dtype) and not any(x is None for x in by):
+        expected.index = expected.index.astype(int)
+
+    msg = "DataFrame.groupby with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        grouped = df.groupby(by=by, axis=axis_1, sort=sort, dropna=dropna)
+    result = grouped.size()
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize("by", ["A", "B", ["A", "B"]])
+@pytest.mark.parametrize("sort", [True, False])
+def test_size_sort(sort, by):
+    df = DataFrame(np.random.default_rng(2).choice(20, (1000, 3)), columns=list("ABC"))
+    left = df.groupby(by=by, sort=sort).size()
+    right = df.groupby(by=by, sort=sort)["C"].apply(lambda a: a.shape[0])
+    tm.assert_series_equal(left, right, check_names=False)
+
+
+def test_size_series_dataframe():
+    # https://github.com/pandas-dev/pandas/issues/11699
+    df = DataFrame(columns=["A", "B"])
+    out = Series(dtype="int64", index=Index([], name="A"))
+    tm.assert_series_equal(df.groupby("A").size(), out)
+
+
+def test_size_groupby_all_null():
+    # https://github.com/pandas-dev/pandas/issues/23050
+    # Assert no 'Value Error : Length of passed values is 2, index implies 0'
+    df = DataFrame({"A": [None, None]})  # all-null groups
+    result = df.groupby("A").size()
+    expected = Series(dtype="int64", index=Index([], name="A"))
+    tm.assert_series_equal(result, expected)
+
+
+def test_size_period_index():
+    # https://github.com/pandas-dev/pandas/issues/34010
+    ser = Series([1], index=PeriodIndex(["2000"], name="A", freq="D"))
+    grp = ser.groupby(level="A")
+    result = grp.size()
+    tm.assert_series_equal(result, ser)
+
+
+@pytest.mark.parametrize("as_index", [True, False])
+def test_size_on_categorical(as_index):
+    df = DataFrame([[1, 1], [2, 2]], columns=["A", "B"])
+    df["A"] = df["A"].astype("category")
+    result = df.groupby(["A", "B"], as_index=as_index, observed=False).size()
+
+    expected = DataFrame(
+        [[1, 1, 1], [1, 2, 0], [2, 1, 0], [2, 2, 1]], columns=["A", "B", "size"]
+    )
+    expected["A"] = expected["A"].astype("category")
+    if as_index:
+        expected = expected.set_index(["A", "B"])["size"].rename(None)
+
+    tm.assert_equal(result, expected)
+
+
+@pytest.mark.parametrize("dtype", ["Int64", "Float64", "boolean"])
+def test_size_series_masked_type_returns_Int64(dtype):
+    # GH 54132
+    ser = Series([1, 1, 1], index=["a", "a", "b"], dtype=dtype)
+    result = ser.groupby(level=0).size()
+    expected = Series([2, 1], dtype="Int64", index=["a", "b"])
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "dtype",
+    [
+        object,
+        pytest.param("string[pyarrow_numpy]", marks=td.skip_if_no("pyarrow")),
+        pytest.param("string[pyarrow]", marks=td.skip_if_no("pyarrow")),
+    ],
+)
+def test_size_strings(dtype):
+    # GH#55627
+    df = DataFrame({"a": ["a", "a", "b"], "b": "a"}, dtype=dtype)
+    result = df.groupby("a")["b"].size()
+    exp_dtype = "Int64" if dtype == "string[pyarrow]" else "int64"
+    expected = Series(
+        [2, 1],
+        index=Index(["a", "b"], name="a", dtype=dtype),
+        name="b",
+        dtype=exp_dtype,
+    )
+    tm.assert_series_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/methods/test_skew.py

@@ -0,0 +1,27 @@
+import numpy as np
+
+import pandas as pd
+import pandas._testing as tm
+
+
+def test_groupby_skew_equivalence():
+    # Test that that groupby skew method (which uses libgroupby.group_skew)
+    #  matches the results of operating group-by-group (which uses nanops.nanskew)
+    nrows = 1000
+    ngroups = 3
+    ncols = 2
+    nan_frac = 0.05
+
+    arr = np.random.default_rng(2).standard_normal((nrows, ncols))
+    arr[np.random.default_rng(2).random(nrows) < nan_frac] = np.nan
+
+    df = pd.DataFrame(arr)
+    grps = np.random.default_rng(2).integers(0, ngroups, size=nrows)
+    gb = df.groupby(grps)
+
+    result = gb.skew()
+
+    grpwise = [grp.skew().to_frame(i).T for i, grp in gb]
+    expected = pd.concat(grpwise, axis=0)
+    expected.index = expected.index.astype(result.index.dtype)  # 32bit builds
+    tm.assert_frame_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_all_methods.py

@@ -0,0 +1,83 @@
+"""
+Tests that apply to all groupby operation methods.
+
+The only tests that should appear here are those that use the `groupby_func` fixture.
+Even if it does use that fixture, prefer a more specific test file if it available
+such as:
+
+ - test_categorical
+ - test_groupby_dropna
+ - test_groupby_subclass
+ - test_raises
+"""
+
+import pytest
+
+import pandas as pd
+from pandas import DataFrame
+import pandas._testing as tm
+from pandas.tests.groupby import get_groupby_method_args
+
+
+def test_multiindex_group_all_columns_when_empty(groupby_func):
+    # GH 32464
+    df = DataFrame({"a": [], "b": [], "c": []}).set_index(["a", "b", "c"])
+    gb = df.groupby(["a", "b", "c"], group_keys=False)
+    method = getattr(gb, groupby_func)
+    args = get_groupby_method_args(groupby_func, df)
+
+    warn = FutureWarning if groupby_func == "fillna" else None
+    warn_msg = "DataFrameGroupBy.fillna is deprecated"
+    with tm.assert_produces_warning(warn, match=warn_msg):
+        result = method(*args).index
+    expected = df.index
+    tm.assert_index_equal(result, expected)
+
+
+def test_duplicate_columns(request, groupby_func, as_index):
+    # GH#50806
+    if groupby_func == "corrwith":
+        msg = "GH#50845 - corrwith fails when there are duplicate columns"
+        request.applymarker(pytest.mark.xfail(reason=msg))
+    df = DataFrame([[1, 3, 6], [1, 4, 7], [2, 5, 8]], columns=list("abb"))
+    args = get_groupby_method_args(groupby_func, df)
+    gb = df.groupby("a", as_index=as_index)
+    warn = FutureWarning if groupby_func == "fillna" else None
+    warn_msg = "DataFrameGroupBy.fillna is deprecated"
+    with tm.assert_produces_warning(warn, match=warn_msg):
+        result = getattr(gb, groupby_func)(*args)
+
+    expected_df = df.set_axis(["a", "b", "c"], axis=1)
+    expected_args = get_groupby_method_args(groupby_func, expected_df)
+    expected_gb = expected_df.groupby("a", as_index=as_index)
+    warn = FutureWarning if groupby_func == "fillna" else None
+    warn_msg = "DataFrameGroupBy.fillna is deprecated"
+    with tm.assert_produces_warning(warn, match=warn_msg):
+        expected = getattr(expected_gb, groupby_func)(*expected_args)
+    if groupby_func not in ("size", "ngroup", "cumcount"):
+        expected = expected.rename(columns={"c": "b"})
+    tm.assert_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "idx",
+    [
+        pd.Index(["a", "a"], name="foo"),
+        pd.MultiIndex.from_tuples((("a", "a"), ("a", "a")), names=["foo", "bar"]),
+    ],
+)
+def test_dup_labels_output_shape(groupby_func, idx):
+    if groupby_func in {"size", "ngroup", "cumcount"}:
+        pytest.skip(f"Not applicable for {groupby_func}")
+
+    df = DataFrame([[1, 1]], columns=idx)
+    grp_by = df.groupby([0])
+
+    args = get_groupby_method_args(groupby_func, df)
+    warn = FutureWarning if groupby_func == "fillna" else None
+    warn_msg = "DataFrameGroupBy.fillna is deprecated"
+    with tm.assert_produces_warning(warn, match=warn_msg):
+        result = getattr(grp_by, groupby_func)(*args)
+
+    assert result.shape == (1, 2)
+    tm.assert_index_equal(result.columns, idx)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_api.py

@@ -0,0 +1,265 @@
+"""
+Tests of the groupby API, including internal consistency and with other pandas objects.
+
+Tests in this file should only check the existence, names, and arguments of groupby
+methods. It should not test the results of any groupby operation.
+"""
+
+import inspect
+
+import pytest
+
+from pandas import (
+    DataFrame,
+    Series,
+)
+from pandas.core.groupby.base import (
+    groupby_other_methods,
+    reduction_kernels,
+    transformation_kernels,
+)
+from pandas.core.groupby.generic import (
+    DataFrameGroupBy,
+    SeriesGroupBy,
+)
+
+
+def test_tab_completion(multiindex_dataframe_random_data):
+    grp = multiindex_dataframe_random_data.groupby(level="second")
+    results = {v for v in dir(grp) if not v.startswith("_")}
+    expected = {
+        "A",
+        "B",
+        "C",
+        "agg",
+        "aggregate",
+        "apply",
+        "boxplot",
+        "filter",
+        "first",
+        "get_group",
+        "groups",
+        "hist",
+        "indices",
+        "last",
+        "max",
+        "mean",
+        "median",
+        "min",
+        "ngroups",
+        "nth",
+        "ohlc",
+        "plot",
+        "prod",
+        "size",
+        "std",
+        "sum",
+        "transform",
+        "var",
+        "sem",
+        "count",
+        "nunique",
+        "head",
+        "describe",
+        "cummax",
+        "quantile",
+        "rank",
+        "cumprod",
+        "tail",
+        "resample",
+        "cummin",
+        "fillna",
+        "cumsum",
+        "cumcount",
+        "ngroup",
+        "all",
+        "shift",
+        "skew",
+        "take",
+        "pct_change",
+        "any",
+        "corr",
+        "corrwith",
+        "cov",
+        "dtypes",
+        "ndim",
+        "diff",
+        "idxmax",
+        "idxmin",
+        "ffill",
+        "bfill",
+        "rolling",
+        "expanding",
+        "pipe",
+        "sample",
+        "ewm",
+        "value_counts",
+    }
+    assert results == expected
+
+
+def test_all_methods_categorized(multiindex_dataframe_random_data):
+    grp = multiindex_dataframe_random_data.groupby(
+        multiindex_dataframe_random_data.iloc[:, 0]
+    )
+    names = {_ for _ in dir(grp) if not _.startswith("_")} - set(
+        multiindex_dataframe_random_data.columns
+    )
+    new_names = set(names)
+    new_names -= reduction_kernels
+    new_names -= transformation_kernels
+    new_names -= groupby_other_methods
+
+    assert not reduction_kernels & transformation_kernels
+    assert not reduction_kernels & groupby_other_methods
+    assert not transformation_kernels & groupby_other_methods
+
+    # new public method?
+    if new_names:
+        msg = f"""
+There are uncategorized methods defined on the Grouper class:
+{new_names}.
+
+Was a new method recently added?
+
+Every public method On Grouper must appear in exactly one the
+following three lists defined in pandas.core.groupby.base:
+- `reduction_kernels`
+- `transformation_kernels`
+- `groupby_other_methods`
+see the comments in pandas/core/groupby/base.py for guidance on
+how to fix this test.
+        """
+        raise AssertionError(msg)
+
+    # removed a public method?
+    all_categorized = reduction_kernels | transformation_kernels | groupby_other_methods
+    if names != all_categorized:
+        msg = f"""
+Some methods which are supposed to be on the Grouper class
+are missing:
+{all_categorized - names}.
+
+They're still defined in one of the lists that live in pandas/core/groupby/base.py.
+If you removed a method, you should update them
+"""
+        raise AssertionError(msg)
+
+
+def test_frame_consistency(groupby_func):
+    # GH#48028
+    if groupby_func in ("first", "last"):
+        msg = "first and last are entirely different between frame and groupby"
+        pytest.skip(reason=msg)
+
+    if groupby_func in ("cumcount", "ngroup"):
+        assert not hasattr(DataFrame, groupby_func)
+        return
+
+    frame_method = getattr(DataFrame, groupby_func)
+    gb_method = getattr(DataFrameGroupBy, groupby_func)
+    result = set(inspect.signature(gb_method).parameters)
+    if groupby_func == "size":
+        # "size" is a method on GroupBy but property on DataFrame:
+        expected = {"self"}
+    else:
+        expected = set(inspect.signature(frame_method).parameters)
+
+    # Exclude certain arguments from result and expected depending on the operation
+    # Some of these may be purposeful inconsistencies between the APIs
+    exclude_expected, exclude_result = set(), set()
+    if groupby_func in ("any", "all"):
+        exclude_expected = {"kwargs", "bool_only", "axis"}
+    elif groupby_func in ("count",):
+        exclude_expected = {"numeric_only", "axis"}
+    elif groupby_func in ("nunique",):
+        exclude_expected = {"axis"}
+    elif groupby_func in ("max", "min"):
+        exclude_expected = {"axis", "kwargs", "skipna"}
+        exclude_result = {"min_count", "engine", "engine_kwargs"}
+    elif groupby_func in ("mean", "std", "sum", "var"):
+        exclude_expected = {"axis", "kwargs", "skipna"}
+        exclude_result = {"engine", "engine_kwargs"}
+    elif groupby_func in ("median", "prod", "sem"):
+        exclude_expected = {"axis", "kwargs", "skipna"}
+    elif groupby_func in ("backfill", "bfill", "ffill", "pad"):
+        exclude_expected = {"downcast", "inplace", "axis", "limit_area"}
+    elif groupby_func in ("cummax", "cummin"):
+        exclude_expected = {"skipna", "args"}
+        exclude_result = {"numeric_only"}
+    elif groupby_func in ("cumprod", "cumsum"):
+        exclude_expected = {"skipna"}
+    elif groupby_func in ("pct_change",):
+        exclude_expected = {"kwargs"}
+        exclude_result = {"axis"}
+    elif groupby_func in ("rank",):
+        exclude_expected = {"numeric_only"}
+    elif groupby_func in ("quantile",):
+        exclude_expected = {"method", "axis"}
+
+    # Ensure excluded arguments are actually in the signatures
+    assert result & exclude_result == exclude_result
+    assert expected & exclude_expected == exclude_expected
+
+    result -= exclude_result
+    expected -= exclude_expected
+    assert result == expected
+
+
+def test_series_consistency(request, groupby_func):
+    # GH#48028
+    if groupby_func in ("first", "last"):
+        pytest.skip("first and last are entirely different between Series and groupby")
+
+    if groupby_func in ("cumcount", "corrwith", "ngroup"):
+        assert not hasattr(Series, groupby_func)
+        return
+
+    series_method = getattr(Series, groupby_func)
+    gb_method = getattr(SeriesGroupBy, groupby_func)
+    result = set(inspect.signature(gb_method).parameters)
+    if groupby_func == "size":
+        # "size" is a method on GroupBy but property on Series
+        expected = {"self"}
+    else:
+        expected = set(inspect.signature(series_method).parameters)
+
+    # Exclude certain arguments from result and expected depending on the operation
+    # Some of these may be purposeful inconsistencies between the APIs
+    exclude_expected, exclude_result = set(), set()
+    if groupby_func in ("any", "all"):
+        exclude_expected = {"kwargs", "bool_only", "axis"}
+    elif groupby_func in ("diff",):
+        exclude_result = {"axis"}
+    elif groupby_func in ("max", "min"):
+        exclude_expected = {"axis", "kwargs", "skipna"}
+        exclude_result = {"min_count", "engine", "engine_kwargs"}
+    elif groupby_func in ("mean", "std", "sum", "var"):
+        exclude_expected = {"axis", "kwargs", "skipna"}
+        exclude_result = {"engine", "engine_kwargs"}
+    elif groupby_func in ("median", "prod", "sem"):
+        exclude_expected = {"axis", "kwargs", "skipna"}
+    elif groupby_func in ("backfill", "bfill", "ffill", "pad"):
+        exclude_expected = {"downcast", "inplace", "axis", "limit_area"}
+    elif groupby_func in ("cummax", "cummin"):
+        exclude_expected = {"skipna", "args"}
+        exclude_result = {"numeric_only"}
+    elif groupby_func in ("cumprod", "cumsum"):
+        exclude_expected = {"skipna"}
+    elif groupby_func in ("pct_change",):
+        exclude_expected = {"kwargs"}
+        exclude_result = {"axis"}
+    elif groupby_func in ("rank",):
+        exclude_expected = {"numeric_only"}
+    elif groupby_func in ("idxmin", "idxmax"):
+        exclude_expected = {"args", "kwargs"}
+    elif groupby_func in ("quantile",):
+        exclude_result = {"numeric_only"}
+
+    # Ensure excluded arguments are actually in the signatures
+    assert result & exclude_result == exclude_result
+    assert expected & exclude_expected == exclude_expected
+
+    result -= exclude_result
+    expected -= exclude_expected
+    assert result == expected

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_apply_mutate.py

@@ -0,0 +1,163 @@
+import numpy as np
+
+import pandas as pd
+import pandas._testing as tm
+
+
+def test_group_by_copy():
+    # GH#44803
+    df = pd.DataFrame(
+        {
+            "name": ["Alice", "Bob", "Carl"],
+            "age": [20, 21, 20],
+        }
+    ).set_index("name")
+
+    msg = "DataFrameGroupBy.apply operated on the grouping columns"
+    with tm.assert_produces_warning(DeprecationWarning, match=msg):
+        grp_by_same_value = df.groupby(["age"], group_keys=False).apply(
+            lambda group: group
+        )
+    msg = "DataFrameGroupBy.apply operated on the grouping columns"
+    with tm.assert_produces_warning(DeprecationWarning, match=msg):
+        grp_by_copy = df.groupby(["age"], group_keys=False).apply(
+            lambda group: group.copy()
+        )
+    tm.assert_frame_equal(grp_by_same_value, grp_by_copy)
+
+
+def test_mutate_groups():
+    # GH3380
+
+    df = pd.DataFrame(
+        {
+            "cat1": ["a"] * 8 + ["b"] * 6,
+            "cat2": ["c"] * 2
+            + ["d"] * 2
+            + ["e"] * 2
+            + ["f"] * 2
+            + ["c"] * 2
+            + ["d"] * 2
+            + ["e"] * 2,
+            "cat3": [f"g{x}" for x in range(1, 15)],
+            "val": np.random.default_rng(2).integers(100, size=14),
+        }
+    )
+
+    def f_copy(x):
+        x = x.copy()
+        x["rank"] = x.val.rank(method="min")
+        return x.groupby("cat2")["rank"].min()
+
+    def f_no_copy(x):
+        x["rank"] = x.val.rank(method="min")
+        return x.groupby("cat2")["rank"].min()
+
+    msg = "DataFrameGroupBy.apply operated on the grouping columns"
+    with tm.assert_produces_warning(DeprecationWarning, match=msg):
+        grpby_copy = df.groupby("cat1").apply(f_copy)
+    with tm.assert_produces_warning(DeprecationWarning, match=msg):
+        grpby_no_copy = df.groupby("cat1").apply(f_no_copy)
+    tm.assert_series_equal(grpby_copy, grpby_no_copy)
+
+
+def test_no_mutate_but_looks_like():
+    # GH 8467
+    # first show's mutation indicator
+    # second does not, but should yield the same results
+    df = pd.DataFrame({"key": [1, 1, 1, 2, 2, 2, 3, 3, 3], "value": range(9)})
+
+    msg = "DataFrameGroupBy.apply operated on the grouping columns"
+    with tm.assert_produces_warning(DeprecationWarning, match=msg):
+        result1 = df.groupby("key", group_keys=True).apply(lambda x: x[:].key)
+    with tm.assert_produces_warning(DeprecationWarning, match=msg):
+        result2 = df.groupby("key", group_keys=True).apply(lambda x: x.key)
+    tm.assert_series_equal(result1, result2)
+
+
+def test_apply_function_with_indexing(warn_copy_on_write):
+    # GH: 33058
+    df = pd.DataFrame(
+        {"col1": ["A", "A", "A", "B", "B", "B"], "col2": [1, 2, 3, 4, 5, 6]}
+    )
+
+    def fn(x):
+        x.loc[x.index[-1], "col2"] = 0
+        return x.col2
+
+    msg = "DataFrameGroupBy.apply operated on the grouping columns"
+    with tm.assert_produces_warning(
+        DeprecationWarning, match=msg, raise_on_extra_warnings=not warn_copy_on_write
+    ):
+        result = df.groupby(["col1"], as_index=False).apply(fn)
+    expected = pd.Series(
+        [1, 2, 0, 4, 5, 0],
+        index=pd.MultiIndex.from_tuples(
+            [(0, 0), (0, 1), (0, 2), (1, 3), (1, 4), (1, 5)]
+        ),
+        name="col2",
+    )
+    tm.assert_series_equal(result, expected)
+
+
+def test_apply_mutate_columns_multiindex():
+    # GH 12652
+    df = pd.DataFrame(
+        {
+            ("C", "julian"): [1, 2, 3],
+            ("B", "geoffrey"): [1, 2, 3],
+            ("A", "julian"): [1, 2, 3],
+            ("B", "julian"): [1, 2, 3],
+            ("A", "geoffrey"): [1, 2, 3],
+            ("C", "geoffrey"): [1, 2, 3],
+        },
+        columns=pd.MultiIndex.from_tuples(
+            [
+                ("A", "julian"),
+                ("A", "geoffrey"),
+                ("B", "julian"),
+                ("B", "geoffrey"),
+                ("C", "julian"),
+                ("C", "geoffrey"),
+            ]
+        ),
+    )
+
+    def add_column(grouped):
+        name = grouped.columns[0][1]
+        grouped["sum", name] = grouped.sum(axis=1)
+        return grouped
+
+    msg = "DataFrame.groupby with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        gb = df.groupby(level=1, axis=1)
+    result = gb.apply(add_column)
+    expected = pd.DataFrame(
+        [
+            [1, 1, 1, 3, 1, 1, 1, 3],
+            [2, 2, 2, 6, 2, 2, 2, 6],
+            [
+                3,
+                3,
+                3,
+                9,
+                3,
+                3,
+                3,
+                9,
+            ],
+        ],
+        columns=pd.MultiIndex.from_tuples(
+            [
+                ("geoffrey", "A", "geoffrey"),
+                ("geoffrey", "B", "geoffrey"),
+                ("geoffrey", "C", "geoffrey"),
+                ("geoffrey", "sum", "geoffrey"),
+                ("julian", "A", "julian"),
+                ("julian", "B", "julian"),
+                ("julian", "C", "julian"),
+                ("julian", "sum", "julian"),
+            ]
+        ),
+    )
+    tm.assert_frame_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_bin_groupby.py

@@ -0,0 +1,65 @@
+import numpy as np
+import pytest
+
+from pandas._libs import lib
+import pandas.util._test_decorators as td
+
+import pandas as pd
+import pandas._testing as tm
+
+
+def assert_block_lengths(x):
+    assert len(x) == len(x._mgr.blocks[0].mgr_locs)
+    return 0
+
+
+def cumsum_max(x):
+    x.cumsum().max()
+    return 0
+
+
+@pytest.mark.parametrize(
+    "func",
+    [
+        cumsum_max,
+        pytest.param(assert_block_lengths, marks=td.skip_array_manager_invalid_test),
+    ],
+)
+def test_mgr_locs_updated(func):
+    # https://github.com/pandas-dev/pandas/issues/31802
+    # Some operations may require creating new blocks, which requires
+    # valid mgr_locs
+    df = pd.DataFrame({"A": ["a", "a", "a"], "B": ["a", "b", "b"], "C": [1, 1, 1]})
+    result = df.groupby(["A", "B"]).agg(func)
+    expected = pd.DataFrame(
+        {"C": [0, 0]},
+        index=pd.MultiIndex.from_product([["a"], ["a", "b"]], names=["A", "B"]),
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "binner,closed,expected",
+    [
+        (
+            np.array([0, 3, 6, 9], dtype=np.int64),
+            "left",
+            np.array([2, 5, 6], dtype=np.int64),
+        ),
+        (
+            np.array([0, 3, 6, 9], dtype=np.int64),
+            "right",
+            np.array([3, 6, 6], dtype=np.int64),
+        ),
+        (np.array([0, 3, 6], dtype=np.int64), "left", np.array([2, 5], dtype=np.int64)),
+        (
+            np.array([0, 3, 6], dtype=np.int64),
+            "right",
+            np.array([3, 6], dtype=np.int64),
+        ),
+    ],
+)
+def test_generate_bins(binner, closed, expected):
+    values = np.array([1, 2, 3, 4, 5, 6], dtype=np.int64)
+    result = lib.generate_bins_dt64(values, binner, closed=closed)
+    tm.assert_numpy_array_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_counting.py

@@ -0,0 +1,394 @@
+from itertools import product
+from string import ascii_lowercase
+
+import numpy as np
+import pytest
+
+from pandas import (
+    DataFrame,
+    Index,
+    MultiIndex,
+    Period,
+    Series,
+    Timedelta,
+    Timestamp,
+    date_range,
+)
+import pandas._testing as tm
+
+
+class TestCounting:
+    def test_cumcount(self):
+        df = DataFrame([["a"], ["a"], ["a"], ["b"], ["a"]], columns=["A"])
+        g = df.groupby("A")
+        sg = g.A
+
+        expected = Series([0, 1, 2, 0, 3])
+
+        tm.assert_series_equal(expected, g.cumcount())
+        tm.assert_series_equal(expected, sg.cumcount())
+
+    def test_cumcount_empty(self):
+        ge = DataFrame().groupby(level=0)
+        se = Series(dtype=object).groupby(level=0)
+
+        # edge case, as this is usually considered float
+        e = Series(dtype="int64")
+
+        tm.assert_series_equal(e, ge.cumcount())
+        tm.assert_series_equal(e, se.cumcount())
+
+    def test_cumcount_dupe_index(self):
+        df = DataFrame(
+            [["a"], ["a"], ["a"], ["b"], ["a"]], columns=["A"], index=[0] * 5
+        )
+        g = df.groupby("A")
+        sg = g.A
+
+        expected = Series([0, 1, 2, 0, 3], index=[0] * 5)
+
+        tm.assert_series_equal(expected, g.cumcount())
+        tm.assert_series_equal(expected, sg.cumcount())
+
+    def test_cumcount_mi(self):
+        mi = MultiIndex.from_tuples([[0, 1], [1, 2], [2, 2], [2, 2], [1, 0]])
+        df = DataFrame([["a"], ["a"], ["a"], ["b"], ["a"]], columns=["A"], index=mi)
+        g = df.groupby("A")
+        sg = g.A
+
+        expected = Series([0, 1, 2, 0, 3], index=mi)
+
+        tm.assert_series_equal(expected, g.cumcount())
+        tm.assert_series_equal(expected, sg.cumcount())
+
+    def test_cumcount_groupby_not_col(self):
+        df = DataFrame(
+            [["a"], ["a"], ["a"], ["b"], ["a"]], columns=["A"], index=[0] * 5
+        )
+        g = df.groupby([0, 0, 0, 1, 0])
+        sg = g.A
+
+        expected = Series([0, 1, 2, 0, 3], index=[0] * 5)
+
+        tm.assert_series_equal(expected, g.cumcount())
+        tm.assert_series_equal(expected, sg.cumcount())
+
+    def test_ngroup(self):
+        df = DataFrame({"A": list("aaaba")})
+        g = df.groupby("A")
+        sg = g.A
+
+        expected = Series([0, 0, 0, 1, 0])
+
+        tm.assert_series_equal(expected, g.ngroup())
+        tm.assert_series_equal(expected, sg.ngroup())
+
+    def test_ngroup_distinct(self):
+        df = DataFrame({"A": list("abcde")})
+        g = df.groupby("A")
+        sg = g.A
+
+        expected = Series(range(5), dtype="int64")
+
+        tm.assert_series_equal(expected, g.ngroup())
+        tm.assert_series_equal(expected, sg.ngroup())
+
+    def test_ngroup_one_group(self):
+        df = DataFrame({"A": [0] * 5})
+        g = df.groupby("A")
+        sg = g.A
+
+        expected = Series([0] * 5)
+
+        tm.assert_series_equal(expected, g.ngroup())
+        tm.assert_series_equal(expected, sg.ngroup())
+
+    def test_ngroup_empty(self):
+        ge = DataFrame().groupby(level=0)
+        se = Series(dtype=object).groupby(level=0)
+
+        # edge case, as this is usually considered float
+        e = Series(dtype="int64")
+
+        tm.assert_series_equal(e, ge.ngroup())
+        tm.assert_series_equal(e, se.ngroup())
+
+    def test_ngroup_series_matches_frame(self):
+        df = DataFrame({"A": list("aaaba")})
+        s = Series(list("aaaba"))
+
+        tm.assert_series_equal(df.groupby(s).ngroup(), s.groupby(s).ngroup())
+
+    def test_ngroup_dupe_index(self):
+        df = DataFrame({"A": list("aaaba")}, index=[0] * 5)
+        g = df.groupby("A")
+        sg = g.A
+
+        expected = Series([0, 0, 0, 1, 0], index=[0] * 5)
+
+        tm.assert_series_equal(expected, g.ngroup())
+        tm.assert_series_equal(expected, sg.ngroup())
+
+    def test_ngroup_mi(self):
+        mi = MultiIndex.from_tuples([[0, 1], [1, 2], [2, 2], [2, 2], [1, 0]])
+        df = DataFrame({"A": list("aaaba")}, index=mi)
+        g = df.groupby("A")
+        sg = g.A
+        expected = Series([0, 0, 0, 1, 0], index=mi)
+
+        tm.assert_series_equal(expected, g.ngroup())
+        tm.assert_series_equal(expected, sg.ngroup())
+
+    def test_ngroup_groupby_not_col(self):
+        df = DataFrame({"A": list("aaaba")}, index=[0] * 5)
+        g = df.groupby([0, 0, 0, 1, 0])
+        sg = g.A
+
+        expected = Series([0, 0, 0, 1, 0], index=[0] * 5)
+
+        tm.assert_series_equal(expected, g.ngroup())
+        tm.assert_series_equal(expected, sg.ngroup())
+
+    def test_ngroup_descending(self):
+        df = DataFrame(["a", "a", "b", "a", "b"], columns=["A"])
+        g = df.groupby(["A"])
+
+        ascending = Series([0, 0, 1, 0, 1])
+        descending = Series([1, 1, 0, 1, 0])
+
+        tm.assert_series_equal(descending, (g.ngroups - 1) - ascending)
+        tm.assert_series_equal(ascending, g.ngroup(ascending=True))
+        tm.assert_series_equal(descending, g.ngroup(ascending=False))
+
+    def test_ngroup_matches_cumcount(self):
+        # verify one manually-worked out case works
+        df = DataFrame(
+            [["a", "x"], ["a", "y"], ["b", "x"], ["a", "x"], ["b", "y"]],
+            columns=["A", "X"],
+        )
+        g = df.groupby(["A", "X"])
+        g_ngroup = g.ngroup()
+        g_cumcount = g.cumcount()
+        expected_ngroup = Series([0, 1, 2, 0, 3])
+        expected_cumcount = Series([0, 0, 0, 1, 0])
+
+        tm.assert_series_equal(g_ngroup, expected_ngroup)
+        tm.assert_series_equal(g_cumcount, expected_cumcount)
+
+    def test_ngroup_cumcount_pair(self):
+        # brute force comparison for all small series
+        for p in product(range(3), repeat=4):
+            df = DataFrame({"a": p})
+            g = df.groupby(["a"])
+
+            order = sorted(set(p))
+            ngroupd = [order.index(val) for val in p]
+            cumcounted = [p[:i].count(val) for i, val in enumerate(p)]
+
+            tm.assert_series_equal(g.ngroup(), Series(ngroupd))
+            tm.assert_series_equal(g.cumcount(), Series(cumcounted))
+
+    def test_ngroup_respects_groupby_order(self, sort):
+        df = DataFrame({"a": np.random.default_rng(2).choice(list("abcdef"), 100)})
+        g = df.groupby("a", sort=sort)
+        df["group_id"] = -1
+        df["group_index"] = -1
+
+        for i, (_, group) in enumerate(g):
+            df.loc[group.index, "group_id"] = i
+            for j, ind in enumerate(group.index):
+                df.loc[ind, "group_index"] = j
+
+        tm.assert_series_equal(Series(df["group_id"].values), g.ngroup())
+        tm.assert_series_equal(Series(df["group_index"].values), g.cumcount())
+
+    @pytest.mark.parametrize(
+        "datetimelike",
+        [
+            [Timestamp(f"2016-05-{i:02d} 20:09:25+00:00") for i in range(1, 4)],
+            [Timestamp(f"2016-05-{i:02d} 20:09:25") for i in range(1, 4)],
+            [Timestamp(f"2016-05-{i:02d} 20:09:25", tz="UTC") for i in range(1, 4)],
+            [Timedelta(x, unit="h") for x in range(1, 4)],
+            [Period(freq="2W", year=2017, month=x) for x in range(1, 4)],
+        ],
+    )
+    def test_count_with_datetimelike(self, datetimelike):
+        # test for #13393, where DataframeGroupBy.count() fails
+        # when counting a datetimelike column.
+
+        df = DataFrame({"x": ["a", "a", "b"], "y": datetimelike})
+        res = df.groupby("x").count()
+        expected = DataFrame({"y": [2, 1]}, index=["a", "b"])
+        expected.index.name = "x"
+        tm.assert_frame_equal(expected, res)
+
+    def test_count_with_only_nans_in_first_group(self):
+        # GH21956
+        df = DataFrame({"A": [np.nan, np.nan], "B": ["a", "b"], "C": [1, 2]})
+        result = df.groupby(["A", "B"]).C.count()
+        mi = MultiIndex(levels=[[], ["a", "b"]], codes=[[], []], names=["A", "B"])
+        expected = Series([], index=mi, dtype=np.int64, name="C")
+        tm.assert_series_equal(result, expected, check_index_type=False)
+
+    def test_count_groupby_column_with_nan_in_groupby_column(self):
+        # https://github.com/pandas-dev/pandas/issues/32841
+        df = DataFrame({"A": [1, 1, 1, 1, 1], "B": [5, 4, np.nan, 3, 0]})
+        res = df.groupby(["B"]).count()
+        expected = DataFrame(
+            index=Index([0.0, 3.0, 4.0, 5.0], name="B"), data={"A": [1, 1, 1, 1]}
+        )
+        tm.assert_frame_equal(expected, res)
+
+    def test_groupby_count_dateparseerror(self):
+        dr = date_range(start="1/1/2012", freq="5min", periods=10)
+
+        # BAD Example, datetimes first
+        ser = Series(np.arange(10), index=[dr, np.arange(10)])
+        grouped = ser.groupby(lambda x: x[1] % 2 == 0)
+        result = grouped.count()
+
+        ser = Series(np.arange(10), index=[np.arange(10), dr])
+        grouped = ser.groupby(lambda x: x[0] % 2 == 0)
+        expected = grouped.count()
+
+        tm.assert_series_equal(result, expected)
+
+
+def test_groupby_timedelta_cython_count():
+    df = DataFrame(
+        {"g": list("ab" * 2), "delta": np.arange(4).astype("timedelta64[ns]")}
+    )
+    expected = Series([2, 2], index=Index(["a", "b"], name="g"), name="delta")
+    result = df.groupby("g").delta.count()
+    tm.assert_series_equal(expected, result)
+
+
+def test_count():
+    n = 1 << 15
+    dr = date_range("2015-08-30", periods=n // 10, freq="min")
+
+    df = DataFrame(
+        {
+            "1st": np.random.default_rng(2).choice(list(ascii_lowercase), n),
+            "2nd": np.random.default_rng(2).integers(0, 5, n),
+            "3rd": np.random.default_rng(2).standard_normal(n).round(3),
+            "4th": np.random.default_rng(2).integers(-10, 10, n),
+            "5th": np.random.default_rng(2).choice(dr, n),
+            "6th": np.random.default_rng(2).standard_normal(n).round(3),
+            "7th": np.random.default_rng(2).standard_normal(n).round(3),
+            "8th": np.random.default_rng(2).choice(dr, n)
+            - np.random.default_rng(2).choice(dr, 1),
+            "9th": np.random.default_rng(2).choice(list(ascii_lowercase), n),
+        }
+    )
+
+    for col in df.columns.drop(["1st", "2nd", "4th"]):
+        df.loc[np.random.default_rng(2).choice(n, n // 10), col] = np.nan
+
+    df["9th"] = df["9th"].astype("category")
+
+    for key in ["1st", "2nd", ["1st", "2nd"]]:
+        left = df.groupby(key).count()
+        msg = "DataFrameGroupBy.apply operated on the grouping columns"
+        with tm.assert_produces_warning(DeprecationWarning, match=msg):
+            right = df.groupby(key).apply(DataFrame.count).drop(key, axis=1)
+        tm.assert_frame_equal(left, right)
+
+
+def test_count_non_nulls():
+    # GH#5610
+    # count counts non-nulls
+    df = DataFrame(
+        [[1, 2, "foo"], [1, np.nan, "bar"], [3, np.nan, np.nan]],
+        columns=["A", "B", "C"],
+    )
+
+    count_as = df.groupby("A").count()
+    count_not_as = df.groupby("A", as_index=False).count()
+
+    expected = DataFrame([[1, 2], [0, 0]], columns=["B", "C"], index=[1, 3])
+    expected.index.name = "A"
+    tm.assert_frame_equal(count_not_as, expected.reset_index())
+    tm.assert_frame_equal(count_as, expected)
+
+    count_B = df.groupby("A")["B"].count()
+    tm.assert_series_equal(count_B, expected["B"])
+
+
+def test_count_object():
+    df = DataFrame({"a": ["a"] * 3 + ["b"] * 3, "c": [2] * 3 + [3] * 3})
+    result = df.groupby("c").a.count()
+    expected = Series([3, 3], index=Index([2, 3], name="c"), name="a")
+    tm.assert_series_equal(result, expected)
+
+    df = DataFrame({"a": ["a", np.nan, np.nan] + ["b"] * 3, "c": [2] * 3 + [3] * 3})
+    result = df.groupby("c").a.count()
+    expected = Series([1, 3], index=Index([2, 3], name="c"), name="a")
+    tm.assert_series_equal(result, expected)
+
+
+def test_count_cross_type():
+    # GH8169
+    # Set float64 dtype to avoid upcast when setting nan below
+    vals = np.hstack(
+        (
+            np.random.default_rng(2).integers(0, 5, (100, 2)),
+            np.random.default_rng(2).integers(0, 2, (100, 2)),
+        )
+    ).astype("float64")
+
+    df = DataFrame(vals, columns=["a", "b", "c", "d"])
+    df[df == 2] = np.nan
+    expected = df.groupby(["c", "d"]).count()
+
+    for t in ["float32", "object"]:
+        df["a"] = df["a"].astype(t)
+        df["b"] = df["b"].astype(t)
+        result = df.groupby(["c", "d"]).count()
+        tm.assert_frame_equal(result, expected)
+
+
+def test_lower_int_prec_count():
+    df = DataFrame(
+        {
+            "a": np.array([0, 1, 2, 100], np.int8),
+            "b": np.array([1, 2, 3, 6], np.uint32),
+            "c": np.array([4, 5, 6, 8], np.int16),
+            "grp": list("ab" * 2),
+        }
+    )
+    result = df.groupby("grp").count()
+    expected = DataFrame(
+        {"a": [2, 2], "b": [2, 2], "c": [2, 2]}, index=Index(list("ab"), name="grp")
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+def test_count_uses_size_on_exception():
+    class RaisingObjectException(Exception):
+        pass
+
+    class RaisingObject:
+        def __init__(self, msg="I will raise inside Cython") -> None:
+            super().__init__()
+            self.msg = msg
+
+        def __eq__(self, other):
+            # gets called in Cython to check that raising calls the method
+            raise RaisingObjectException(self.msg)
+
+    df = DataFrame({"a": [RaisingObject() for _ in range(4)], "grp": list("ab" * 2)})
+    result = df.groupby("grp").count()
+    expected = DataFrame({"a": [2, 2]}, index=Index(list("ab"), name="grp"))
+    tm.assert_frame_equal(result, expected)
+
+
+def test_count_arrow_string_array(any_string_dtype):
+    # GH#54751
+    pytest.importorskip("pyarrow")
+    df = DataFrame(
+        {"a": [1, 2, 3], "b": Series(["a", "b", "a"], dtype=any_string_dtype)}
+    )
+    result = df.groupby("a").count()
+    expected = DataFrame({"b": 1}, index=Index([1, 2, 3], name="a"))
+    tm.assert_frame_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_cumulative.py

@@ -0,0 +1,319 @@
+import numpy as np
+import pytest
+
+from pandas.errors import UnsupportedFunctionCall
+import pandas.util._test_decorators as td
+
+import pandas as pd
+from pandas import (
+    DataFrame,
+    Series,
+)
+import pandas._testing as tm
+
+
+@pytest.fixture(
+    params=[np.int32, np.int64, np.float32, np.float64, "Int64", "Float64"],
+    ids=["np.int32", "np.int64", "np.float32", "np.float64", "Int64", "Float64"],
+)
+def dtypes_for_minmax(request):
+    """
+    Fixture of dtypes with min and max values used for testing
+    cummin and cummax
+    """
+    dtype = request.param
+
+    np_type = dtype
+    if dtype == "Int64":
+        np_type = np.int64
+    elif dtype == "Float64":
+        np_type = np.float64
+
+    min_val = (
+        np.iinfo(np_type).min
+        if np.dtype(np_type).kind == "i"
+        else np.finfo(np_type).min
+    )
+    max_val = (
+        np.iinfo(np_type).max
+        if np.dtype(np_type).kind == "i"
+        else np.finfo(np_type).max
+    )
+
+    return (dtype, min_val, max_val)
+
+
+def test_groupby_cumprod():
+    # GH 4095
+    df = DataFrame({"key": ["b"] * 10, "value": 2})
+
+    actual = df.groupby("key")["value"].cumprod()
+    expected = df.groupby("key", group_keys=False)["value"].apply(lambda x: x.cumprod())
+    expected.name = "value"
+    tm.assert_series_equal(actual, expected)
+
+    df = DataFrame({"key": ["b"] * 100, "value": 2})
+    df["value"] = df["value"].astype(float)
+    actual = df.groupby("key")["value"].cumprod()
+    expected = df.groupby("key", group_keys=False)["value"].apply(lambda x: x.cumprod())
+    expected.name = "value"
+    tm.assert_series_equal(actual, expected)
+
+
+@pytest.mark.skip_ubsan
+def test_groupby_cumprod_overflow():
+    # GH#37493 if we overflow we return garbage consistent with numpy
+    df = DataFrame({"key": ["b"] * 4, "value": 100_000})
+    actual = df.groupby("key")["value"].cumprod()
+    expected = Series(
+        [100_000, 10_000_000_000, 1_000_000_000_000_000, 7766279631452241920],
+        name="value",
+    )
+    tm.assert_series_equal(actual, expected)
+
+    numpy_result = df.groupby("key", group_keys=False)["value"].apply(
+        lambda x: x.cumprod()
+    )
+    numpy_result.name = "value"
+    tm.assert_series_equal(actual, numpy_result)
+
+
+def test_groupby_cumprod_nan_influences_other_columns():
+    # GH#48064
+    df = DataFrame(
+        {
+            "a": 1,
+            "b": [1, np.nan, 2],
+            "c": [1, 2, 3.0],
+        }
+    )
+    result = df.groupby("a").cumprod(numeric_only=True, skipna=False)
+    expected = DataFrame({"b": [1, np.nan, np.nan], "c": [1, 2, 6.0]})
+    tm.assert_frame_equal(result, expected)
+
+
+def test_cummin(dtypes_for_minmax):
+    dtype = dtypes_for_minmax[0]
+    min_val = dtypes_for_minmax[1]
+
+    # GH 15048
+    base_df = DataFrame({"A": [1, 1, 1, 1, 2, 2, 2, 2], "B": [3, 4, 3, 2, 2, 3, 2, 1]})
+    expected_mins = [3, 3, 3, 2, 2, 2, 2, 1]
+
+    df = base_df.astype(dtype)
+
+    expected = DataFrame({"B": expected_mins}).astype(dtype)
+    result = df.groupby("A").cummin()
+    tm.assert_frame_equal(result, expected)
+    result = df.groupby("A", group_keys=False).B.apply(lambda x: x.cummin()).to_frame()
+    tm.assert_frame_equal(result, expected)
+
+    # Test w/ min value for dtype
+    df.loc[[2, 6], "B"] = min_val
+    df.loc[[1, 5], "B"] = min_val + 1
+    expected.loc[[2, 3, 6, 7], "B"] = min_val
+    expected.loc[[1, 5], "B"] = min_val + 1  # should not be rounded to min_val
+    result = df.groupby("A").cummin()
+    tm.assert_frame_equal(result, expected, check_exact=True)
+    expected = (
+        df.groupby("A", group_keys=False).B.apply(lambda x: x.cummin()).to_frame()
+    )
+    tm.assert_frame_equal(result, expected, check_exact=True)
+
+    # Test nan in some values
+    # Explicit cast to float to avoid implicit cast when setting nan
+    base_df = base_df.astype({"B": "float"})
+    base_df.loc[[0, 2, 4, 6], "B"] = np.nan
+    expected = DataFrame({"B": [np.nan, 4, np.nan, 2, np.nan, 3, np.nan, 1]})
+    result = base_df.groupby("A").cummin()
+    tm.assert_frame_equal(result, expected)
+    expected = (
+        base_df.groupby("A", group_keys=False).B.apply(lambda x: x.cummin()).to_frame()
+    )
+    tm.assert_frame_equal(result, expected)
+
+    # GH 15561
+    df = DataFrame({"a": [1], "b": pd.to_datetime(["2001"])})
+    expected = Series(pd.to_datetime("2001"), index=[0], name="b")
+
+    result = df.groupby("a")["b"].cummin()
+    tm.assert_series_equal(expected, result)
+
+    # GH 15635
+    df = DataFrame({"a": [1, 2, 1], "b": [1, 2, 2]})
+    result = df.groupby("a").b.cummin()
+    expected = Series([1, 2, 1], name="b")
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize("method", ["cummin", "cummax"])
+@pytest.mark.parametrize("dtype", ["UInt64", "Int64", "Float64", "float", "boolean"])
+def test_cummin_max_all_nan_column(method, dtype):
+    base_df = DataFrame({"A": [1, 1, 1, 1, 2, 2, 2, 2], "B": [np.nan] * 8})
+    base_df["B"] = base_df["B"].astype(dtype)
+    grouped = base_df.groupby("A")
+
+    expected = DataFrame({"B": [np.nan] * 8}, dtype=dtype)
+    result = getattr(grouped, method)()
+    tm.assert_frame_equal(expected, result)
+
+    result = getattr(grouped["B"], method)().to_frame()
+    tm.assert_frame_equal(expected, result)
+
+
+def test_cummax(dtypes_for_minmax):
+    dtype = dtypes_for_minmax[0]
+    max_val = dtypes_for_minmax[2]
+
+    # GH 15048
+    base_df = DataFrame({"A": [1, 1, 1, 1, 2, 2, 2, 2], "B": [3, 4, 3, 2, 2, 3, 2, 1]})
+    expected_maxs = [3, 4, 4, 4, 2, 3, 3, 3]
+
+    df = base_df.astype(dtype)
+
+    expected = DataFrame({"B": expected_maxs}).astype(dtype)
+    result = df.groupby("A").cummax()
+    tm.assert_frame_equal(result, expected)
+    result = df.groupby("A", group_keys=False).B.apply(lambda x: x.cummax()).to_frame()
+    tm.assert_frame_equal(result, expected)
+
+    # Test w/ max value for dtype
+    df.loc[[2, 6], "B"] = max_val
+    expected.loc[[2, 3, 6, 7], "B"] = max_val
+    result = df.groupby("A").cummax()
+    tm.assert_frame_equal(result, expected)
+    expected = (
+        df.groupby("A", group_keys=False).B.apply(lambda x: x.cummax()).to_frame()
+    )
+    tm.assert_frame_equal(result, expected)
+
+    # Test nan in some values
+    # Explicit cast to float to avoid implicit cast when setting nan
+    base_df = base_df.astype({"B": "float"})
+    base_df.loc[[0, 2, 4, 6], "B"] = np.nan
+    expected = DataFrame({"B": [np.nan, 4, np.nan, 4, np.nan, 3, np.nan, 3]})
+    result = base_df.groupby("A").cummax()
+    tm.assert_frame_equal(result, expected)
+    expected = (
+        base_df.groupby("A", group_keys=False).B.apply(lambda x: x.cummax()).to_frame()
+    )
+    tm.assert_frame_equal(result, expected)
+
+    # GH 15561
+    df = DataFrame({"a": [1], "b": pd.to_datetime(["2001"])})
+    expected = Series(pd.to_datetime("2001"), index=[0], name="b")
+
+    result = df.groupby("a")["b"].cummax()
+    tm.assert_series_equal(expected, result)
+
+    # GH 15635
+    df = DataFrame({"a": [1, 2, 1], "b": [2, 1, 1]})
+    result = df.groupby("a").b.cummax()
+    expected = Series([2, 1, 2], name="b")
+    tm.assert_series_equal(result, expected)
+
+
+def test_cummax_i8_at_implementation_bound():
+    # the minimum value used to be treated as NPY_NAT+1 instead of NPY_NAT
+    #  for int64 dtype GH#46382
+    ser = Series([pd.NaT._value + n for n in range(5)])
+    df = DataFrame({"A": 1, "B": ser, "C": ser._values.view("M8[ns]")})
+    gb = df.groupby("A")
+
+    res = gb.cummax()
+    exp = df[["B", "C"]]
+    tm.assert_frame_equal(res, exp)
+
+
+@pytest.mark.parametrize("method", ["cummin", "cummax"])
+@pytest.mark.parametrize("dtype", ["float", "Int64", "Float64"])
+@pytest.mark.parametrize(
+    "groups,expected_data",
+    [
+        ([1, 1, 1], [1, None, None]),
+        ([1, 2, 3], [1, None, 2]),
+        ([1, 3, 3], [1, None, None]),
+    ],
+)
+def test_cummin_max_skipna(method, dtype, groups, expected_data):
+    # GH-34047
+    df = DataFrame({"a": Series([1, None, 2], dtype=dtype)})
+    orig = df.copy()
+    gb = df.groupby(groups)["a"]
+
+    result = getattr(gb, method)(skipna=False)
+    expected = Series(expected_data, dtype=dtype, name="a")
+
+    # check we didn't accidentally alter df
+    tm.assert_frame_equal(df, orig)
+
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize("method", ["cummin", "cummax"])
+def test_cummin_max_skipna_multiple_cols(method):
+    # Ensure missing value in "a" doesn't cause "b" to be nan-filled
+    df = DataFrame({"a": [np.nan, 2.0, 2.0], "b": [2.0, 2.0, 2.0]})
+    gb = df.groupby([1, 1, 1])[["a", "b"]]
+
+    result = getattr(gb, method)(skipna=False)
+    expected = DataFrame({"a": [np.nan, np.nan, np.nan], "b": [2.0, 2.0, 2.0]})
+
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("func", ["cumprod", "cumsum"])
+def test_numpy_compat(func):
+    # see gh-12811
+    df = DataFrame({"A": [1, 2, 1], "B": [1, 2, 3]})
+    g = df.groupby("A")
+
+    msg = "numpy operations are not valid with groupby"
+
+    with pytest.raises(UnsupportedFunctionCall, match=msg):
+        getattr(g, func)(1, 2, 3)
+    with pytest.raises(UnsupportedFunctionCall, match=msg):
+        getattr(g, func)(foo=1)
+
+
+@td.skip_if_32bit
+@pytest.mark.parametrize("method", ["cummin", "cummax"])
+@pytest.mark.parametrize(
+    "dtype,val", [("UInt64", np.iinfo("uint64").max), ("Int64", 2**53 + 1)]
+)
+def test_nullable_int_not_cast_as_float(method, dtype, val):
+    data = [val, pd.NA]
+    df = DataFrame({"grp": [1, 1], "b": data}, dtype=dtype)
+    grouped = df.groupby("grp")
+
+    result = grouped.transform(method)
+    expected = DataFrame({"b": data}, dtype=dtype)
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_cython_api2():
+    # this takes the fast apply path
+
+    # cumsum (GH5614)
+    df = DataFrame([[1, 2, np.nan], [1, np.nan, 9], [3, 4, 9]], columns=["A", "B", "C"])
+    expected = DataFrame([[2, np.nan], [np.nan, 9], [4, 9]], columns=["B", "C"])
+    result = df.groupby("A").cumsum()
+    tm.assert_frame_equal(result, expected)
+
+    # GH 5755 - cumsum is a transformer and should ignore as_index
+    result = df.groupby("A", as_index=False).cumsum()
+    tm.assert_frame_equal(result, expected)
+
+    # GH 13994
+    msg = "DataFrameGroupBy.cumsum with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        result = df.groupby("A").cumsum(axis=1)
+    expected = df.cumsum(axis=1)
+    tm.assert_frame_equal(result, expected)
+
+    msg = "DataFrameGroupBy.cumprod with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        result = df.groupby("A").cumprod(axis=1)
+    expected = df.cumprod(axis=1)
+    tm.assert_frame_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_filters.py

@@ -0,0 +1,636 @@
+from string import ascii_lowercase
+
+import numpy as np
+import pytest
+
+import pandas as pd
+from pandas import (
+    DataFrame,
+    Series,
+    Timestamp,
+)
+import pandas._testing as tm
+
+
+def test_filter_series():
+    s = Series([1, 3, 20, 5, 22, 24, 7])
+    expected_odd = Series([1, 3, 5, 7], index=[0, 1, 3, 6])
+    expected_even = Series([20, 22, 24], index=[2, 4, 5])
+    grouper = s.apply(lambda x: x % 2)
+    grouped = s.groupby(grouper)
+    tm.assert_series_equal(grouped.filter(lambda x: x.mean() < 10), expected_odd)
+    tm.assert_series_equal(grouped.filter(lambda x: x.mean() > 10), expected_even)
+    # Test dropna=False.
+    tm.assert_series_equal(
+        grouped.filter(lambda x: x.mean() < 10, dropna=False),
+        expected_odd.reindex(s.index),
+    )
+    tm.assert_series_equal(
+        grouped.filter(lambda x: x.mean() > 10, dropna=False),
+        expected_even.reindex(s.index),
+    )
+
+
+def test_filter_single_column_df():
+    df = DataFrame([1, 3, 20, 5, 22, 24, 7])
+    expected_odd = DataFrame([1, 3, 5, 7], index=[0, 1, 3, 6])
+    expected_even = DataFrame([20, 22, 24], index=[2, 4, 5])
+    grouper = df[0].apply(lambda x: x % 2)
+    grouped = df.groupby(grouper)
+    tm.assert_frame_equal(grouped.filter(lambda x: x.mean() < 10), expected_odd)
+    tm.assert_frame_equal(grouped.filter(lambda x: x.mean() > 10), expected_even)
+    # Test dropna=False.
+    tm.assert_frame_equal(
+        grouped.filter(lambda x: x.mean() < 10, dropna=False),
+        expected_odd.reindex(df.index),
+    )
+    tm.assert_frame_equal(
+        grouped.filter(lambda x: x.mean() > 10, dropna=False),
+        expected_even.reindex(df.index),
+    )
+
+
+def test_filter_multi_column_df():
+    df = DataFrame({"A": [1, 12, 12, 1], "B": [1, 1, 1, 1]})
+    grouper = df["A"].apply(lambda x: x % 2)
+    grouped = df.groupby(grouper)
+    expected = DataFrame({"A": [12, 12], "B": [1, 1]}, index=[1, 2])
+    tm.assert_frame_equal(
+        grouped.filter(lambda x: x["A"].sum() - x["B"].sum() > 10), expected
+    )
+
+
+def test_filter_mixed_df():
+    df = DataFrame({"A": [1, 12, 12, 1], "B": "a b c d".split()})
+    grouper = df["A"].apply(lambda x: x % 2)
+    grouped = df.groupby(grouper)
+    expected = DataFrame({"A": [12, 12], "B": ["b", "c"]}, index=[1, 2])
+    tm.assert_frame_equal(grouped.filter(lambda x: x["A"].sum() > 10), expected)
+
+
+def test_filter_out_all_groups():
+    s = Series([1, 3, 20, 5, 22, 24, 7])
+    grouper = s.apply(lambda x: x % 2)
+    grouped = s.groupby(grouper)
+    tm.assert_series_equal(grouped.filter(lambda x: x.mean() > 1000), s[[]])
+    df = DataFrame({"A": [1, 12, 12, 1], "B": "a b c d".split()})
+    grouper = df["A"].apply(lambda x: x % 2)
+    grouped = df.groupby(grouper)
+    tm.assert_frame_equal(grouped.filter(lambda x: x["A"].sum() > 1000), df.loc[[]])
+
+
+def test_filter_out_no_groups():
+    s = Series([1, 3, 20, 5, 22, 24, 7])
+    grouper = s.apply(lambda x: x % 2)
+    grouped = s.groupby(grouper)
+    filtered = grouped.filter(lambda x: x.mean() > 0)
+    tm.assert_series_equal(filtered, s)
+    df = DataFrame({"A": [1, 12, 12, 1], "B": "a b c d".split()})
+    grouper = df["A"].apply(lambda x: x % 2)
+    grouped = df.groupby(grouper)
+    filtered = grouped.filter(lambda x: x["A"].mean() > 0)
+    tm.assert_frame_equal(filtered, df)
+
+
+def test_filter_out_all_groups_in_df():
+    # GH12768
+    df = DataFrame({"a": [1, 1, 2], "b": [1, 2, 0]})
+    res = df.groupby("a")
+    res = res.filter(lambda x: x["b"].sum() > 5, dropna=False)
+    expected = DataFrame({"a": [np.nan] * 3, "b": [np.nan] * 3})
+    tm.assert_frame_equal(expected, res)
+
+    df = DataFrame({"a": [1, 1, 2], "b": [1, 2, 0]})
+    res = df.groupby("a")
+    res = res.filter(lambda x: x["b"].sum() > 5, dropna=True)
+    expected = DataFrame({"a": [], "b": []}, dtype="int64")
+    tm.assert_frame_equal(expected, res)
+
+
+def test_filter_condition_raises():
+    def raise_if_sum_is_zero(x):
+        if x.sum() == 0:
+            raise ValueError
+        return x.sum() > 0
+
+    s = Series([-1, 0, 1, 2])
+    grouper = s.apply(lambda x: x % 2)
+    grouped = s.groupby(grouper)
+    msg = "the filter must return a boolean result"
+    with pytest.raises(TypeError, match=msg):
+        grouped.filter(raise_if_sum_is_zero)
+
+
+def test_filter_with_axis_in_groupby():
+    # issue 11041
+    index = pd.MultiIndex.from_product([range(10), [0, 1]])
+    data = DataFrame(np.arange(100).reshape(-1, 20), columns=index, dtype="int64")
+
+    msg = "DataFrame.groupby with axis=1"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        gb = data.groupby(level=0, axis=1)
+    result = gb.filter(lambda x: x.iloc[0, 0] > 10)
+    expected = data.iloc[:, 12:20]
+    tm.assert_frame_equal(result, expected)
+
+
+def test_filter_bad_shapes():
+    df = DataFrame({"A": np.arange(8), "B": list("aabbbbcc"), "C": np.arange(8)})
+    s = df["B"]
+    g_df = df.groupby("B")
+    g_s = s.groupby(s)
+
+    f = lambda x: x
+    msg = "filter function returned a DataFrame, but expected a scalar bool"
+    with pytest.raises(TypeError, match=msg):
+        g_df.filter(f)
+    msg = "the filter must return a boolean result"
+    with pytest.raises(TypeError, match=msg):
+        g_s.filter(f)
+
+    f = lambda x: x == 1
+    msg = "filter function returned a DataFrame, but expected a scalar bool"
+    with pytest.raises(TypeError, match=msg):
+        g_df.filter(f)
+    msg = "the filter must return a boolean result"
+    with pytest.raises(TypeError, match=msg):
+        g_s.filter(f)
+
+    f = lambda x: np.outer(x, x)
+    msg = "can't multiply sequence by non-int of type 'str'"
+    with pytest.raises(TypeError, match=msg):
+        g_df.filter(f)
+    msg = "the filter must return a boolean result"
+    with pytest.raises(TypeError, match=msg):
+        g_s.filter(f)
+
+
+def test_filter_nan_is_false():
+    df = DataFrame({"A": np.arange(8), "B": list("aabbbbcc"), "C": np.arange(8)})
+    s = df["B"]
+    g_df = df.groupby(df["B"])
+    g_s = s.groupby(s)
+
+    f = lambda x: np.nan
+    tm.assert_frame_equal(g_df.filter(f), df.loc[[]])
+    tm.assert_series_equal(g_s.filter(f), s[[]])
+
+
+def test_filter_pdna_is_false():
+    # in particular, dont raise in filter trying to call bool(pd.NA)
+    df = DataFrame({"A": np.arange(8), "B": list("aabbbbcc"), "C": np.arange(8)})
+    ser = df["B"]
+    g_df = df.groupby(df["B"])
+    g_s = ser.groupby(ser)
+
+    func = lambda x: pd.NA
+    res = g_df.filter(func)
+    tm.assert_frame_equal(res, df.loc[[]])
+    res = g_s.filter(func)
+    tm.assert_series_equal(res, ser[[]])
+
+
+def test_filter_against_workaround_ints():
+    # Series of ints
+    s = Series(np.random.default_rng(2).integers(0, 100, 100))
+    grouper = s.apply(lambda x: np.round(x, -1))
+    grouped = s.groupby(grouper)
+    f = lambda x: x.mean() > 10
+
+    old_way = s[grouped.transform(f).astype("bool")]
+    new_way = grouped.filter(f)
+    tm.assert_series_equal(new_way.sort_values(), old_way.sort_values())
+
+
+def test_filter_against_workaround_floats():
+    # Series of floats
+    s = 100 * Series(np.random.default_rng(2).random(100))
+    grouper = s.apply(lambda x: np.round(x, -1))
+    grouped = s.groupby(grouper)
+    f = lambda x: x.mean() > 10
+    old_way = s[grouped.transform(f).astype("bool")]
+    new_way = grouped.filter(f)
+    tm.assert_series_equal(new_way.sort_values(), old_way.sort_values())
+
+
+def test_filter_against_workaround_dataframe():
+    # Set up DataFrame of ints, floats, strings.
+    letters = np.array(list(ascii_lowercase))
+    N = 100
+    random_letters = letters.take(
+        np.random.default_rng(2).integers(0, 26, N, dtype=int)
+    )
+    df = DataFrame(
+        {
+            "ints": Series(np.random.default_rng(2).integers(0, 100, N)),
+            "floats": N / 10 * Series(np.random.default_rng(2).random(N)),
+            "letters": Series(random_letters),
+        }
+    )
+
+    # Group by ints; filter on floats.
+    grouped = df.groupby("ints")
+    old_way = df[grouped.floats.transform(lambda x: x.mean() > N / 20).astype("bool")]
+    new_way = grouped.filter(lambda x: x["floats"].mean() > N / 20)
+    tm.assert_frame_equal(new_way, old_way)
+
+    # Group by floats (rounded); filter on strings.
+    grouper = df.floats.apply(lambda x: np.round(x, -1))
+    grouped = df.groupby(grouper)
+    old_way = df[grouped.letters.transform(lambda x: len(x) < N / 10).astype("bool")]
+    new_way = grouped.filter(lambda x: len(x.letters) < N / 10)
+    tm.assert_frame_equal(new_way, old_way)
+
+    # Group by strings; filter on ints.
+    grouped = df.groupby("letters")
+    old_way = df[grouped.ints.transform(lambda x: x.mean() > N / 20).astype("bool")]
+    new_way = grouped.filter(lambda x: x["ints"].mean() > N / 20)
+    tm.assert_frame_equal(new_way, old_way)
+
+
+def test_filter_using_len():
+    # BUG GH4447
+    df = DataFrame({"A": np.arange(8), "B": list("aabbbbcc"), "C": np.arange(8)})
+    grouped = df.groupby("B")
+    actual = grouped.filter(lambda x: len(x) > 2)
+    expected = DataFrame(
+        {"A": np.arange(2, 6), "B": list("bbbb"), "C": np.arange(2, 6)},
+        index=np.arange(2, 6, dtype=np.int64),
+    )
+    tm.assert_frame_equal(actual, expected)
+
+    actual = grouped.filter(lambda x: len(x) > 4)
+    expected = df.loc[[]]
+    tm.assert_frame_equal(actual, expected)
+
+    # Series have always worked properly, but we'll test anyway.
+    s = df["B"]
+    grouped = s.groupby(s)
+    actual = grouped.filter(lambda x: len(x) > 2)
+    expected = Series(4 * ["b"], index=np.arange(2, 6, dtype=np.int64), name="B")
+    tm.assert_series_equal(actual, expected)
+
+    actual = grouped.filter(lambda x: len(x) > 4)
+    expected = s[[]]
+    tm.assert_series_equal(actual, expected)
+
+
+def test_filter_maintains_ordering():
+    # Simple case: index is sequential. #4621
+    df = DataFrame(
+        {"pid": [1, 1, 1, 2, 2, 3, 3, 3], "tag": [23, 45, 62, 24, 45, 34, 25, 62]}
+    )
+    s = df["pid"]
+    grouped = df.groupby("tag")
+    actual = grouped.filter(lambda x: len(x) > 1)
+    expected = df.iloc[[1, 2, 4, 7]]
+    tm.assert_frame_equal(actual, expected)
+
+    grouped = s.groupby(df["tag"])
+    actual = grouped.filter(lambda x: len(x) > 1)
+    expected = s.iloc[[1, 2, 4, 7]]
+    tm.assert_series_equal(actual, expected)
+
+    # Now index is sequentially decreasing.
+    df.index = np.arange(len(df) - 1, -1, -1)
+    s = df["pid"]
+    grouped = df.groupby("tag")
+    actual = grouped.filter(lambda x: len(x) > 1)
+    expected = df.iloc[[1, 2, 4, 7]]
+    tm.assert_frame_equal(actual, expected)
+
+    grouped = s.groupby(df["tag"])
+    actual = grouped.filter(lambda x: len(x) > 1)
+    expected = s.iloc[[1, 2, 4, 7]]
+    tm.assert_series_equal(actual, expected)
+
+    # Index is shuffled.
+    SHUFFLED = [4, 6, 7, 2, 1, 0, 5, 3]
+    df.index = df.index[SHUFFLED]
+    s = df["pid"]
+    grouped = df.groupby("tag")
+    actual = grouped.filter(lambda x: len(x) > 1)
+    expected = df.iloc[[1, 2, 4, 7]]
+    tm.assert_frame_equal(actual, expected)
+
+    grouped = s.groupby(df["tag"])
+    actual = grouped.filter(lambda x: len(x) > 1)
+    expected = s.iloc[[1, 2, 4, 7]]
+    tm.assert_series_equal(actual, expected)
+
+
+def test_filter_multiple_timestamp():
+    # GH 10114
+    df = DataFrame(
+        {
+            "A": np.arange(5, dtype="int64"),
+            "B": ["foo", "bar", "foo", "bar", "bar"],
+            "C": Timestamp("20130101"),
+        }
+    )
+
+    grouped = df.groupby(["B", "C"])
+
+    result = grouped["A"].filter(lambda x: True)
+    tm.assert_series_equal(df["A"], result)
+
+    result = grouped["A"].transform(len)
+    expected = Series([2, 3, 2, 3, 3], name="A")
+    tm.assert_series_equal(result, expected)
+
+    result = grouped.filter(lambda x: True)
+    tm.assert_frame_equal(df, result)
+
+    result = grouped.transform("sum")
+    expected = DataFrame({"A": [2, 8, 2, 8, 8]})
+    tm.assert_frame_equal(result, expected)
+
+    result = grouped.transform(len)
+    expected = DataFrame({"A": [2, 3, 2, 3, 3]})
+    tm.assert_frame_equal(result, expected)
+
+
+def test_filter_and_transform_with_non_unique_int_index():
+    # GH4620
+    index = [1, 1, 1, 2, 1, 1, 0, 1]
+    df = DataFrame(
+        {"pid": [1, 1, 1, 2, 2, 3, 3, 3], "tag": [23, 45, 62, 24, 45, 34, 25, 62]},
+        index=index,
+    )
+    grouped_df = df.groupby("tag")
+    ser = df["pid"]
+    grouped_ser = ser.groupby(df["tag"])
+    expected_indexes = [1, 2, 4, 7]
+
+    # Filter DataFrame
+    actual = grouped_df.filter(lambda x: len(x) > 1)
+    expected = df.iloc[expected_indexes]
+    tm.assert_frame_equal(actual, expected)
+
+    actual = grouped_df.filter(lambda x: len(x) > 1, dropna=False)
+    # Cast to avoid upcast when setting nan below
+    expected = df.copy().astype("float64")
+    expected.iloc[[0, 3, 5, 6]] = np.nan
+    tm.assert_frame_equal(actual, expected)
+
+    # Filter Series
+    actual = grouped_ser.filter(lambda x: len(x) > 1)
+    expected = ser.take(expected_indexes)
+    tm.assert_series_equal(actual, expected)
+
+    actual = grouped_ser.filter(lambda x: len(x) > 1, dropna=False)
+    expected = Series([np.nan, 1, 1, np.nan, 2, np.nan, np.nan, 3], index, name="pid")
+    # ^ made manually because this can get confusing!
+    tm.assert_series_equal(actual, expected)
+
+    # Transform Series
+    actual = grouped_ser.transform(len)
+    expected = Series([1, 2, 2, 1, 2, 1, 1, 2], index, name="pid")
+    tm.assert_series_equal(actual, expected)
+
+    # Transform (a column from) DataFrameGroupBy
+    actual = grouped_df.pid.transform(len)
+    tm.assert_series_equal(actual, expected)
+
+
+def test_filter_and_transform_with_multiple_non_unique_int_index():
+    # GH4620
+    index = [1, 1, 1, 2, 0, 0, 0, 1]
+    df = DataFrame(
+        {"pid": [1, 1, 1, 2, 2, 3, 3, 3], "tag": [23, 45, 62, 24, 45, 34, 25, 62]},
+        index=index,
+    )
+    grouped_df = df.groupby("tag")
+    ser = df["pid"]
+    grouped_ser = ser.groupby(df["tag"])
+    expected_indexes = [1, 2, 4, 7]
+
+    # Filter DataFrame
+    actual = grouped_df.filter(lambda x: len(x) > 1)
+    expected = df.iloc[expected_indexes]
+    tm.assert_frame_equal(actual, expected)
+
+    actual = grouped_df.filter(lambda x: len(x) > 1, dropna=False)
+    # Cast to avoid upcast when setting nan below
+    expected = df.copy().astype("float64")
+    expected.iloc[[0, 3, 5, 6]] = np.nan
+    tm.assert_frame_equal(actual, expected)
+
+    # Filter Series
+    actual = grouped_ser.filter(lambda x: len(x) > 1)
+    expected = ser.take(expected_indexes)
+    tm.assert_series_equal(actual, expected)
+
+    actual = grouped_ser.filter(lambda x: len(x) > 1, dropna=False)
+    expected = Series([np.nan, 1, 1, np.nan, 2, np.nan, np.nan, 3], index, name="pid")
+    # ^ made manually because this can get confusing!
+    tm.assert_series_equal(actual, expected)
+
+    # Transform Series
+    actual = grouped_ser.transform(len)
+    expected = Series([1, 2, 2, 1, 2, 1, 1, 2], index, name="pid")
+    tm.assert_series_equal(actual, expected)
+
+    # Transform (a column from) DataFrameGroupBy
+    actual = grouped_df.pid.transform(len)
+    tm.assert_series_equal(actual, expected)
+
+
+def test_filter_and_transform_with_non_unique_float_index():
+    # GH4620
+    index = np.array([1, 1, 1, 2, 1, 1, 0, 1], dtype=float)
+    df = DataFrame(
+        {"pid": [1, 1, 1, 2, 2, 3, 3, 3], "tag": [23, 45, 62, 24, 45, 34, 25, 62]},
+        index=index,
+    )
+    grouped_df = df.groupby("tag")
+    ser = df["pid"]
+    grouped_ser = ser.groupby(df["tag"])
+    expected_indexes = [1, 2, 4, 7]
+
+    # Filter DataFrame
+    actual = grouped_df.filter(lambda x: len(x) > 1)
+    expected = df.iloc[expected_indexes]
+    tm.assert_frame_equal(actual, expected)
+
+    actual = grouped_df.filter(lambda x: len(x) > 1, dropna=False)
+    # Cast to avoid upcast when setting nan below
+    expected = df.copy().astype("float64")
+    expected.iloc[[0, 3, 5, 6]] = np.nan
+    tm.assert_frame_equal(actual, expected)
+
+    # Filter Series
+    actual = grouped_ser.filter(lambda x: len(x) > 1)
+    expected = ser.take(expected_indexes)
+    tm.assert_series_equal(actual, expected)
+
+    actual = grouped_ser.filter(lambda x: len(x) > 1, dropna=False)
+    expected = Series([np.nan, 1, 1, np.nan, 2, np.nan, np.nan, 3], index, name="pid")
+    # ^ made manually because this can get confusing!
+    tm.assert_series_equal(actual, expected)
+
+    # Transform Series
+    actual = grouped_ser.transform(len)
+    expected = Series([1, 2, 2, 1, 2, 1, 1, 2], index, name="pid")
+    tm.assert_series_equal(actual, expected)
+
+    # Transform (a column from) DataFrameGroupBy
+    actual = grouped_df.pid.transform(len)
+    tm.assert_series_equal(actual, expected)
+
+
+def test_filter_and_transform_with_non_unique_timestamp_index():
+    # GH4620
+    t0 = Timestamp("2013-09-30 00:05:00")
+    t1 = Timestamp("2013-10-30 00:05:00")
+    t2 = Timestamp("2013-11-30 00:05:00")
+    index = [t1, t1, t1, t2, t1, t1, t0, t1]
+    df = DataFrame(
+        {"pid": [1, 1, 1, 2, 2, 3, 3, 3], "tag": [23, 45, 62, 24, 45, 34, 25, 62]},
+        index=index,
+    )
+    grouped_df = df.groupby("tag")
+    ser = df["pid"]
+    grouped_ser = ser.groupby(df["tag"])
+    expected_indexes = [1, 2, 4, 7]
+
+    # Filter DataFrame
+    actual = grouped_df.filter(lambda x: len(x) > 1)
+    expected = df.iloc[expected_indexes]
+    tm.assert_frame_equal(actual, expected)
+
+    actual = grouped_df.filter(lambda x: len(x) > 1, dropna=False)
+    # Cast to avoid upcast when setting nan below
+    expected = df.copy().astype("float64")
+    expected.iloc[[0, 3, 5, 6]] = np.nan
+    tm.assert_frame_equal(actual, expected)
+
+    # Filter Series
+    actual = grouped_ser.filter(lambda x: len(x) > 1)
+    expected = ser.take(expected_indexes)
+    tm.assert_series_equal(actual, expected)
+
+    actual = grouped_ser.filter(lambda x: len(x) > 1, dropna=False)
+    expected = Series([np.nan, 1, 1, np.nan, 2, np.nan, np.nan, 3], index, name="pid")
+    # ^ made manually because this can get confusing!
+    tm.assert_series_equal(actual, expected)
+
+    # Transform Series
+    actual = grouped_ser.transform(len)
+    expected = Series([1, 2, 2, 1, 2, 1, 1, 2], index, name="pid")
+    tm.assert_series_equal(actual, expected)
+
+    # Transform (a column from) DataFrameGroupBy
+    actual = grouped_df.pid.transform(len)
+    tm.assert_series_equal(actual, expected)
+
+
+def test_filter_and_transform_with_non_unique_string_index():
+    # GH4620
+    index = list("bbbcbbab")
+    df = DataFrame(
+        {"pid": [1, 1, 1, 2, 2, 3, 3, 3], "tag": [23, 45, 62, 24, 45, 34, 25, 62]},
+        index=index,
+    )
+    grouped_df = df.groupby("tag")
+    ser = df["pid"]
+    grouped_ser = ser.groupby(df["tag"])
+    expected_indexes = [1, 2, 4, 7]
+
+    # Filter DataFrame
+    actual = grouped_df.filter(lambda x: len(x) > 1)
+    expected = df.iloc[expected_indexes]
+    tm.assert_frame_equal(actual, expected)
+
+    actual = grouped_df.filter(lambda x: len(x) > 1, dropna=False)
+    # Cast to avoid upcast when setting nan below
+    expected = df.copy().astype("float64")
+    expected.iloc[[0, 3, 5, 6]] = np.nan
+    tm.assert_frame_equal(actual, expected)
+
+    # Filter Series
+    actual = grouped_ser.filter(lambda x: len(x) > 1)
+    expected = ser.take(expected_indexes)
+    tm.assert_series_equal(actual, expected)
+
+    actual = grouped_ser.filter(lambda x: len(x) > 1, dropna=False)
+    expected = Series([np.nan, 1, 1, np.nan, 2, np.nan, np.nan, 3], index, name="pid")
+    # ^ made manually because this can get confusing!
+    tm.assert_series_equal(actual, expected)
+
+    # Transform Series
+    actual = grouped_ser.transform(len)
+    expected = Series([1, 2, 2, 1, 2, 1, 1, 2], index, name="pid")
+    tm.assert_series_equal(actual, expected)
+
+    # Transform (a column from) DataFrameGroupBy
+    actual = grouped_df.pid.transform(len)
+    tm.assert_series_equal(actual, expected)
+
+
+def test_filter_has_access_to_grouped_cols():
+    df = DataFrame([[1, 2], [1, 3], [5, 6]], columns=["A", "B"])
+    g = df.groupby("A")
+    # previously didn't have access to col A #????
+    filt = g.filter(lambda x: x["A"].sum() == 2)
+    tm.assert_frame_equal(filt, df.iloc[[0, 1]])
+
+
+def test_filter_enforces_scalarness():
+    df = DataFrame(
+        [
+            ["best", "a", "x"],
+            ["worst", "b", "y"],
+            ["best", "c", "x"],
+            ["best", "d", "y"],
+            ["worst", "d", "y"],
+            ["worst", "d", "y"],
+            ["best", "d", "z"],
+        ],
+        columns=["a", "b", "c"],
+    )
+    with pytest.raises(TypeError, match="filter function returned a.*"):
+        df.groupby("c").filter(lambda g: g["a"] == "best")
+
+
+def test_filter_non_bool_raises():
+    df = DataFrame(
+        [
+            ["best", "a", 1],
+            ["worst", "b", 1],
+            ["best", "c", 1],
+            ["best", "d", 1],
+            ["worst", "d", 1],
+            ["worst", "d", 1],
+            ["best", "d", 1],
+        ],
+        columns=["a", "b", "c"],
+    )
+    with pytest.raises(TypeError, match="filter function returned a.*"):
+        df.groupby("a").filter(lambda g: g.c.mean())
+
+
+def test_filter_dropna_with_empty_groups():
+    # GH 10780
+    data = Series(np.random.default_rng(2).random(9), index=np.repeat([1, 2, 3], 3))
+    grouped = data.groupby(level=0)
+    result_false = grouped.filter(lambda x: x.mean() > 1, dropna=False)
+    expected_false = Series([np.nan] * 9, index=np.repeat([1, 2, 3], 3))
+    tm.assert_series_equal(result_false, expected_false)
+
+    result_true = grouped.filter(lambda x: x.mean() > 1, dropna=True)
+    expected_true = Series(index=pd.Index([], dtype=int), dtype=np.float64)
+    tm.assert_series_equal(result_true, expected_true)
+
+
+def test_filter_consistent_result_before_after_agg_func():
+    # GH 17091
+    df = DataFrame({"data": range(6), "key": list("ABCABC")})
+    grouper = df.groupby("key")
+    result = grouper.filter(lambda x: True)
+    expected = DataFrame({"data": range(6), "key": list("ABCABC")})
+    tm.assert_frame_equal(result, expected)
+
+    grouper.sum()
+    result = grouper.filter(lambda x: True)
+    tm.assert_frame_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_groupby_dropna.py

@@ -0,0 +1,696 @@
+import numpy as np
+import pytest
+
+from pandas.compat.pyarrow import pa_version_under10p1
+
+from pandas.core.dtypes.missing import na_value_for_dtype
+
+import pandas as pd
+import pandas._testing as tm
+from pandas.tests.groupby import get_groupby_method_args
+
+
+@pytest.mark.parametrize(
+    "dropna, tuples, outputs",
+    [
+        (
+            True,
+            [["A", "B"], ["B", "A"]],
+            {"c": [13.0, 123.23], "d": [13.0, 123.0], "e": [13.0, 1.0]},
+        ),
+        (
+            False,
+            [["A", "B"], ["A", np.nan], ["B", "A"]],
+            {
+                "c": [13.0, 12.3, 123.23],
+                "d": [13.0, 233.0, 123.0],
+                "e": [13.0, 12.0, 1.0],
+            },
+        ),
+    ],
+)
+def test_groupby_dropna_multi_index_dataframe_nan_in_one_group(
+    dropna, tuples, outputs, nulls_fixture
+):
+    # GH 3729 this is to test that NA is in one group
+    df_list = [
+        ["A", "B", 12, 12, 12],
+        ["A", nulls_fixture, 12.3, 233.0, 12],
+        ["B", "A", 123.23, 123, 1],
+        ["A", "B", 1, 1, 1.0],
+    ]
+    df = pd.DataFrame(df_list, columns=["a", "b", "c", "d", "e"])
+    grouped = df.groupby(["a", "b"], dropna=dropna).sum()
+
+    mi = pd.MultiIndex.from_tuples(tuples, names=list("ab"))
+
+    # Since right now, by default MI will drop NA from levels when we create MI
+    # via `from_*`, so we need to add NA for level manually afterwards.
+    if not dropna:
+        mi = mi.set_levels(["A", "B", np.nan], level="b")
+    expected = pd.DataFrame(outputs, index=mi)
+
+    tm.assert_frame_equal(grouped, expected)
+
+
+@pytest.mark.parametrize(
+    "dropna, tuples, outputs",
+    [
+        (
+            True,
+            [["A", "B"], ["B", "A"]],
+            {"c": [12.0, 123.23], "d": [12.0, 123.0], "e": [12.0, 1.0]},
+        ),
+        (
+            False,
+            [["A", "B"], ["A", np.nan], ["B", "A"], [np.nan, "B"]],
+            {
+                "c": [12.0, 13.3, 123.23, 1.0],
+                "d": [12.0, 234.0, 123.0, 1.0],
+                "e": [12.0, 13.0, 1.0, 1.0],
+            },
+        ),
+    ],
+)
+def test_groupby_dropna_multi_index_dataframe_nan_in_two_groups(
+    dropna, tuples, outputs, nulls_fixture, nulls_fixture2
+):
+    # GH 3729 this is to test that NA in different groups with different representations
+    df_list = [
+        ["A", "B", 12, 12, 12],
+        ["A", nulls_fixture, 12.3, 233.0, 12],
+        ["B", "A", 123.23, 123, 1],
+        [nulls_fixture2, "B", 1, 1, 1.0],
+        ["A", nulls_fixture2, 1, 1, 1.0],
+    ]
+    df = pd.DataFrame(df_list, columns=["a", "b", "c", "d", "e"])
+    grouped = df.groupby(["a", "b"], dropna=dropna).sum()
+
+    mi = pd.MultiIndex.from_tuples(tuples, names=list("ab"))
+
+    # Since right now, by default MI will drop NA from levels when we create MI
+    # via `from_*`, so we need to add NA for level manually afterwards.
+    if not dropna:
+        mi = mi.set_levels([["A", "B", np.nan], ["A", "B", np.nan]])
+    expected = pd.DataFrame(outputs, index=mi)
+
+    tm.assert_frame_equal(grouped, expected)
+
+
+@pytest.mark.parametrize(
+    "dropna, idx, outputs",
+    [
+        (True, ["A", "B"], {"b": [123.23, 13.0], "c": [123.0, 13.0], "d": [1.0, 13.0]}),
+        (
+            False,
+            ["A", "B", np.nan],
+            {
+                "b": [123.23, 13.0, 12.3],
+                "c": [123.0, 13.0, 233.0],
+                "d": [1.0, 13.0, 12.0],
+            },
+        ),
+    ],
+)
+def test_groupby_dropna_normal_index_dataframe(dropna, idx, outputs):
+    # GH 3729
+    df_list = [
+        ["B", 12, 12, 12],
+        [None, 12.3, 233.0, 12],
+        ["A", 123.23, 123, 1],
+        ["B", 1, 1, 1.0],
+    ]
+    df = pd.DataFrame(df_list, columns=["a", "b", "c", "d"])
+    grouped = df.groupby("a", dropna=dropna).sum()
+
+    expected = pd.DataFrame(outputs, index=pd.Index(idx, dtype="object", name="a"))
+
+    tm.assert_frame_equal(grouped, expected)
+
+
+@pytest.mark.parametrize(
+    "dropna, idx, expected",
+    [
+        (True, ["a", "a", "b", np.nan], pd.Series([3, 3], index=["a", "b"])),
+        (
+            False,
+            ["a", "a", "b", np.nan],
+            pd.Series([3, 3, 3], index=["a", "b", np.nan]),
+        ),
+    ],
+)
+def test_groupby_dropna_series_level(dropna, idx, expected):
+    ser = pd.Series([1, 2, 3, 3], index=idx)
+
+    result = ser.groupby(level=0, dropna=dropna).sum()
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "dropna, expected",
+    [
+        (True, pd.Series([210.0, 350.0], index=["a", "b"], name="Max Speed")),
+        (
+            False,
+            pd.Series([210.0, 350.0, 20.0], index=["a", "b", np.nan], name="Max Speed"),
+        ),
+    ],
+)
+def test_groupby_dropna_series_by(dropna, expected):
+    ser = pd.Series(
+        [390.0, 350.0, 30.0, 20.0],
+        index=["Falcon", "Falcon", "Parrot", "Parrot"],
+        name="Max Speed",
+    )
+
+    result = ser.groupby(["a", "b", "a", np.nan], dropna=dropna).mean()
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize("dropna", (False, True))
+def test_grouper_dropna_propagation(dropna):
+    # GH 36604
+    df = pd.DataFrame({"A": [0, 0, 1, None], "B": [1, 2, 3, None]})
+    gb = df.groupby("A", dropna=dropna)
+    assert gb._grouper.dropna == dropna
+
+
+@pytest.mark.parametrize(
+    "index",
+    [
+        pd.RangeIndex(0, 4),
+        list("abcd"),
+        pd.MultiIndex.from_product([(1, 2), ("R", "B")], names=["num", "col"]),
+    ],
+)
+def test_groupby_dataframe_slice_then_transform(dropna, index):
+    # GH35014 & GH35612
+    expected_data = {"B": [2, 2, 1, np.nan if dropna else 1]}
+
+    df = pd.DataFrame({"A": [0, 0, 1, None], "B": [1, 2, 3, None]}, index=index)
+    gb = df.groupby("A", dropna=dropna)
+
+    result = gb.transform(len)
+    expected = pd.DataFrame(expected_data, index=index)
+    tm.assert_frame_equal(result, expected)
+
+    result = gb[["B"]].transform(len)
+    expected = pd.DataFrame(expected_data, index=index)
+    tm.assert_frame_equal(result, expected)
+
+    result = gb["B"].transform(len)
+    expected = pd.Series(expected_data["B"], index=index, name="B")
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "dropna, tuples, outputs",
+    [
+        (
+            True,
+            [["A", "B"], ["B", "A"]],
+            {"c": [13.0, 123.23], "d": [12.0, 123.0], "e": [1.0, 1.0]},
+        ),
+        (
+            False,
+            [["A", "B"], ["A", np.nan], ["B", "A"]],
+            {
+                "c": [13.0, 12.3, 123.23],
+                "d": [12.0, 233.0, 123.0],
+                "e": [1.0, 12.0, 1.0],
+            },
+        ),
+    ],
+)
+def test_groupby_dropna_multi_index_dataframe_agg(dropna, tuples, outputs):
+    # GH 3729
+    df_list = [
+        ["A", "B", 12, 12, 12],
+        ["A", None, 12.3, 233.0, 12],
+        ["B", "A", 123.23, 123, 1],
+        ["A", "B", 1, 1, 1.0],
+    ]
+    df = pd.DataFrame(df_list, columns=["a", "b", "c", "d", "e"])
+    agg_dict = {"c": "sum", "d": "max", "e": "min"}
+    grouped = df.groupby(["a", "b"], dropna=dropna).agg(agg_dict)
+
+    mi = pd.MultiIndex.from_tuples(tuples, names=list("ab"))
+
+    # Since right now, by default MI will drop NA from levels when we create MI
+    # via `from_*`, so we need to add NA for level manually afterwards.
+    if not dropna:
+        mi = mi.set_levels(["A", "B", np.nan], level="b")
+    expected = pd.DataFrame(outputs, index=mi)
+
+    tm.assert_frame_equal(grouped, expected)
+
+
+@pytest.mark.arm_slow
+@pytest.mark.parametrize(
+    "datetime1, datetime2",
+    [
+        (pd.Timestamp("2020-01-01"), pd.Timestamp("2020-02-01")),
+        (pd.Timedelta("-2 days"), pd.Timedelta("-1 days")),
+        (pd.Period("2020-01-01"), pd.Period("2020-02-01")),
+    ],
+)
+@pytest.mark.parametrize("dropna, values", [(True, [12, 3]), (False, [12, 3, 6])])
+def test_groupby_dropna_datetime_like_data(
+    dropna, values, datetime1, datetime2, unique_nulls_fixture, unique_nulls_fixture2
+):
+    # 3729
+    df = pd.DataFrame(
+        {
+            "values": [1, 2, 3, 4, 5, 6],
+            "dt": [
+                datetime1,
+                unique_nulls_fixture,
+                datetime2,
+                unique_nulls_fixture2,
+                datetime1,
+                datetime1,
+            ],
+        }
+    )
+
+    if dropna:
+        indexes = [datetime1, datetime2]
+    else:
+        indexes = [datetime1, datetime2, np.nan]
+
+    grouped = df.groupby("dt", dropna=dropna).agg({"values": "sum"})
+    expected = pd.DataFrame({"values": values}, index=pd.Index(indexes, name="dt"))
+
+    tm.assert_frame_equal(grouped, expected)
+
+
+@pytest.mark.parametrize(
+    "dropna, data, selected_data, levels",
+    [
+        pytest.param(
+            False,
+            {"groups": ["a", "a", "b", np.nan], "values": [10, 10, 20, 30]},
+            {"values": [0, 1, 0, 0]},
+            ["a", "b", np.nan],
+            id="dropna_false_has_nan",
+        ),
+        pytest.param(
+            True,
+            {"groups": ["a", "a", "b", np.nan], "values": [10, 10, 20, 30]},
+            {"values": [0, 1, 0]},
+            None,
+            id="dropna_true_has_nan",
+        ),
+        pytest.param(
+            # no nan in "groups"; dropna=True|False should be same.
+            False,
+            {"groups": ["a", "a", "b", "c"], "values": [10, 10, 20, 30]},
+            {"values": [0, 1, 0, 0]},
+            None,
+            id="dropna_false_no_nan",
+        ),
+        pytest.param(
+            # no nan in "groups"; dropna=True|False should be same.
+            True,
+            {"groups": ["a", "a", "b", "c"], "values": [10, 10, 20, 30]},
+            {"values": [0, 1, 0, 0]},
+            None,
+            id="dropna_true_no_nan",
+        ),
+    ],
+)
+def test_groupby_apply_with_dropna_for_multi_index(dropna, data, selected_data, levels):
+    # GH 35889
+
+    df = pd.DataFrame(data)
+    gb = df.groupby("groups", dropna=dropna)
+    msg = "DataFrameGroupBy.apply operated on the grouping columns"
+    with tm.assert_produces_warning(DeprecationWarning, match=msg):
+        result = gb.apply(lambda grp: pd.DataFrame({"values": range(len(grp))}))
+
+    mi_tuples = tuple(zip(data["groups"], selected_data["values"]))
+    mi = pd.MultiIndex.from_tuples(mi_tuples, names=["groups", None])
+    # Since right now, by default MI will drop NA from levels when we create MI
+    # via `from_*`, so we need to add NA for level manually afterwards.
+    if not dropna and levels:
+        mi = mi.set_levels(levels, level="groups")
+
+    expected = pd.DataFrame(selected_data, index=mi)
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("input_index", [None, ["a"], ["a", "b"]])
+@pytest.mark.parametrize("keys", [["a"], ["a", "b"]])
+@pytest.mark.parametrize("series", [True, False])
+def test_groupby_dropna_with_multiindex_input(input_index, keys, series):
+    # GH#46783
+    obj = pd.DataFrame(
+        {
+            "a": [1, np.nan],
+            "b": [1, 1],
+            "c": [2, 3],
+        }
+    )
+
+    expected = obj.set_index(keys)
+    if series:
+        expected = expected["c"]
+    elif input_index == ["a", "b"] and keys == ["a"]:
+        # Column b should not be aggregated
+        expected = expected[["c"]]
+
+    if input_index is not None:
+        obj = obj.set_index(input_index)
+    gb = obj.groupby(keys, dropna=False)
+    if series:
+        gb = gb["c"]
+    result = gb.sum()
+
+    tm.assert_equal(result, expected)
+
+
+def test_groupby_nan_included():
+    # GH 35646
+    data = {"group": ["g1", np.nan, "g1", "g2", np.nan], "B": [0, 1, 2, 3, 4]}
+    df = pd.DataFrame(data)
+    grouped = df.groupby("group", dropna=False)
+    result = grouped.indices
+    dtype = np.intp
+    expected = {
+        "g1": np.array([0, 2], dtype=dtype),
+        "g2": np.array([3], dtype=dtype),
+        np.nan: np.array([1, 4], dtype=dtype),
+    }
+    for result_values, expected_values in zip(result.values(), expected.values()):
+        tm.assert_numpy_array_equal(result_values, expected_values)
+    assert np.isnan(list(result.keys())[2])
+    assert list(result.keys())[0:2] == ["g1", "g2"]
+
+
+def test_groupby_drop_nan_with_multi_index():
+    # GH 39895
+    df = pd.DataFrame([[np.nan, 0, 1]], columns=["a", "b", "c"])
+    df = df.set_index(["a", "b"])
+    result = df.groupby(["a", "b"], dropna=False).first()
+    expected = df
+    tm.assert_frame_equal(result, expected)
+
+
+# sequence_index enumerates all strings made up of x, y, z of length 4
+@pytest.mark.parametrize("sequence_index", range(3**4))
+@pytest.mark.parametrize(
+    "dtype",
+    [
+        None,
+        "UInt8",
+        "Int8",
+        "UInt16",
+        "Int16",
+        "UInt32",
+        "Int32",
+        "UInt64",
+        "Int64",
+        "Float32",
+        "Int64",
+        "Float64",
+        "category",
+        "string",
+        pytest.param(
+            "string[pyarrow]",
+            marks=pytest.mark.skipif(
+                pa_version_under10p1, reason="pyarrow is not installed"
+            ),
+        ),
+        "datetime64[ns]",
+        "period[d]",
+        "Sparse[float]",
+    ],
+)
+@pytest.mark.parametrize("test_series", [True, False])
+def test_no_sort_keep_na(sequence_index, dtype, test_series, as_index):
+    # GH#46584, GH#48794
+
+    # Convert sequence_index into a string sequence, e.g. 5 becomes "xxyz"
+    # This sequence is used for the grouper.
+    sequence = "".join(
+        [{0: "x", 1: "y", 2: "z"}[sequence_index // (3**k) % 3] for k in range(4)]
+    )
+
+    # Unique values to use for grouper, depends on dtype
+    if dtype in ("string", "string[pyarrow]"):
+        uniques = {"x": "x", "y": "y", "z": pd.NA}
+    elif dtype in ("datetime64[ns]", "period[d]"):
+        uniques = {"x": "2016-01-01", "y": "2017-01-01", "z": pd.NA}
+    else:
+        uniques = {"x": 1, "y": 2, "z": np.nan}
+
+    df = pd.DataFrame(
+        {
+            "key": pd.Series([uniques[label] for label in sequence], dtype=dtype),
+            "a": [0, 1, 2, 3],
+        }
+    )
+    gb = df.groupby("key", dropna=False, sort=False, as_index=as_index, observed=False)
+    if test_series:
+        gb = gb["a"]
+    result = gb.sum()
+
+    # Manually compute the groupby sum, use the labels "x", "y", and "z" to avoid
+    # issues with hashing np.nan
+    summed = {}
+    for idx, label in enumerate(sequence):
+        summed[label] = summed.get(label, 0) + idx
+    if dtype == "category":
+        index = pd.CategoricalIndex(
+            [uniques[e] for e in summed],
+            df["key"].cat.categories,
+            name="key",
+        )
+    elif isinstance(dtype, str) and dtype.startswith("Sparse"):
+        index = pd.Index(
+            pd.array([uniques[label] for label in summed], dtype=dtype), name="key"
+        )
+    else:
+        index = pd.Index([uniques[label] for label in summed], dtype=dtype, name="key")
+    expected = pd.Series(summed.values(), index=index, name="a", dtype=None)
+    if not test_series:
+        expected = expected.to_frame()
+    if not as_index:
+        expected = expected.reset_index()
+        if dtype is not None and dtype.startswith("Sparse"):
+            expected["key"] = expected["key"].astype(dtype)
+
+    tm.assert_equal(result, expected)
+
+
+@pytest.mark.parametrize("test_series", [True, False])
+@pytest.mark.parametrize("dtype", [object, None])
+def test_null_is_null_for_dtype(
+    sort, dtype, nulls_fixture, nulls_fixture2, test_series
+):
+    # GH#48506 - groups should always result in using the null for the dtype
+    df = pd.DataFrame({"a": [1, 2]})
+    groups = pd.Series([nulls_fixture, nulls_fixture2], dtype=dtype)
+    obj = df["a"] if test_series else df
+    gb = obj.groupby(groups, dropna=False, sort=sort)
+    result = gb.sum()
+    index = pd.Index([na_value_for_dtype(groups.dtype)])
+    expected = pd.DataFrame({"a": [3]}, index=index)
+    if test_series:
+        tm.assert_series_equal(result, expected["a"])
+    else:
+        tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("index_kind", ["range", "single", "multi"])
+def test_categorical_reducers(reduction_func, observed, sort, as_index, index_kind):
+    # Ensure there is at least one null value by appending to the end
+    values = np.append(np.random.default_rng(2).choice([1, 2, None], size=19), None)
+    df = pd.DataFrame(
+        {"x": pd.Categorical(values, categories=[1, 2, 3]), "y": range(20)}
+    )
+
+    # Strategy: Compare to dropna=True by filling null values with a new code
+    df_filled = df.copy()
+    df_filled["x"] = pd.Categorical(values, categories=[1, 2, 3, 4]).fillna(4)
+
+    if index_kind == "range":
+        keys = ["x"]
+    elif index_kind == "single":
+        keys = ["x"]
+        df = df.set_index("x")
+        df_filled = df_filled.set_index("x")
+    else:
+        keys = ["x", "x2"]
+        df["x2"] = df["x"]
+        df = df.set_index(["x", "x2"])
+        df_filled["x2"] = df_filled["x"]
+        df_filled = df_filled.set_index(["x", "x2"])
+    args = get_groupby_method_args(reduction_func, df)
+    args_filled = get_groupby_method_args(reduction_func, df_filled)
+    if reduction_func == "corrwith" and index_kind == "range":
+        # Don't include the grouping columns so we can call reset_index
+        args = (args[0].drop(columns=keys),)
+        args_filled = (args_filled[0].drop(columns=keys),)
+
+    gb_keepna = df.groupby(
+        keys, dropna=False, observed=observed, sort=sort, as_index=as_index
+    )
+
+    if not observed and reduction_func in ["idxmin", "idxmax"]:
+        with pytest.raises(
+            ValueError, match="empty group due to unobserved categories"
+        ):
+            getattr(gb_keepna, reduction_func)(*args)
+        return
+
+    gb_filled = df_filled.groupby(keys, observed=observed, sort=sort, as_index=True)
+    expected = getattr(gb_filled, reduction_func)(*args_filled).reset_index()
+    expected["x"] = expected["x"].cat.remove_categories([4])
+    if index_kind == "multi":
+        expected["x2"] = expected["x2"].cat.remove_categories([4])
+    if as_index:
+        if index_kind == "multi":
+            expected = expected.set_index(["x", "x2"])
+        else:
+            expected = expected.set_index("x")
+    elif index_kind != "range" and reduction_func != "size":
+        # size, unlike other methods, has the desired behavior in GH#49519
+        expected = expected.drop(columns="x")
+        if index_kind == "multi":
+            expected = expected.drop(columns="x2")
+    if reduction_func in ("idxmax", "idxmin") and index_kind != "range":
+        # expected was computed with a RangeIndex; need to translate to index values
+        values = expected["y"].values.tolist()
+        if index_kind == "single":
+            values = [np.nan if e == 4 else e for e in values]
+            expected["y"] = pd.Categorical(values, categories=[1, 2, 3])
+        else:
+            values = [(np.nan, np.nan) if e == (4, 4) else e for e in values]
+            expected["y"] = values
+    if reduction_func == "size":
+        # size, unlike other methods, has the desired behavior in GH#49519
+        expected = expected.rename(columns={0: "size"})
+        if as_index:
+            expected = expected["size"].rename(None)
+
+    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)
+
+
+def test_categorical_transformers(
+    request, transformation_func, observed, sort, as_index
+):
+    # GH#36327
+    if transformation_func == "fillna":
+        msg = "GH#49651 fillna may incorrectly reorders results when dropna=False"
+        request.applymarker(pytest.mark.xfail(reason=msg, strict=False))
+
+    values = np.append(np.random.default_rng(2).choice([1, 2, None], size=19), None)
+    df = pd.DataFrame(
+        {"x": pd.Categorical(values, categories=[1, 2, 3]), "y": range(20)}
+    )
+    args = get_groupby_method_args(transformation_func, df)
+
+    # Compute result for null group
+    null_group_values = df[df["x"].isnull()]["y"]
+    if transformation_func == "cumcount":
+        null_group_data = list(range(len(null_group_values)))
+    elif transformation_func == "ngroup":
+        if sort:
+            if observed:
+                na_group = df["x"].nunique(dropna=False) - 1
+            else:
+                # TODO: Should this be 3?
+                na_group = df["x"].nunique(dropna=False) - 1
+        else:
+            na_group = df.iloc[: null_group_values.index[0]]["x"].nunique()
+        null_group_data = len(null_group_values) * [na_group]
+    else:
+        null_group_data = getattr(null_group_values, transformation_func)(*args)
+    null_group_result = pd.DataFrame({"y": null_group_data})
+
+    gb_keepna = df.groupby(
+        "x", dropna=False, observed=observed, sort=sort, as_index=as_index
+    )
+    gb_dropna = df.groupby("x", dropna=True, observed=observed, sort=sort)
+
+    msg = "The default fill_method='ffill' in DataFrameGroupBy.pct_change is deprecated"
+    if transformation_func == "pct_change":
+        with tm.assert_produces_warning(FutureWarning, match=msg):
+            result = getattr(gb_keepna, "pct_change")(*args)
+    else:
+        result = getattr(gb_keepna, transformation_func)(*args)
+    expected = getattr(gb_dropna, transformation_func)(*args)
+
+    for iloc, value in zip(
+        df[df["x"].isnull()].index.tolist(), null_group_result.values.ravel()
+    ):
+        if expected.ndim == 1:
+            expected.iloc[iloc] = value
+        else:
+            expected.iloc[iloc, 0] = value
+    if transformation_func == "ngroup":
+        expected[df["x"].notnull() & expected.ge(na_group)] += 1
+    if transformation_func not in ("rank", "diff", "pct_change", "shift"):
+        expected = expected.astype("int64")
+
+    tm.assert_equal(result, expected)
+
+
+@pytest.mark.parametrize("method", ["head", "tail"])
+def test_categorical_head_tail(method, observed, sort, as_index):
+    # GH#36327
+    values = np.random.default_rng(2).choice([1, 2, None], 30)
+    df = pd.DataFrame(
+        {"x": pd.Categorical(values, categories=[1, 2, 3]), "y": range(len(values))}
+    )
+    gb = df.groupby("x", dropna=False, observed=observed, sort=sort, as_index=as_index)
+    result = getattr(gb, method)()
+
+    if method == "tail":
+        values = values[::-1]
+    # Take the top 5 values from each group
+    mask = (
+        ((values == 1) & ((values == 1).cumsum() <= 5))
+        | ((values == 2) & ((values == 2).cumsum() <= 5))
+        # flake8 doesn't like the vectorized check for None, thinks we should use `is`
+        | ((values == None) & ((values == None).cumsum() <= 5))  # noqa: E711
+    )
+    if method == "tail":
+        mask = mask[::-1]
+    expected = df[mask]
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_categorical_agg():
+    # GH#36327
+    values = np.random.default_rng(2).choice([1, 2, None], 30)
+    df = pd.DataFrame(
+        {"x": pd.Categorical(values, categories=[1, 2, 3]), "y": range(len(values))}
+    )
+    gb = df.groupby("x", dropna=False, observed=False)
+    result = gb.agg(lambda x: x.sum())
+    expected = gb.sum()
+    tm.assert_frame_equal(result, expected)
+
+
+def test_categorical_transform():
+    # GH#36327
+    values = np.random.default_rng(2).choice([1, 2, None], 30)
+    df = pd.DataFrame(
+        {"x": pd.Categorical(values, categories=[1, 2, 3]), "y": range(len(values))}
+    )
+    gb = df.groupby("x", dropna=False, observed=False)
+    result = gb.transform(lambda x: x.sum())
+    expected = gb.transform("sum")
+    tm.assert_frame_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_groupby_subclass.py

@@ -0,0 +1,135 @@
+from datetime import datetime
+
+import numpy as np
+import pytest
+
+from pandas import (
+    DataFrame,
+    Index,
+    Series,
+)
+import pandas._testing as tm
+from pandas.tests.groupby import get_groupby_method_args
+
+pytestmark = pytest.mark.filterwarnings(
+    "ignore:Passing a BlockManager|Passing a SingleBlockManager:DeprecationWarning"
+)
+
+
+@pytest.mark.parametrize(
+    "obj",
+    [
+        tm.SubclassedDataFrame({"A": np.arange(0, 10)}),
+        tm.SubclassedSeries(np.arange(0, 10), name="A"),
+    ],
+)
+def test_groupby_preserves_subclass(obj, groupby_func):
+    # GH28330 -- preserve subclass through groupby operations
+
+    if isinstance(obj, Series) and groupby_func in {"corrwith"}:
+        pytest.skip(f"Not applicable for Series and {groupby_func}")
+
+    grouped = obj.groupby(np.arange(0, 10))
+
+    # Groups should preserve subclass type
+    assert isinstance(grouped.get_group(0), type(obj))
+
+    args = get_groupby_method_args(groupby_func, obj)
+
+    warn = FutureWarning if groupby_func == "fillna" else None
+    msg = f"{type(grouped).__name__}.fillna is deprecated"
+    with tm.assert_produces_warning(warn, match=msg, raise_on_extra_warnings=False):
+        result1 = getattr(grouped, groupby_func)(*args)
+    with tm.assert_produces_warning(warn, match=msg, raise_on_extra_warnings=False):
+        result2 = grouped.agg(groupby_func, *args)
+
+    # Reduction or transformation kernels should preserve type
+    slices = {"ngroup", "cumcount", "size"}
+    if isinstance(obj, DataFrame) and groupby_func in slices:
+        assert isinstance(result1, tm.SubclassedSeries)
+    else:
+        assert isinstance(result1, type(obj))
+
+    # Confirm .agg() groupby operations return same results
+    if isinstance(result1, DataFrame):
+        tm.assert_frame_equal(result1, result2)
+    else:
+        tm.assert_series_equal(result1, result2)
+
+
+def test_groupby_preserves_metadata():
+    # GH-37343
+    custom_df = tm.SubclassedDataFrame({"a": [1, 2, 3], "b": [1, 1, 2], "c": [7, 8, 9]})
+    assert "testattr" in custom_df._metadata
+    custom_df.testattr = "hello"
+    for _, group_df in custom_df.groupby("c"):
+        assert group_df.testattr == "hello"
+
+    # GH-45314
+    def func(group):
+        assert isinstance(group, tm.SubclassedDataFrame)
+        assert hasattr(group, "testattr")
+        assert group.testattr == "hello"
+        return group.testattr
+
+    msg = "DataFrameGroupBy.apply operated on the grouping columns"
+    with tm.assert_produces_warning(
+        DeprecationWarning,
+        match=msg,
+        raise_on_extra_warnings=False,
+        check_stacklevel=False,
+    ):
+        result = custom_df.groupby("c").apply(func)
+    expected = tm.SubclassedSeries(["hello"] * 3, index=Index([7, 8, 9], name="c"))
+    tm.assert_series_equal(result, expected)
+
+    result = custom_df.groupby("c").apply(func, include_groups=False)
+    tm.assert_series_equal(result, expected)
+
+    # https://github.com/pandas-dev/pandas/pull/56761
+    result = custom_df.groupby("c")[["a", "b"]].apply(func)
+    tm.assert_series_equal(result, expected)
+
+    def func2(group):
+        assert isinstance(group, tm.SubclassedSeries)
+        assert hasattr(group, "testattr")
+        return group.testattr
+
+    custom_series = tm.SubclassedSeries([1, 2, 3])
+    custom_series.testattr = "hello"
+    result = custom_series.groupby(custom_df["c"]).apply(func2)
+    tm.assert_series_equal(result, expected)
+    result = custom_series.groupby(custom_df["c"]).agg(func2)
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize("obj", [DataFrame, tm.SubclassedDataFrame])
+def test_groupby_resample_preserves_subclass(obj):
+    # GH28330 -- preserve subclass through groupby.resample()
+
+    df = obj(
+        {
+            "Buyer": "Carl Carl Carl Carl Joe Carl".split(),
+            "Quantity": [18, 3, 5, 1, 9, 3],
+            "Date": [
+                datetime(2013, 9, 1, 13, 0),
+                datetime(2013, 9, 1, 13, 5),
+                datetime(2013, 10, 1, 20, 0),
+                datetime(2013, 10, 3, 10, 0),
+                datetime(2013, 12, 2, 12, 0),
+                datetime(2013, 9, 2, 14, 0),
+            ],
+        }
+    )
+    df = df.set_index("Date")
+
+    # Confirm groupby.resample() preserves dataframe type
+    msg = "DataFrameGroupBy.resample operated on the grouping columns"
+    with tm.assert_produces_warning(
+        DeprecationWarning,
+        match=msg,
+        raise_on_extra_warnings=False,
+        check_stacklevel=False,
+    ):
+        result = df.groupby("Buyer").resample("5D").sum()
+    assert isinstance(result, obj)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_index_as_string.py

@@ -0,0 +1,85 @@
+import numpy as np
+import pytest
+
+import pandas as pd
+import pandas._testing as tm
+
+
+@pytest.fixture(params=[["inner"], ["inner", "outer"]])
+def frame(request):
+    levels = request.param
+    df = pd.DataFrame(
+        {
+            "outer": ["a", "a", "a", "b", "b", "b"],
+            "inner": [1, 2, 3, 1, 2, 3],
+            "A": np.arange(6),
+            "B": ["one", "one", "two", "two", "one", "one"],
+        }
+    )
+    if levels:
+        df = df.set_index(levels)
+
+    return df
+
+
+@pytest.fixture()
+def series():
+    df = pd.DataFrame(
+        {
+            "outer": ["a", "a", "a", "b", "b", "b"],
+            "inner": [1, 2, 3, 1, 2, 3],
+            "A": np.arange(6),
+            "B": ["one", "one", "two", "two", "one", "one"],
+        }
+    )
+    s = df.set_index(["outer", "inner", "B"])["A"]
+
+    return s
+
+
+@pytest.mark.parametrize(
+    "key_strs,groupers",
+    [
+        ("inner", pd.Grouper(level="inner")),  # Index name
+        (["inner"], [pd.Grouper(level="inner")]),  # List of index name
+        (["B", "inner"], ["B", pd.Grouper(level="inner")]),  # Column and index
+        (["inner", "B"], [pd.Grouper(level="inner"), "B"]),  # Index and column
+    ],
+)
+def test_grouper_index_level_as_string(frame, key_strs, groupers):
+    if "B" not in key_strs or "outer" in frame.columns:
+        result = frame.groupby(key_strs).mean(numeric_only=True)
+        expected = frame.groupby(groupers).mean(numeric_only=True)
+    else:
+        result = frame.groupby(key_strs).mean()
+        expected = frame.groupby(groupers).mean()
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "levels",
+    [
+        "inner",
+        "outer",
+        "B",
+        ["inner"],
+        ["outer"],
+        ["B"],
+        ["inner", "outer"],
+        ["outer", "inner"],
+        ["inner", "outer", "B"],
+        ["B", "outer", "inner"],
+    ],
+)
+def test_grouper_index_level_as_string_series(series, levels):
+    # Compute expected result
+    if isinstance(levels, list):
+        groupers = [pd.Grouper(level=lv) for lv in levels]
+    else:
+        groupers = pd.Grouper(level=levels)
+
+    expected = series.groupby(groupers).mean()
+
+    # Compute and check result
+    result = series.groupby(levels).mean()
+    tm.assert_series_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_indexing.py

@@ -0,0 +1,333 @@
+# Test GroupBy._positional_selector positional grouped indexing GH#42864
+
+import numpy as np
+import pytest
+
+import pandas as pd
+import pandas._testing as tm
+
+
+@pytest.mark.parametrize(
+    "arg, expected_rows",
+    [
+        [0, [0, 1, 4]],
+        [2, [5]],
+        [5, []],
+        [-1, [3, 4, 7]],
+        [-2, [1, 6]],
+        [-6, []],
+    ],
+)
+def test_int(slice_test_df, slice_test_grouped, arg, expected_rows):
+    # Test single integer
+    result = slice_test_grouped._positional_selector[arg]
+    expected = slice_test_df.iloc[expected_rows]
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_slice(slice_test_df, slice_test_grouped):
+    # Test single slice
+    result = slice_test_grouped._positional_selector[0:3:2]
+    expected = slice_test_df.iloc[[0, 1, 4, 5]]
+
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "arg, expected_rows",
+    [
+        [[0, 2], [0, 1, 4, 5]],
+        [[0, 2, -1], [0, 1, 3, 4, 5, 7]],
+        [range(0, 3, 2), [0, 1, 4, 5]],
+        [{0, 2}, [0, 1, 4, 5]],
+    ],
+    ids=[
+        "list",
+        "negative",
+        "range",
+        "set",
+    ],
+)
+def test_list(slice_test_df, slice_test_grouped, arg, expected_rows):
+    # Test lists of integers and integer valued iterables
+    result = slice_test_grouped._positional_selector[arg]
+    expected = slice_test_df.iloc[expected_rows]
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_ints(slice_test_df, slice_test_grouped):
+    # Test tuple of ints
+    result = slice_test_grouped._positional_selector[0, 2, -1]
+    expected = slice_test_df.iloc[[0, 1, 3, 4, 5, 7]]
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_slices(slice_test_df, slice_test_grouped):
+    # Test tuple of slices
+    result = slice_test_grouped._positional_selector[:2, -2:]
+    expected = slice_test_df.iloc[[0, 1, 2, 3, 4, 6, 7]]
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_mix(slice_test_df, slice_test_grouped):
+    # Test mixed tuple of ints and slices
+    result = slice_test_grouped._positional_selector[0, 1, -2:]
+    expected = slice_test_df.iloc[[0, 1, 2, 3, 4, 6, 7]]
+
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "arg, expected_rows",
+    [
+        [0, [0, 1, 4]],
+        [[0, 2, -1], [0, 1, 3, 4, 5, 7]],
+        [(slice(None, 2), slice(-2, None)), [0, 1, 2, 3, 4, 6, 7]],
+    ],
+)
+def test_as_index(slice_test_df, arg, expected_rows):
+    # Test the default as_index behaviour
+    result = slice_test_df.groupby("Group", sort=False)._positional_selector[arg]
+    expected = slice_test_df.iloc[expected_rows]
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_doc_examples():
+    # Test the examples in the documentation
+    df = pd.DataFrame(
+        [["a", 1], ["a", 2], ["a", 3], ["b", 4], ["b", 5]], columns=["A", "B"]
+    )
+
+    grouped = df.groupby("A", as_index=False)
+
+    result = grouped._positional_selector[1:2]
+    expected = pd.DataFrame([["a", 2], ["b", 5]], columns=["A", "B"], index=[1, 4])
+
+    tm.assert_frame_equal(result, expected)
+
+    result = grouped._positional_selector[1, -1]
+    expected = pd.DataFrame(
+        [["a", 2], ["a", 3], ["b", 5]], columns=["A", "B"], index=[1, 2, 4]
+    )
+
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.fixture()
+def multiindex_data():
+    rng = np.random.default_rng(2)
+    ndates = 100
+    nitems = 20
+    dates = pd.date_range("20130101", periods=ndates, freq="D")
+    items = [f"item {i}" for i in range(nitems)]
+
+    data = {}
+    for date in dates:
+        nitems_for_date = nitems - rng.integers(0, 12)
+        levels = [
+            (item, rng.integers(0, 10000) / 100, rng.integers(0, 10000) / 100)
+            for item in items[:nitems_for_date]
+        ]
+        levels.sort(key=lambda x: x[1])
+        data[date] = levels
+
+    return data
+
+
+def _make_df_from_data(data):
+    rows = {}
+    for date in data:
+        for level in data[date]:
+            rows[(date, level[0])] = {"A": level[1], "B": level[2]}
+
+    df = pd.DataFrame.from_dict(rows, orient="index")
+    df.index.names = ("Date", "Item")
+    return df
+
+
+def test_multiindex(multiindex_data):
+    # Test the multiindex mentioned as the use-case in the documentation
+    df = _make_df_from_data(multiindex_data)
+    result = df.groupby("Date", as_index=False).nth(slice(3, -3))
+
+    sliced = {date: multiindex_data[date][3:-3] for date in multiindex_data}
+    expected = _make_df_from_data(sliced)
+
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("arg", [1, 5, 30, 1000, -1, -5, -30, -1000])
+@pytest.mark.parametrize("method", ["head", "tail"])
+@pytest.mark.parametrize("simulated", [True, False])
+def test_against_head_and_tail(arg, method, simulated):
+    # Test gives the same results as grouped head and tail
+    n_groups = 100
+    n_rows_per_group = 30
+
+    data = {
+        "group": [
+            f"group {g}" for j in range(n_rows_per_group) for g in range(n_groups)
+        ],
+        "value": [
+            f"group {g} row {j}"
+            for j in range(n_rows_per_group)
+            for g in range(n_groups)
+        ],
+    }
+    df = pd.DataFrame(data)
+    grouped = df.groupby("group", as_index=False)
+    size = arg if arg >= 0 else n_rows_per_group + arg
+
+    if method == "head":
+        result = grouped._positional_selector[:arg]
+
+        if simulated:
+            indices = [
+                j * n_groups + i
+                for j in range(size)
+                for i in range(n_groups)
+                if j * n_groups + i < n_groups * n_rows_per_group
+            ]
+            expected = df.iloc[indices]
+
+        else:
+            expected = grouped.head(arg)
+
+    else:
+        result = grouped._positional_selector[-arg:]
+
+        if simulated:
+            indices = [
+                (n_rows_per_group + j - size) * n_groups + i
+                for j in range(size)
+                for i in range(n_groups)
+                if (n_rows_per_group + j - size) * n_groups + i >= 0
+            ]
+            expected = df.iloc[indices]
+
+        else:
+            expected = grouped.tail(arg)
+
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("start", [None, 0, 1, 10, -1, -10])
+@pytest.mark.parametrize("stop", [None, 0, 1, 10, -1, -10])
+@pytest.mark.parametrize("step", [None, 1, 5])
+def test_against_df_iloc(start, stop, step):
+    # Test that a single group gives the same results as DataFrame.iloc
+    n_rows = 30
+
+    data = {
+        "group": ["group 0"] * n_rows,
+        "value": list(range(n_rows)),
+    }
+    df = pd.DataFrame(data)
+    grouped = df.groupby("group", as_index=False)
+
+    result = grouped._positional_selector[start:stop:step]
+    expected = df.iloc[start:stop:step]
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_series():
+    # Test grouped Series
+    ser = pd.Series([1, 2, 3, 4, 5], index=["a", "a", "a", "b", "b"])
+    grouped = ser.groupby(level=0)
+    result = grouped._positional_selector[1:2]
+    expected = pd.Series([2, 5], index=["a", "b"])
+
+    tm.assert_series_equal(result, expected)
+
+
+@pytest.mark.parametrize("step", [1, 2, 3, 4, 5])
+def test_step(step):
+    # Test slice with various step values
+    data = [["x", f"x{i}"] for i in range(5)]
+    data += [["y", f"y{i}"] for i in range(4)]
+    data += [["z", f"z{i}"] for i in range(3)]
+    df = pd.DataFrame(data, columns=["A", "B"])
+
+    grouped = df.groupby("A", as_index=False)
+
+    result = grouped._positional_selector[::step]
+
+    data = [["x", f"x{i}"] for i in range(0, 5, step)]
+    data += [["y", f"y{i}"] for i in range(0, 4, step)]
+    data += [["z", f"z{i}"] for i in range(0, 3, step)]
+
+    index = [0 + i for i in range(0, 5, step)]
+    index += [5 + i for i in range(0, 4, step)]
+    index += [9 + i for i in range(0, 3, step)]
+
+    expected = pd.DataFrame(data, columns=["A", "B"], index=index)
+
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.fixture()
+def column_group_df():
+    return pd.DataFrame(
+        [[0, 1, 2, 3, 4, 5, 6], [0, 0, 1, 0, 1, 0, 2]],
+        columns=["A", "B", "C", "D", "E", "F", "G"],
+    )
+
+
+def test_column_axis(column_group_df):
+    msg = "DataFrame.groupby with axis=1"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        g = column_group_df.groupby(column_group_df.iloc[1], axis=1)
+    result = g._positional_selector[1:-1]
+    expected = column_group_df.iloc[:, [1, 3]]
+
+    tm.assert_frame_equal(result, expected)
+
+
+def test_columns_on_iter():
+    # GitHub issue #44821
+    df = pd.DataFrame({k: range(10) for k in "ABC"})
+
+    # Group-by and select columns
+    cols = ["A", "B"]
+    for _, dg in df.groupby(df.A < 4)[cols]:
+        tm.assert_index_equal(dg.columns, pd.Index(cols))
+        assert "C" not in dg.columns
+
+
+@pytest.mark.parametrize("func", [list, pd.Index, pd.Series, np.array])
+def test_groupby_duplicated_columns(func):
+    # GH#44924
+    df = pd.DataFrame(
+        {
+            "A": [1, 2],
+            "B": [3, 3],
+            "C": ["G", "G"],
+        }
+    )
+    result = df.groupby("C")[func(["A", "B", "A"])].mean()
+    expected = pd.DataFrame(
+        [[1.5, 3.0, 1.5]], columns=["A", "B", "A"], index=pd.Index(["G"], name="C")
+    )
+    tm.assert_frame_equal(result, expected)
+
+
+def test_groupby_get_nonexisting_groups():
+    # GH#32492
+    df = pd.DataFrame(
+        data={
+            "A": ["a1", "a2", None],
+            "B": ["b1", "b2", "b1"],
+            "val": [1, 2, 3],
+        }
+    )
+    grps = df.groupby(by=["A", "B"])
+
+    msg = "('a2', 'b1')"
+    with pytest.raises(KeyError, match=msg):
+        grps.get_group(("a2", "b1"))

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_libgroupby.py

@@ -0,0 +1,331 @@
+import numpy as np
+import pytest
+
+from pandas._libs import groupby as libgroupby
+from pandas._libs.groupby import (
+    group_cumprod,
+    group_cumsum,
+    group_mean,
+    group_sum,
+    group_var,
+)
+
+from pandas.core.dtypes.common import ensure_platform_int
+
+from pandas import isna
+import pandas._testing as tm
+
+
+class GroupVarTestMixin:
+    def test_group_var_generic_1d(self):
+        prng = np.random.default_rng(2)
+
+        out = (np.nan * np.ones((5, 1))).astype(self.dtype)
+        counts = np.zeros(5, dtype="int64")
+        values = 10 * prng.random((15, 1)).astype(self.dtype)
+        labels = np.tile(np.arange(5), (3,)).astype("intp")
+
+        expected_out = (
+            np.squeeze(values).reshape((5, 3), order="F").std(axis=1, ddof=1) ** 2
+        )[:, np.newaxis]
+        expected_counts = counts + 3
+
+        self.algo(out, counts, values, labels)
+        assert np.allclose(out, expected_out, self.rtol)
+        tm.assert_numpy_array_equal(counts, expected_counts)
+
+    def test_group_var_generic_1d_flat_labels(self):
+        prng = np.random.default_rng(2)
+
+        out = (np.nan * np.ones((1, 1))).astype(self.dtype)
+        counts = np.zeros(1, dtype="int64")
+        values = 10 * prng.random((5, 1)).astype(self.dtype)
+        labels = np.zeros(5, dtype="intp")
+
+        expected_out = np.array([[values.std(ddof=1) ** 2]])
+        expected_counts = counts + 5
+
+        self.algo(out, counts, values, labels)
+
+        assert np.allclose(out, expected_out, self.rtol)
+        tm.assert_numpy_array_equal(counts, expected_counts)
+
+    def test_group_var_generic_2d_all_finite(self):
+        prng = np.random.default_rng(2)
+
+        out = (np.nan * np.ones((5, 2))).astype(self.dtype)
+        counts = np.zeros(5, dtype="int64")
+        values = 10 * prng.random((10, 2)).astype(self.dtype)
+        labels = np.tile(np.arange(5), (2,)).astype("intp")
+
+        expected_out = np.std(values.reshape(2, 5, 2), ddof=1, axis=0) ** 2
+        expected_counts = counts + 2
+
+        self.algo(out, counts, values, labels)
+        assert np.allclose(out, expected_out, self.rtol)
+        tm.assert_numpy_array_equal(counts, expected_counts)
+
+    def test_group_var_generic_2d_some_nan(self):
+        prng = np.random.default_rng(2)
+
+        out = (np.nan * np.ones((5, 2))).astype(self.dtype)
+        counts = np.zeros(5, dtype="int64")
+        values = 10 * prng.random((10, 2)).astype(self.dtype)
+        values[:, 1] = np.nan
+        labels = np.tile(np.arange(5), (2,)).astype("intp")
+
+        expected_out = np.vstack(
+            [
+                values[:, 0].reshape(5, 2, order="F").std(ddof=1, axis=1) ** 2,
+                np.nan * np.ones(5),
+            ]
+        ).T.astype(self.dtype)
+        expected_counts = counts + 2
+
+        self.algo(out, counts, values, labels)
+        tm.assert_almost_equal(out, expected_out, rtol=0.5e-06)
+        tm.assert_numpy_array_equal(counts, expected_counts)
+
+    def test_group_var_constant(self):
+        # Regression test from GH 10448.
+
+        out = np.array([[np.nan]], dtype=self.dtype)
+        counts = np.array([0], dtype="int64")
+        values = 0.832845131556193 * np.ones((3, 1), dtype=self.dtype)
+        labels = np.zeros(3, dtype="intp")
+
+        self.algo(out, counts, values, labels)
+
+        assert counts[0] == 3
+        assert out[0, 0] >= 0
+        tm.assert_almost_equal(out[0, 0], 0.0)
+
+
+class TestGroupVarFloat64(GroupVarTestMixin):
+    __test__ = True
+
+    algo = staticmethod(group_var)
+    dtype = np.float64
+    rtol = 1e-5
+
+    def test_group_var_large_inputs(self):
+        prng = np.random.default_rng(2)
+
+        out = np.array([[np.nan]], dtype=self.dtype)
+        counts = np.array([0], dtype="int64")
+        values = (prng.random(10**6) + 10**12).astype(self.dtype)
+        values.shape = (10**6, 1)
+        labels = np.zeros(10**6, dtype="intp")
+
+        self.algo(out, counts, values, labels)
+
+        assert counts[0] == 10**6
+        tm.assert_almost_equal(out[0, 0], 1.0 / 12, rtol=0.5e-3)
+
+
+class TestGroupVarFloat32(GroupVarTestMixin):
+    __test__ = True
+
+    algo = staticmethod(group_var)
+    dtype = np.float32
+    rtol = 1e-2
+
+
+@pytest.mark.parametrize("dtype", ["float32", "float64"])
+def test_group_ohlc(dtype):
+    obj = np.array(np.random.default_rng(2).standard_normal(20), dtype=dtype)
+
+    bins = np.array([6, 12, 20])
+    out = np.zeros((3, 4), dtype)
+    counts = np.zeros(len(out), dtype=np.int64)
+    labels = ensure_platform_int(np.repeat(np.arange(3), np.diff(np.r_[0, bins])))
+
+    func = libgroupby.group_ohlc
+    func(out, counts, obj[:, None], labels)
+
+    def _ohlc(group):
+        if isna(group).all():
+            return np.repeat(np.nan, 4)
+        return [group[0], group.max(), group.min(), group[-1]]
+
+    expected = np.array([_ohlc(obj[:6]), _ohlc(obj[6:12]), _ohlc(obj[12:])])
+
+    tm.assert_almost_equal(out, expected)
+    tm.assert_numpy_array_equal(counts, np.array([6, 6, 8], dtype=np.int64))
+
+    obj[:6] = np.nan
+    func(out, counts, obj[:, None], labels)
+    expected[0] = np.nan
+    tm.assert_almost_equal(out, expected)
+
+
+def _check_cython_group_transform_cumulative(pd_op, np_op, dtype):
+    """
+    Check a group transform that executes a cumulative function.
+
+    Parameters
+    ----------
+    pd_op : callable
+        The pandas cumulative function.
+    np_op : callable
+        The analogous one in NumPy.
+    dtype : type
+        The specified dtype of the data.
+    """
+    is_datetimelike = False
+
+    data = np.array([[1], [2], [3], [4]], dtype=dtype)
+    answer = np.zeros_like(data)
+
+    labels = np.array([0, 0, 0, 0], dtype=np.intp)
+    ngroups = 1
+    pd_op(answer, data, labels, ngroups, is_datetimelike)
+
+    tm.assert_numpy_array_equal(np_op(data), answer[:, 0], check_dtype=False)
+
+
+@pytest.mark.parametrize("np_dtype", ["int64", "uint64", "float32", "float64"])
+def test_cython_group_transform_cumsum(np_dtype):
+    # see gh-4095
+    dtype = np.dtype(np_dtype).type
+    pd_op, np_op = group_cumsum, np.cumsum
+    _check_cython_group_transform_cumulative(pd_op, np_op, dtype)
+
+
+def test_cython_group_transform_cumprod():
+    # see gh-4095
+    dtype = np.float64
+    pd_op, np_op = group_cumprod, np.cumprod
+    _check_cython_group_transform_cumulative(pd_op, np_op, dtype)
+
+
+def test_cython_group_transform_algos():
+    # see gh-4095
+    is_datetimelike = False
+
+    # with nans
+    labels = np.array([0, 0, 0, 0, 0], dtype=np.intp)
+    ngroups = 1
+
+    data = np.array([[1], [2], [3], [np.nan], [4]], dtype="float64")
+    actual = np.zeros_like(data)
+    actual.fill(np.nan)
+    group_cumprod(actual, data, labels, ngroups, is_datetimelike)
+    expected = np.array([1, 2, 6, np.nan, 24], dtype="float64")
+    tm.assert_numpy_array_equal(actual[:, 0], expected)
+
+    actual = np.zeros_like(data)
+    actual.fill(np.nan)
+    group_cumsum(actual, data, labels, ngroups, is_datetimelike)
+    expected = np.array([1, 3, 6, np.nan, 10], dtype="float64")
+    tm.assert_numpy_array_equal(actual[:, 0], expected)
+
+    # timedelta
+    is_datetimelike = True
+    data = np.array([np.timedelta64(1, "ns")] * 5, dtype="m8[ns]")[:, None]
+    actual = np.zeros_like(data, dtype="int64")
+    group_cumsum(actual, data.view("int64"), labels, ngroups, is_datetimelike)
+    expected = np.array(
+        [
+            np.timedelta64(1, "ns"),
+            np.timedelta64(2, "ns"),
+            np.timedelta64(3, "ns"),
+            np.timedelta64(4, "ns"),
+            np.timedelta64(5, "ns"),
+        ]
+    )
+    tm.assert_numpy_array_equal(actual[:, 0].view("m8[ns]"), expected)
+
+
+def test_cython_group_mean_datetimelike():
+    actual = np.zeros(shape=(1, 1), dtype="float64")
+    counts = np.array([0], dtype="int64")
+    data = (
+        np.array(
+            [np.timedelta64(2, "ns"), np.timedelta64(4, "ns"), np.timedelta64("NaT")],
+            dtype="m8[ns]",
+        )[:, None]
+        .view("int64")
+        .astype("float64")
+    )
+    labels = np.zeros(len(data), dtype=np.intp)
+
+    group_mean(actual, counts, data, labels, is_datetimelike=True)
+
+    tm.assert_numpy_array_equal(actual[:, 0], np.array([3], dtype="float64"))
+
+
+def test_cython_group_mean_wrong_min_count():
+    actual = np.zeros(shape=(1, 1), dtype="float64")
+    counts = np.zeros(1, dtype="int64")
+    data = np.zeros(1, dtype="float64")[:, None]
+    labels = np.zeros(1, dtype=np.intp)
+
+    with pytest.raises(AssertionError, match="min_count"):
+        group_mean(actual, counts, data, labels, is_datetimelike=True, min_count=0)
+
+
+def test_cython_group_mean_not_datetimelike_but_has_NaT_values():
+    actual = np.zeros(shape=(1, 1), dtype="float64")
+    counts = np.array([0], dtype="int64")
+    data = (
+        np.array(
+            [np.timedelta64("NaT"), np.timedelta64("NaT")],
+            dtype="m8[ns]",
+        )[:, None]
+        .view("int64")
+        .astype("float64")
+    )
+    labels = np.zeros(len(data), dtype=np.intp)
+
+    group_mean(actual, counts, data, labels, is_datetimelike=False)
+
+    tm.assert_numpy_array_equal(
+        actual[:, 0], np.array(np.divide(np.add(data[0], data[1]), 2), dtype="float64")
+    )
+
+
+def test_cython_group_mean_Inf_at_begining_and_end():
+    # GH 50367
+    actual = np.array([[np.nan, np.nan], [np.nan, np.nan]], dtype="float64")
+    counts = np.array([0, 0], dtype="int64")
+    data = np.array(
+        [[np.inf, 1.0], [1.0, 2.0], [2.0, 3.0], [3.0, 4.0], [4.0, 5.0], [5, np.inf]],
+        dtype="float64",
+    )
+    labels = np.array([0, 1, 0, 1, 0, 1], dtype=np.intp)
+
+    group_mean(actual, counts, data, labels, is_datetimelike=False)
+
+    expected = np.array([[np.inf, 3], [3, np.inf]], dtype="float64")
+
+    tm.assert_numpy_array_equal(
+        actual,
+        expected,
+    )
+
+
+@pytest.mark.parametrize(
+    "values, out",
+    [
+        ([[np.inf], [np.inf], [np.inf]], [[np.inf], [np.inf]]),
+        ([[np.inf], [np.inf], [-np.inf]], [[np.inf], [np.nan]]),
+        ([[np.inf], [-np.inf], [np.inf]], [[np.inf], [np.nan]]),
+        ([[np.inf], [-np.inf], [-np.inf]], [[np.inf], [-np.inf]]),
+    ],
+)
+def test_cython_group_sum_Inf_at_begining_and_end(values, out):
+    # GH #53606
+    actual = np.array([[np.nan], [np.nan]], dtype="float64")
+    counts = np.array([0, 0], dtype="int64")
+    data = np.array(values, dtype="float64")
+    labels = np.array([0, 1, 1], dtype=np.intp)
+
+    group_sum(actual, counts, data, labels, None, is_datetimelike=False)
+
+    expected = np.array(out, dtype="float64")
+
+    tm.assert_numpy_array_equal(
+        actual,
+        expected,
+    )

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_missing.py

@@ -0,0 +1,163 @@
+import numpy as np
+import pytest
+
+import pandas as pd
+from pandas import (
+    DataFrame,
+    Index,
+    date_range,
+)
+import pandas._testing as tm
+
+
+@pytest.mark.parametrize("func", ["ffill", "bfill"])
+def test_groupby_column_index_name_lost_fill_funcs(func):
+    # GH: 29764 groupby loses index sometimes
+    df = DataFrame(
+        [[1, 1.0, -1.0], [1, np.nan, np.nan], [1, 2.0, -2.0]],
+        columns=Index(["type", "a", "b"], name="idx"),
+    )
+    df_grouped = df.groupby(["type"])[["a", "b"]]
+    result = getattr(df_grouped, func)().columns
+    expected = Index(["a", "b"], name="idx")
+    tm.assert_index_equal(result, expected)
+
+
+@pytest.mark.parametrize("func", ["ffill", "bfill"])
+def test_groupby_fill_duplicate_column_names(func):
+    # GH: 25610 ValueError with duplicate column names
+    df1 = DataFrame({"field1": [1, 3, 4], "field2": [1, 3, 4]})
+    df2 = DataFrame({"field1": [1, np.nan, 4]})
+    df_grouped = pd.concat([df1, df2], axis=1).groupby(by=["field2"])
+    expected = DataFrame(
+        [[1, 1.0], [3, np.nan], [4, 4.0]], columns=["field1", "field1"]
+    )
+    result = getattr(df_grouped, func)()
+    tm.assert_frame_equal(result, expected)
+
+
+def test_ffill_missing_arguments():
+    # GH 14955
+    df = DataFrame({"a": [1, 2], "b": [1, 1]})
+    msg = "DataFrameGroupBy.fillna is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        with pytest.raises(ValueError, match="Must specify a fill"):
+            df.groupby("b").fillna()
+
+
+@pytest.mark.parametrize(
+    "method, expected", [("ffill", [None, "a", "a"]), ("bfill", ["a", "a", None])]
+)
+def test_fillna_with_string_dtype(method, expected):
+    # GH 40250
+    df = DataFrame({"a": pd.array([None, "a", None], dtype="string"), "b": [0, 0, 0]})
+    grp = df.groupby("b")
+    msg = "DataFrameGroupBy.fillna is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        result = grp.fillna(method=method)
+    expected = DataFrame({"a": pd.array(expected, dtype="string")})
+    tm.assert_frame_equal(result, expected)
+
+
+def test_fill_consistency():
+    # GH9221
+    # pass thru keyword arguments to the generated wrapper
+    # are set if the passed kw is None (only)
+    df = DataFrame(
+        index=pd.MultiIndex.from_product(
+            [["value1", "value2"], date_range("2014-01-01", "2014-01-06")]
+        ),
+        columns=Index(["1", "2"], name="id"),
+    )
+    df["1"] = [
+        np.nan,
+        1,
+        np.nan,
+        np.nan,
+        11,
+        np.nan,
+        np.nan,
+        2,
+        np.nan,
+        np.nan,
+        22,
+        np.nan,
+    ]
+    df["2"] = [
+        np.nan,
+        3,
+        np.nan,
+        np.nan,
+        33,
+        np.nan,
+        np.nan,
+        4,
+        np.nan,
+        np.nan,
+        44,
+        np.nan,
+    ]
+
+    msg = "The 'axis' keyword in DataFrame.groupby is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        expected = df.groupby(level=0, axis=0).fillna(method="ffill")
+
+    msg = "DataFrame.groupby with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        result = df.T.groupby(level=0, axis=1).fillna(method="ffill").T
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("method", ["ffill", "bfill"])
+@pytest.mark.parametrize("dropna", [True, False])
+@pytest.mark.parametrize("has_nan_group", [True, False])
+def test_ffill_handles_nan_groups(dropna, method, has_nan_group):
+    # GH 34725
+
+    df_without_nan_rows = DataFrame([(1, 0.1), (2, 0.2)])
+
+    ridx = [-1, 0, -1, -1, 1, -1]
+    df = df_without_nan_rows.reindex(ridx).reset_index(drop=True)
+
+    group_b = np.nan if has_nan_group else "b"
+    df["group_col"] = pd.Series(["a"] * 3 + [group_b] * 3)
+
+    grouped = df.groupby(by="group_col", dropna=dropna)
+    result = getattr(grouped, method)(limit=None)
+
+    expected_rows = {
+        ("ffill", True, True): [-1, 0, 0, -1, -1, -1],
+        ("ffill", True, False): [-1, 0, 0, -1, 1, 1],
+        ("ffill", False, True): [-1, 0, 0, -1, 1, 1],
+        ("ffill", False, False): [-1, 0, 0, -1, 1, 1],
+        ("bfill", True, True): [0, 0, -1, -1, -1, -1],
+        ("bfill", True, False): [0, 0, -1, 1, 1, -1],
+        ("bfill", False, True): [0, 0, -1, 1, 1, -1],
+        ("bfill", False, False): [0, 0, -1, 1, 1, -1],
+    }
+
+    ridx = expected_rows.get((method, dropna, has_nan_group))
+    expected = df_without_nan_rows.reindex(ridx).reset_index(drop=True)
+    # columns are a 'take' on df.columns, which are object dtype
+    expected.columns = expected.columns.astype(object)
+
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.parametrize("min_count, value", [(2, np.nan), (-1, 1.0)])
+@pytest.mark.parametrize("func", ["first", "last", "max", "min"])
+def test_min_count(func, min_count, value):
+    # GH#37821
+    df = DataFrame({"a": [1] * 3, "b": [1, np.nan, np.nan], "c": [np.nan] * 3})
+    result = getattr(df.groupby("a"), func)(min_count=min_count)
+    expected = DataFrame({"b": [value], "c": [np.nan]}, index=Index([1], name="a"))
+    tm.assert_frame_equal(result, expected)
+
+
+def test_indices_with_missing():
+    # GH 9304
+    df = DataFrame({"a": [1, 1, np.nan], "b": [2, 3, 4], "c": [5, 6, 7]})
+    g = df.groupby(["a", "b"])
+    result = g.indices
+    expected = {(1.0, 2): np.array([0]), (1.0, 3): np.array([1])}
+    assert result == expected

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_numba.py

@@ -0,0 +1,80 @@
+import pytest
+
+from pandas import (
+    DataFrame,
+    Series,
+    option_context,
+)
+import pandas._testing as tm
+
+pytestmark = pytest.mark.single_cpu
+
+pytest.importorskip("numba")
+
+
+@pytest.mark.filterwarnings("ignore")
+# Filter warnings when parallel=True and the function can't be parallelized by Numba
+class TestEngine:
+    def test_cython_vs_numba_frame(
+        self, sort, nogil, parallel, nopython, numba_supported_reductions
+    ):
+        func, kwargs = numba_supported_reductions
+        df = DataFrame({"a": [3, 2, 3, 2], "b": range(4), "c": range(1, 5)})
+        engine_kwargs = {"nogil": nogil, "parallel": parallel, "nopython": nopython}
+        gb = df.groupby("a", sort=sort)
+        result = getattr(gb, func)(
+            engine="numba", engine_kwargs=engine_kwargs, **kwargs
+        )
+        expected = getattr(gb, func)(**kwargs)
+        tm.assert_frame_equal(result, expected)
+
+    def test_cython_vs_numba_getitem(
+        self, sort, nogil, parallel, nopython, numba_supported_reductions
+    ):
+        func, kwargs = numba_supported_reductions
+        df = DataFrame({"a": [3, 2, 3, 2], "b": range(4), "c": range(1, 5)})
+        engine_kwargs = {"nogil": nogil, "parallel": parallel, "nopython": nopython}
+        gb = df.groupby("a", sort=sort)["c"]
+        result = getattr(gb, func)(
+            engine="numba", engine_kwargs=engine_kwargs, **kwargs
+        )
+        expected = getattr(gb, func)(**kwargs)
+        tm.assert_series_equal(result, expected)
+
+    def test_cython_vs_numba_series(
+        self, sort, nogil, parallel, nopython, numba_supported_reductions
+    ):
+        func, kwargs = numba_supported_reductions
+        ser = Series(range(3), index=[1, 2, 1], name="foo")
+        engine_kwargs = {"nogil": nogil, "parallel": parallel, "nopython": nopython}
+        gb = ser.groupby(level=0, sort=sort)
+        result = getattr(gb, func)(
+            engine="numba", engine_kwargs=engine_kwargs, **kwargs
+        )
+        expected = getattr(gb, func)(**kwargs)
+        tm.assert_series_equal(result, expected)
+
+    def test_as_index_false_unsupported(self, numba_supported_reductions):
+        func, kwargs = numba_supported_reductions
+        df = DataFrame({"a": [3, 2, 3, 2], "b": range(4), "c": range(1, 5)})
+        gb = df.groupby("a", as_index=False)
+        with pytest.raises(NotImplementedError, match="as_index=False"):
+            getattr(gb, func)(engine="numba", **kwargs)
+
+    def test_axis_1_unsupported(self, numba_supported_reductions):
+        func, kwargs = numba_supported_reductions
+        df = DataFrame({"a": [3, 2, 3, 2], "b": range(4), "c": range(1, 5)})
+        gb = df.groupby("a", axis=1)
+        with pytest.raises(NotImplementedError, match="axis=1"):
+            getattr(gb, func)(engine="numba", **kwargs)
+
+    def test_no_engine_doesnt_raise(self):
+        # GH55520
+        df = DataFrame({"a": [3, 2, 3, 2], "b": range(4), "c": range(1, 5)})
+        gb = df.groupby("a")
+        # Make sure behavior of functions w/out engine argument don't raise
+        # when the global use_numba option is set
+        with option_context("compute.use_numba", True):
+            res = gb.agg({"b": "first"})
+        expected = gb.agg({"b": "first"})
+        tm.assert_frame_equal(res, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_numeric_only.py

@@ -0,0 +1,521 @@
+import re
+
+import numpy as np
+import pytest
+
+from pandas._libs import lib
+
+import pandas as pd
+from pandas import (
+    DataFrame,
+    Index,
+    Series,
+    Timestamp,
+    date_range,
+)
+import pandas._testing as tm
+from pandas.tests.groupby import get_groupby_method_args
+
+
+class TestNumericOnly:
+    # make sure that we are passing thru kwargs to our agg functions
+
+    @pytest.fixture
+    def df(self):
+        # GH3668
+        # GH5724
+        df = DataFrame(
+            {
+                "group": [1, 1, 2],
+                "int": [1, 2, 3],
+                "float": [4.0, 5.0, 6.0],
+                "string": list("abc"),
+                "category_string": Series(list("abc")).astype("category"),
+                "category_int": [7, 8, 9],
+                "datetime": date_range("20130101", periods=3),
+                "datetimetz": date_range("20130101", periods=3, tz="US/Eastern"),
+                "timedelta": pd.timedelta_range("1 s", periods=3, freq="s"),
+            },
+            columns=[
+                "group",
+                "int",
+                "float",
+                "string",
+                "category_string",
+                "category_int",
+                "datetime",
+                "datetimetz",
+                "timedelta",
+            ],
+        )
+        return df
+
+    @pytest.mark.parametrize("method", ["mean", "median"])
+    def test_averages(self, df, method):
+        # mean / median
+        expected_columns_numeric = Index(["int", "float", "category_int"])
+
+        gb = df.groupby("group")
+        expected = DataFrame(
+            {
+                "category_int": [7.5, 9],
+                "float": [4.5, 6.0],
+                "timedelta": [pd.Timedelta("1.5s"), pd.Timedelta("3s")],
+                "int": [1.5, 3],
+                "datetime": [
+                    Timestamp("2013-01-01 12:00:00"),
+                    Timestamp("2013-01-03 00:00:00"),
+                ],
+                "datetimetz": [
+                    Timestamp("2013-01-01 12:00:00", tz="US/Eastern"),
+                    Timestamp("2013-01-03 00:00:00", tz="US/Eastern"),
+                ],
+            },
+            index=Index([1, 2], name="group"),
+            columns=[
+                "int",
+                "float",
+                "category_int",
+            ],
+        )
+
+        result = getattr(gb, method)(numeric_only=True)
+        tm.assert_frame_equal(result.reindex_like(expected), expected)
+
+        expected_columns = expected.columns
+
+        self._check(df, method, expected_columns, expected_columns_numeric)
+
+    @pytest.mark.parametrize("method", ["min", "max"])
+    def test_extrema(self, df, method):
+        # TODO: min, max *should* handle
+        # categorical (ordered) dtype
+
+        expected_columns = Index(
+            [
+                "int",
+                "float",
+                "string",
+                "category_int",
+                "datetime",
+                "datetimetz",
+                "timedelta",
+            ]
+        )
+        expected_columns_numeric = expected_columns
+
+        self._check(df, method, expected_columns, expected_columns_numeric)
+
+    @pytest.mark.parametrize("method", ["first", "last"])
+    def test_first_last(self, df, method):
+        expected_columns = Index(
+            [
+                "int",
+                "float",
+                "string",
+                "category_string",
+                "category_int",
+                "datetime",
+                "datetimetz",
+                "timedelta",
+            ]
+        )
+        expected_columns_numeric = expected_columns
+
+        self._check(df, method, expected_columns, expected_columns_numeric)
+
+    @pytest.mark.parametrize("method", ["sum", "cumsum"])
+    def test_sum_cumsum(self, df, method):
+        expected_columns_numeric = Index(["int", "float", "category_int"])
+        expected_columns = Index(
+            ["int", "float", "string", "category_int", "timedelta"]
+        )
+        if method == "cumsum":
+            # cumsum loses string
+            expected_columns = Index(["int", "float", "category_int", "timedelta"])
+
+        self._check(df, method, expected_columns, expected_columns_numeric)
+
+    @pytest.mark.parametrize("method", ["prod", "cumprod"])
+    def test_prod_cumprod(self, df, method):
+        expected_columns = Index(["int", "float", "category_int"])
+        expected_columns_numeric = expected_columns
+
+        self._check(df, method, expected_columns, expected_columns_numeric)
+
+    @pytest.mark.parametrize("method", ["cummin", "cummax"])
+    def test_cummin_cummax(self, df, method):
+        # like min, max, but don't include strings
+        expected_columns = Index(
+            ["int", "float", "category_int", "datetime", "datetimetz", "timedelta"]
+        )
+
+        # GH#15561: numeric_only=False set by default like min/max
+        expected_columns_numeric = expected_columns
+
+        self._check(df, method, expected_columns, expected_columns_numeric)
+
+    def _check(self, df, method, expected_columns, expected_columns_numeric):
+        gb = df.groupby("group")
+
+        # object dtypes for transformations are not implemented in Cython and
+        # have no Python fallback
+        exception = NotImplementedError if method.startswith("cum") else TypeError
+
+        if method in ("min", "max", "cummin", "cummax", "cumsum", "cumprod"):
+            # The methods default to numeric_only=False and raise TypeError
+            msg = "|".join(
+                [
+                    "Categorical is not ordered",
+                    f"Cannot perform {method} with non-ordered Categorical",
+                    re.escape(f"agg function failed [how->{method},dtype->object]"),
+                    # cumsum/cummin/cummax/cumprod
+                    "function is not implemented for this dtype",
+                ]
+            )
+            with pytest.raises(exception, match=msg):
+                getattr(gb, method)()
+        elif method in ("sum", "mean", "median", "prod"):
+            msg = "|".join(
+                [
+                    "category type does not support sum operations",
+                    re.escape(f"agg function failed [how->{method},dtype->object]"),
+                    re.escape(f"agg function failed [how->{method},dtype->string]"),
+                ]
+            )
+            with pytest.raises(exception, match=msg):
+                getattr(gb, method)()
+        else:
+            result = getattr(gb, method)()
+            tm.assert_index_equal(result.columns, expected_columns_numeric)
+
+        if method not in ("first", "last"):
+            msg = "|".join(
+                [
+                    "Categorical is not ordered",
+                    "category type does not support",
+                    "function is not implemented for this dtype",
+                    f"Cannot perform {method} with non-ordered Categorical",
+                    re.escape(f"agg function failed [how->{method},dtype->object]"),
+                    re.escape(f"agg function failed [how->{method},dtype->string]"),
+                ]
+            )
+            with pytest.raises(exception, match=msg):
+                getattr(gb, method)(numeric_only=False)
+        else:
+            result = getattr(gb, method)(numeric_only=False)
+            tm.assert_index_equal(result.columns, expected_columns)
+
+
+@pytest.mark.parametrize("numeric_only", [True, False, None])
+def test_axis1_numeric_only(request, groupby_func, numeric_only, using_infer_string):
+    if groupby_func in ("idxmax", "idxmin"):
+        pytest.skip("idxmax and idx_min tested in test_idxmin_idxmax_axis1")
+    if groupby_func in ("corrwith", "skew"):
+        msg = "GH#47723 groupby.corrwith and skew do not correctly implement axis=1"
+        request.applymarker(pytest.mark.xfail(reason=msg))
+
+    df = DataFrame(
+        np.random.default_rng(2).standard_normal((10, 4)), columns=["A", "B", "C", "D"]
+    )
+    df["E"] = "x"
+    groups = [1, 2, 3, 1, 2, 3, 1, 2, 3, 4]
+    gb = df.groupby(groups)
+    method = getattr(gb, groupby_func)
+    args = get_groupby_method_args(groupby_func, df)
+    kwargs = {"axis": 1}
+    if numeric_only is not None:
+        # when numeric_only is None we don't pass any argument
+        kwargs["numeric_only"] = numeric_only
+
+    # Functions without numeric_only and axis args
+    no_args = ("cumprod", "cumsum", "diff", "fillna", "pct_change", "rank", "shift")
+    # Functions with axis args
+    has_axis = (
+        "cumprod",
+        "cumsum",
+        "diff",
+        "pct_change",
+        "rank",
+        "shift",
+        "cummax",
+        "cummin",
+        "idxmin",
+        "idxmax",
+        "fillna",
+    )
+    warn_msg = f"DataFrameGroupBy.{groupby_func} with axis=1 is deprecated"
+    if numeric_only is not None and groupby_func in no_args:
+        msg = "got an unexpected keyword argument 'numeric_only'"
+        if groupby_func in ["cumprod", "cumsum"]:
+            with pytest.raises(TypeError, match=msg):
+                with tm.assert_produces_warning(FutureWarning, match=warn_msg):
+                    method(*args, **kwargs)
+        else:
+            with pytest.raises(TypeError, match=msg):
+                method(*args, **kwargs)
+    elif groupby_func not in has_axis:
+        msg = "got an unexpected keyword argument 'axis'"
+        with pytest.raises(TypeError, match=msg):
+            method(*args, **kwargs)
+    # fillna and shift are successful even on object dtypes
+    elif (numeric_only is None or not numeric_only) and groupby_func not in (
+        "fillna",
+        "shift",
+    ):
+        msgs = (
+            # cummax, cummin, rank
+            "not supported between instances of",
+            # cumprod
+            "can't multiply sequence by non-int of type 'float'",
+            # cumsum, diff, pct_change
+            "unsupported operand type",
+            "has no kernel",
+        )
+        if using_infer_string:
+            import pyarrow as pa
+
+            errs = (TypeError, pa.lib.ArrowNotImplementedError)
+        else:
+            errs = TypeError
+        with pytest.raises(errs, match=f"({'|'.join(msgs)})"):
+            with tm.assert_produces_warning(FutureWarning, match=warn_msg):
+                method(*args, **kwargs)
+    else:
+        with tm.assert_produces_warning(FutureWarning, match=warn_msg):
+            result = method(*args, **kwargs)
+
+        df_expected = df.drop(columns="E").T if numeric_only else df.T
+        expected = getattr(df_expected, groupby_func)(*args).T
+        if groupby_func == "shift" and not numeric_only:
+            # shift with axis=1 leaves the leftmost column as numeric
+            # but transposing for expected gives us object dtype
+            expected = expected.astype(float)
+
+        tm.assert_equal(result, expected)
+
+
+@pytest.mark.parametrize(
+    "kernel, has_arg",
+    [
+        ("all", False),
+        ("any", False),
+        ("bfill", False),
+        ("corr", True),
+        ("corrwith", True),
+        ("cov", True),
+        ("cummax", True),
+        ("cummin", True),
+        ("cumprod", True),
+        ("cumsum", True),
+        ("diff", False),
+        ("ffill", False),
+        ("fillna", False),
+        ("first", True),
+        ("idxmax", True),
+        ("idxmin", True),
+        ("last", True),
+        ("max", True),
+        ("mean", True),
+        ("median", True),
+        ("min", True),
+        ("nth", False),
+        ("nunique", False),
+        ("pct_change", False),
+        ("prod", True),
+        ("quantile", True),
+        ("sem", True),
+        ("skew", True),
+        ("std", True),
+        ("sum", True),
+        ("var", True),
+    ],
+)
+@pytest.mark.parametrize("numeric_only", [True, False, lib.no_default])
+@pytest.mark.parametrize("keys", [["a1"], ["a1", "a2"]])
+def test_numeric_only(kernel, has_arg, numeric_only, keys):
+    # GH#46072
+    # drops_nuisance: Whether the op drops nuisance columns even when numeric_only=False
+    # has_arg: Whether the op has a numeric_only arg
+    df = DataFrame({"a1": [1, 1], "a2": [2, 2], "a3": [5, 6], "b": 2 * [object]})
+
+    args = get_groupby_method_args(kernel, df)
+    kwargs = {} if numeric_only is lib.no_default else {"numeric_only": numeric_only}
+
+    gb = df.groupby(keys)
+    method = getattr(gb, kernel)
+    if has_arg and numeric_only is True:
+        # Cases where b does not appear in the result
+        result = method(*args, **kwargs)
+        assert "b" not in result.columns
+    elif (
+        # kernels that work on any dtype and have numeric_only arg
+        kernel in ("first", "last")
+        or (
+            # kernels that work on any dtype and don't have numeric_only arg
+            kernel in ("any", "all", "bfill", "ffill", "fillna", "nth", "nunique")
+            and numeric_only is lib.no_default
+        )
+    ):
+        warn = FutureWarning if kernel == "fillna" else None
+        msg = "DataFrameGroupBy.fillna is deprecated"
+        with tm.assert_produces_warning(warn, match=msg):
+            result = method(*args, **kwargs)
+        assert "b" in result.columns
+    elif has_arg:
+        assert numeric_only is not True
+        # kernels that are successful on any dtype were above; this will fail
+
+        # object dtypes for transformations are not implemented in Cython and
+        # have no Python fallback
+        exception = NotImplementedError if kernel.startswith("cum") else TypeError
+
+        msg = "|".join(
+            [
+                "not allowed for this dtype",
+                "cannot be performed against 'object' dtypes",
+                # On PY39 message is "a number"; on PY310 and after is "a real number"
+                "must be a string or a.* number",
+                "unsupported operand type",
+                "function is not implemented for this dtype",
+                re.escape(f"agg function failed [how->{kernel},dtype->object]"),
+            ]
+        )
+        if kernel == "idxmin":
+            msg = "'<' not supported between instances of 'type' and 'type'"
+        elif kernel == "idxmax":
+            msg = "'>' not supported between instances of 'type' and 'type'"
+        with pytest.raises(exception, match=msg):
+            method(*args, **kwargs)
+    elif not has_arg and numeric_only is not lib.no_default:
+        with pytest.raises(
+            TypeError, match="got an unexpected keyword argument 'numeric_only'"
+        ):
+            method(*args, **kwargs)
+    else:
+        assert kernel in ("diff", "pct_change")
+        assert numeric_only is lib.no_default
+        # Doesn't have numeric_only argument and fails on nuisance columns
+        with pytest.raises(TypeError, match=r"unsupported operand type"):
+            method(*args, **kwargs)
+
+
+@pytest.mark.filterwarnings("ignore:Downcasting object dtype arrays:FutureWarning")
+@pytest.mark.parametrize("dtype", [bool, int, float, object])
+def test_deprecate_numeric_only_series(dtype, groupby_func, request):
+    # GH#46560
+    grouper = [0, 0, 1]
+
+    ser = Series([1, 0, 0], dtype=dtype)
+    gb = ser.groupby(grouper)
+
+    if groupby_func == "corrwith":
+        # corrwith is not implemented on SeriesGroupBy
+        assert not hasattr(gb, groupby_func)
+        return
+
+    method = getattr(gb, groupby_func)
+
+    expected_ser = Series([1, 0, 0])
+    expected_gb = expected_ser.groupby(grouper)
+    expected_method = getattr(expected_gb, groupby_func)
+
+    args = get_groupby_method_args(groupby_func, ser)
+
+    fails_on_numeric_object = (
+        "corr",
+        "cov",
+        "cummax",
+        "cummin",
+        "cumprod",
+        "cumsum",
+        "quantile",
+    )
+    # ops that give an object result on object input
+    obj_result = (
+        "first",
+        "last",
+        "nth",
+        "bfill",
+        "ffill",
+        "shift",
+        "sum",
+        "diff",
+        "pct_change",
+        "var",
+        "mean",
+        "median",
+        "min",
+        "max",
+        "prod",
+        "skew",
+    )
+
+    # Test default behavior; kernels that fail may be enabled in the future but kernels
+    # that succeed should not be allowed to fail (without deprecation, at least)
+    if groupby_func in fails_on_numeric_object and dtype is object:
+        if groupby_func == "quantile":
+            msg = "cannot be performed against 'object' dtypes"
+        else:
+            msg = "is not supported for object dtype"
+        warn = FutureWarning if groupby_func == "fillna" else None
+        warn_msg = "DataFrameGroupBy.fillna is deprecated"
+        with tm.assert_produces_warning(warn, match=warn_msg):
+            with pytest.raises(TypeError, match=msg):
+                method(*args)
+    elif dtype is object:
+        warn = FutureWarning if groupby_func == "fillna" else None
+        warn_msg = "SeriesGroupBy.fillna is deprecated"
+        with tm.assert_produces_warning(warn, match=warn_msg):
+            result = method(*args)
+        with tm.assert_produces_warning(warn, match=warn_msg):
+            expected = expected_method(*args)
+        if groupby_func in obj_result:
+            expected = expected.astype(object)
+        tm.assert_series_equal(result, expected)
+
+    has_numeric_only = (
+        "first",
+        "last",
+        "max",
+        "mean",
+        "median",
+        "min",
+        "prod",
+        "quantile",
+        "sem",
+        "skew",
+        "std",
+        "sum",
+        "var",
+        "cummax",
+        "cummin",
+        "cumprod",
+        "cumsum",
+    )
+    if groupby_func not in has_numeric_only:
+        msg = "got an unexpected keyword argument 'numeric_only'"
+        with pytest.raises(TypeError, match=msg):
+            method(*args, numeric_only=True)
+    elif dtype is object:
+        msg = "|".join(
+            [
+                "SeriesGroupBy.sem called with numeric_only=True and dtype object",
+                "Series.skew does not allow numeric_only=True with non-numeric",
+                "cum(sum|prod|min|max) is not supported for object dtype",
+                r"Cannot use numeric_only=True with SeriesGroupBy\..* and non-numeric",
+            ]
+        )
+        with pytest.raises(TypeError, match=msg):
+            method(*args, numeric_only=True)
+    elif dtype == bool and groupby_func == "quantile":
+        msg = "Allowing bool dtype in SeriesGroupBy.quantile"
+        with tm.assert_produces_warning(FutureWarning, match=msg):
+            # GH#51424
+            result = method(*args, numeric_only=True)
+            expected = method(*args, numeric_only=False)
+        tm.assert_series_equal(result, expected)
+    else:
+        result = method(*args, numeric_only=True)
+        expected = method(*args, numeric_only=False)
+        tm.assert_series_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_pipe.py

@@ -0,0 +1,80 @@
+import numpy as np
+
+import pandas as pd
+from pandas import (
+    DataFrame,
+    Index,
+)
+import pandas._testing as tm
+
+
+def test_pipe():
+    # Test the pipe method of DataFrameGroupBy.
+    # Issue #17871
+
+    random_state = np.random.default_rng(2)
+
+    df = DataFrame(
+        {
+            "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"],
+            "B": random_state.standard_normal(8),
+            "C": random_state.standard_normal(8),
+        }
+    )
+
+    def f(dfgb):
+        return dfgb.B.max() - dfgb.C.min().min()
+
+    def square(srs):
+        return srs**2
+
+    # Note that the transformations are
+    # GroupBy -> Series
+    # Series -> Series
+    # This then chains the GroupBy.pipe and the
+    # NDFrame.pipe methods
+    result = df.groupby("A").pipe(f).pipe(square)
+
+    index = Index(["bar", "foo"], dtype="object", name="A")
+    expected = pd.Series([3.749306591013693, 6.717707873081384], name="B", index=index)
+
+    tm.assert_series_equal(expected, result)
+
+
+def test_pipe_args():
+    # Test passing args to the pipe method of DataFrameGroupBy.
+    # Issue #17871
+
+    df = DataFrame(
+        {
+            "group": ["A", "A", "B", "B", "C"],
+            "x": [1.0, 2.0, 3.0, 2.0, 5.0],
+            "y": [10.0, 100.0, 1000.0, -100.0, -1000.0],
+        }
+    )
+
+    def f(dfgb, arg1):
+        filtered = dfgb.filter(lambda grp: grp.y.mean() > arg1, dropna=False)
+        return filtered.groupby("group")
+
+    def g(dfgb, arg2):
+        return dfgb.sum() / dfgb.sum().sum() + arg2
+
+    def h(df, arg3):
+        return df.x + df.y - arg3
+
+    result = df.groupby("group").pipe(f, 0).pipe(g, 10).pipe(h, 100)
+
+    # Assert the results here
+    index = Index(["A", "B"], name="group")
+    expected = pd.Series([-79.5160891089, -78.4839108911], index=index)
+
+    tm.assert_series_equal(result, expected)
+
+    # test SeriesGroupby.pipe
+    ser = pd.Series([1, 1, 2, 2, 3, 3])
+    result = ser.groupby(ser).pipe(lambda grp: grp.sum() * grp.count())
+
+    expected = pd.Series([4, 8, 12], index=Index([1, 2, 3], dtype=np.int64))
+
+    tm.assert_series_equal(result, expected)

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_raises.py

@@ -0,0 +1,716 @@
+# Only tests that raise an error and have no better location should go here.
+# Tests for specific groupby methods should go in their respective
+# test file.
+
+import datetime
+import re
+
+import numpy as np
+import pytest
+
+from pandas import (
+    Categorical,
+    DataFrame,
+    Grouper,
+    Series,
+)
+import pandas._testing as tm
+from pandas.tests.groupby import get_groupby_method_args
+
+
+@pytest.fixture(
+    params=[
+        "a",
+        ["a"],
+        ["a", "b"],
+        Grouper(key="a"),
+        lambda x: x % 2,
+        [0, 0, 0, 1, 2, 2, 2, 3, 3],
+        np.array([0, 0, 0, 1, 2, 2, 2, 3, 3]),
+        dict(zip(range(9), [0, 0, 0, 1, 2, 2, 2, 3, 3])),
+        Series([1, 1, 1, 1, 1, 2, 2, 2, 2]),
+        [Series([1, 1, 1, 1, 1, 2, 2, 2, 2]), Series([3, 3, 4, 4, 4, 4, 4, 3, 3])],
+    ]
+)
+def by(request):
+    return request.param
+
+
+@pytest.fixture(params=[True, False])
+def groupby_series(request):
+    return request.param
+
+
+@pytest.fixture
+def df_with_string_col():
+    df = DataFrame(
+        {
+            "a": [1, 1, 1, 1, 1, 2, 2, 2, 2],
+            "b": [3, 3, 4, 4, 4, 4, 4, 3, 3],
+            "c": range(9),
+            "d": list("xyzwtyuio"),
+        }
+    )
+    return df
+
+
+@pytest.fixture
+def df_with_datetime_col():
+    df = DataFrame(
+        {
+            "a": [1, 1, 1, 1, 1, 2, 2, 2, 2],
+            "b": [3, 3, 4, 4, 4, 4, 4, 3, 3],
+            "c": range(9),
+            "d": datetime.datetime(2005, 1, 1, 10, 30, 23, 540000),
+        }
+    )
+    return df
+
+
+@pytest.fixture
+def df_with_timedelta_col():
+    df = DataFrame(
+        {
+            "a": [1, 1, 1, 1, 1, 2, 2, 2, 2],
+            "b": [3, 3, 4, 4, 4, 4, 4, 3, 3],
+            "c": range(9),
+            "d": datetime.timedelta(days=1),
+        }
+    )
+    return df
+
+
+@pytest.fixture
+def df_with_cat_col():
+    df = DataFrame(
+        {
+            "a": [1, 1, 1, 1, 1, 2, 2, 2, 2],
+            "b": [3, 3, 4, 4, 4, 4, 4, 3, 3],
+            "c": range(9),
+            "d": Categorical(
+                ["a", "a", "a", "a", "b", "b", "b", "b", "c"],
+                categories=["a", "b", "c", "d"],
+                ordered=True,
+            ),
+        }
+    )
+    return df
+
+
+def _call_and_check(klass, msg, how, gb, groupby_func, args, warn_msg=""):
+    warn_klass = None if warn_msg == "" else FutureWarning
+    with tm.assert_produces_warning(warn_klass, match=warn_msg):
+        if klass is None:
+            if how == "method":
+                getattr(gb, groupby_func)(*args)
+            elif how == "agg":
+                gb.agg(groupby_func, *args)
+            else:
+                gb.transform(groupby_func, *args)
+        else:
+            with pytest.raises(klass, match=msg):
+                if how == "method":
+                    getattr(gb, groupby_func)(*args)
+                elif how == "agg":
+                    gb.agg(groupby_func, *args)
+                else:
+                    gb.transform(groupby_func, *args)
+
+
+@pytest.mark.parametrize("how", ["method", "agg", "transform"])
+def test_groupby_raises_string(
+    how, by, groupby_series, groupby_func, df_with_string_col
+):
+    df = df_with_string_col
+    args = get_groupby_method_args(groupby_func, df)
+    gb = df.groupby(by=by)
+
+    if groupby_series:
+        gb = gb["d"]
+
+        if groupby_func == "corrwith":
+            assert not hasattr(gb, "corrwith")
+            return
+
+    klass, msg = {
+        "all": (None, ""),
+        "any": (None, ""),
+        "bfill": (None, ""),
+        "corrwith": (TypeError, "Could not convert"),
+        "count": (None, ""),
+        "cumcount": (None, ""),
+        "cummax": (
+            (NotImplementedError, TypeError),
+            "(function|cummax) is not (implemented|supported) for (this|object) dtype",
+        ),
+        "cummin": (
+            (NotImplementedError, TypeError),
+            "(function|cummin) is not (implemented|supported) for (this|object) dtype",
+        ),
+        "cumprod": (
+            (NotImplementedError, TypeError),
+            "(function|cumprod) is not (implemented|supported) for (this|object) dtype",
+        ),
+        "cumsum": (
+            (NotImplementedError, TypeError),
+            "(function|cumsum) is not (implemented|supported) for (this|object) dtype",
+        ),
+        "diff": (TypeError, "unsupported operand type"),
+        "ffill": (None, ""),
+        "fillna": (None, ""),
+        "first": (None, ""),
+        "idxmax": (None, ""),
+        "idxmin": (None, ""),
+        "last": (None, ""),
+        "max": (None, ""),
+        "mean": (
+            TypeError,
+            re.escape("agg function failed [how->mean,dtype->object]"),
+        ),
+        "median": (
+            TypeError,
+            re.escape("agg function failed [how->median,dtype->object]"),
+        ),
+        "min": (None, ""),
+        "ngroup": (None, ""),
+        "nunique": (None, ""),
+        "pct_change": (TypeError, "unsupported operand type"),
+        "prod": (
+            TypeError,
+            re.escape("agg function failed [how->prod,dtype->object]"),
+        ),
+        "quantile": (TypeError, "cannot be performed against 'object' dtypes!"),
+        "rank": (None, ""),
+        "sem": (ValueError, "could not convert string to float"),
+        "shift": (None, ""),
+        "size": (None, ""),
+        "skew": (ValueError, "could not convert string to float"),
+        "std": (ValueError, "could not convert string to float"),
+        "sum": (None, ""),
+        "var": (
+            TypeError,
+            re.escape("agg function failed [how->var,dtype->"),
+        ),
+    }[groupby_func]
+
+    if groupby_func == "fillna":
+        kind = "Series" if groupby_series else "DataFrame"
+        warn_msg = f"{kind}GroupBy.fillna is deprecated"
+    else:
+        warn_msg = ""
+    _call_and_check(klass, msg, how, gb, groupby_func, args, warn_msg)
+
+
+@pytest.mark.parametrize("how", ["agg", "transform"])
+def test_groupby_raises_string_udf(how, by, groupby_series, df_with_string_col):
+    df = df_with_string_col
+    gb = df.groupby(by=by)
+
+    if groupby_series:
+        gb = gb["d"]
+
+    def func(x):
+        raise TypeError("Test error message")
+
+    with pytest.raises(TypeError, match="Test error message"):
+        getattr(gb, how)(func)
+
+
+@pytest.mark.parametrize("how", ["agg", "transform"])
+@pytest.mark.parametrize("groupby_func_np", [np.sum, np.mean])
+def test_groupby_raises_string_np(
+    how, by, groupby_series, groupby_func_np, df_with_string_col
+):
+    # GH#50749
+    df = df_with_string_col
+    gb = df.groupby(by=by)
+
+    if groupby_series:
+        gb = gb["d"]
+
+    klass, msg = {
+        np.sum: (None, ""),
+        np.mean: (
+            TypeError,
+            re.escape("agg function failed [how->mean,dtype->object]"),
+        ),
+    }[groupby_func_np]
+
+    if groupby_series:
+        warn_msg = "using SeriesGroupBy.[sum|mean]"
+    else:
+        warn_msg = "using DataFrameGroupBy.[sum|mean]"
+    _call_and_check(klass, msg, how, gb, groupby_func_np, (), warn_msg=warn_msg)
+
+
+@pytest.mark.parametrize("how", ["method", "agg", "transform"])
+def test_groupby_raises_datetime(
+    how, by, groupby_series, groupby_func, df_with_datetime_col
+):
+    df = df_with_datetime_col
+    args = get_groupby_method_args(groupby_func, df)
+    gb = df.groupby(by=by)
+
+    if groupby_series:
+        gb = gb["d"]
+
+        if groupby_func == "corrwith":
+            assert not hasattr(gb, "corrwith")
+            return
+
+    klass, msg = {
+        "all": (None, ""),
+        "any": (None, ""),
+        "bfill": (None, ""),
+        "corrwith": (TypeError, "cannot perform __mul__ with this index type"),
+        "count": (None, ""),
+        "cumcount": (None, ""),
+        "cummax": (None, ""),
+        "cummin": (None, ""),
+        "cumprod": (TypeError, "datetime64 type does not support cumprod operations"),
+        "cumsum": (TypeError, "datetime64 type does not support cumsum operations"),
+        "diff": (None, ""),
+        "ffill": (None, ""),
+        "fillna": (None, ""),
+        "first": (None, ""),
+        "idxmax": (None, ""),
+        "idxmin": (None, ""),
+        "last": (None, ""),
+        "max": (None, ""),
+        "mean": (None, ""),
+        "median": (None, ""),
+        "min": (None, ""),
+        "ngroup": (None, ""),
+        "nunique": (None, ""),
+        "pct_change": (TypeError, "cannot perform __truediv__ with this index type"),
+        "prod": (TypeError, "datetime64 type does not support prod"),
+        "quantile": (None, ""),
+        "rank": (None, ""),
+        "sem": (None, ""),
+        "shift": (None, ""),
+        "size": (None, ""),
+        "skew": (
+            TypeError,
+            "|".join(
+                [
+                    r"dtype datetime64\[ns\] does not support reduction",
+                    "datetime64 type does not support skew operations",
+                ]
+            ),
+        ),
+        "std": (None, ""),
+        "sum": (TypeError, "datetime64 type does not support sum operations"),
+        "var": (TypeError, "datetime64 type does not support var operations"),
+    }[groupby_func]
+
+    if groupby_func in ["any", "all"]:
+        warn_msg = f"'{groupby_func}' with datetime64 dtypes is deprecated"
+    elif groupby_func == "fillna":
+        kind = "Series" if groupby_series else "DataFrame"
+        warn_msg = f"{kind}GroupBy.fillna is deprecated"
+    else:
+        warn_msg = ""
+    _call_and_check(klass, msg, how, gb, groupby_func, args, warn_msg=warn_msg)
+
+
+@pytest.mark.parametrize("how", ["agg", "transform"])
+def test_groupby_raises_datetime_udf(how, by, groupby_series, df_with_datetime_col):
+    df = df_with_datetime_col
+    gb = df.groupby(by=by)
+
+    if groupby_series:
+        gb = gb["d"]
+
+    def func(x):
+        raise TypeError("Test error message")
+
+    with pytest.raises(TypeError, match="Test error message"):
+        getattr(gb, how)(func)
+
+
+@pytest.mark.parametrize("how", ["agg", "transform"])
+@pytest.mark.parametrize("groupby_func_np", [np.sum, np.mean])
+def test_groupby_raises_datetime_np(
+    how, by, groupby_series, groupby_func_np, df_with_datetime_col
+):
+    # GH#50749
+    df = df_with_datetime_col
+    gb = df.groupby(by=by)
+
+    if groupby_series:
+        gb = gb["d"]
+
+    klass, msg = {
+        np.sum: (TypeError, "datetime64 type does not support sum operations"),
+        np.mean: (None, ""),
+    }[groupby_func_np]
+
+    if groupby_series:
+        warn_msg = "using SeriesGroupBy.[sum|mean]"
+    else:
+        warn_msg = "using DataFrameGroupBy.[sum|mean]"
+    _call_and_check(klass, msg, how, gb, groupby_func_np, (), warn_msg=warn_msg)
+
+
+@pytest.mark.parametrize("func", ["prod", "cumprod", "skew", "var"])
+def test_groupby_raises_timedelta(func, df_with_timedelta_col):
+    df = df_with_timedelta_col
+    gb = df.groupby(by="a")
+
+    _call_and_check(
+        TypeError,
+        "timedelta64 type does not support .* operations",
+        "method",
+        gb,
+        func,
+        [],
+    )
+
+
+@pytest.mark.parametrize("how", ["method", "agg", "transform"])
+def test_groupby_raises_category(
+    how, by, groupby_series, groupby_func, using_copy_on_write, df_with_cat_col
+):
+    # GH#50749
+    df = df_with_cat_col
+    args = get_groupby_method_args(groupby_func, df)
+    gb = df.groupby(by=by)
+
+    if groupby_series:
+        gb = gb["d"]
+
+        if groupby_func == "corrwith":
+            assert not hasattr(gb, "corrwith")
+            return
+
+    klass, msg = {
+        "all": (None, ""),
+        "any": (None, ""),
+        "bfill": (None, ""),
+        "corrwith": (
+            TypeError,
+            r"unsupported operand type\(s\) for \*: 'Categorical' and 'int'",
+        ),
+        "count": (None, ""),
+        "cumcount": (None, ""),
+        "cummax": (
+            (NotImplementedError, TypeError),
+            "(category type does not support cummax operations|"
+            "category dtype not supported|"
+            "cummax is not supported for category dtype)",
+        ),
+        "cummin": (
+            (NotImplementedError, TypeError),
+            "(category type does not support cummin operations|"
+            "category dtype not supported|"
+            "cummin is not supported for category dtype)",
+        ),
+        "cumprod": (
+            (NotImplementedError, TypeError),
+            "(category type does not support cumprod operations|"
+            "category dtype not supported|"
+            "cumprod is not supported for category dtype)",
+        ),
+        "cumsum": (
+            (NotImplementedError, TypeError),
+            "(category type does not support cumsum operations|"
+            "category dtype not supported|"
+            "cumsum is not supported for category dtype)",
+        ),
+        "diff": (
+            TypeError,
+            r"unsupported operand type\(s\) for -: 'Categorical' and 'Categorical'",
+        ),
+        "ffill": (None, ""),
+        "fillna": (
+            TypeError,
+            r"Cannot setitem on a Categorical with a new category \(0\), "
+            "set the categories first",
+        )
+        if not using_copy_on_write
+        else (None, ""),  # no-op with CoW
+        "first": (None, ""),
+        "idxmax": (None, ""),
+        "idxmin": (None, ""),
+        "last": (None, ""),
+        "max": (None, ""),
+        "mean": (
+            TypeError,
+            "|".join(
+                [
+                    "'Categorical' .* does not support reduction 'mean'",
+                    "category dtype does not support aggregation 'mean'",
+                ]
+            ),
+        ),
+        "median": (
+            TypeError,
+            "|".join(
+                [
+                    "'Categorical' .* does not support reduction 'median'",
+                    "category dtype does not support aggregation 'median'",
+                ]
+            ),
+        ),
+        "min": (None, ""),
+        "ngroup": (None, ""),
+        "nunique": (None, ""),
+        "pct_change": (
+            TypeError,
+            r"unsupported operand type\(s\) for /: 'Categorical' and 'Categorical'",
+        ),
+        "prod": (TypeError, "category type does not support prod operations"),
+        "quantile": (TypeError, "No matching signature found"),
+        "rank": (None, ""),
+        "sem": (
+            TypeError,
+            "|".join(
+                [
+                    "'Categorical' .* does not support reduction 'sem'",
+                    "category dtype does not support aggregation 'sem'",
+                ]
+            ),
+        ),
+        "shift": (None, ""),
+        "size": (None, ""),
+        "skew": (
+            TypeError,
+            "|".join(
+                [
+                    "dtype category does not support reduction 'skew'",
+                    "category type does not support skew operations",
+                ]
+            ),
+        ),
+        "std": (
+            TypeError,
+            "|".join(
+                [
+                    "'Categorical' .* does not support reduction 'std'",
+                    "category dtype does not support aggregation 'std'",
+                ]
+            ),
+        ),
+        "sum": (TypeError, "category type does not support sum operations"),
+        "var": (
+            TypeError,
+            "|".join(
+                [
+                    "'Categorical' .* does not support reduction 'var'",
+                    "category dtype does not support aggregation 'var'",
+                ]
+            ),
+        ),
+    }[groupby_func]
+
+    if groupby_func == "fillna":
+        kind = "Series" if groupby_series else "DataFrame"
+        warn_msg = f"{kind}GroupBy.fillna is deprecated"
+    else:
+        warn_msg = ""
+    _call_and_check(klass, msg, how, gb, groupby_func, args, warn_msg)
+
+
+@pytest.mark.parametrize("how", ["agg", "transform"])
+def test_groupby_raises_category_udf(how, by, groupby_series, df_with_cat_col):
+    # GH#50749
+    df = df_with_cat_col
+    gb = df.groupby(by=by)
+
+    if groupby_series:
+        gb = gb["d"]
+
+    def func(x):
+        raise TypeError("Test error message")
+
+    with pytest.raises(TypeError, match="Test error message"):
+        getattr(gb, how)(func)
+
+
+@pytest.mark.parametrize("how", ["agg", "transform"])
+@pytest.mark.parametrize("groupby_func_np", [np.sum, np.mean])
+def test_groupby_raises_category_np(
+    how, by, groupby_series, groupby_func_np, df_with_cat_col
+):
+    # GH#50749
+    df = df_with_cat_col
+    gb = df.groupby(by=by)
+
+    if groupby_series:
+        gb = gb["d"]
+
+    klass, msg = {
+        np.sum: (TypeError, "category type does not support sum operations"),
+        np.mean: (
+            TypeError,
+            "category dtype does not support aggregation 'mean'",
+        ),
+    }[groupby_func_np]
+
+    if groupby_series:
+        warn_msg = "using SeriesGroupBy.[sum|mean]"
+    else:
+        warn_msg = "using DataFrameGroupBy.[sum|mean]"
+    _call_and_check(klass, msg, how, gb, groupby_func_np, (), warn_msg=warn_msg)
+
+
+@pytest.mark.parametrize("how", ["method", "agg", "transform"])
+def test_groupby_raises_category_on_category(
+    how,
+    by,
+    groupby_series,
+    groupby_func,
+    observed,
+    using_copy_on_write,
+    df_with_cat_col,
+):
+    # GH#50749
+    df = df_with_cat_col
+    df["a"] = Categorical(
+        ["a", "a", "a", "a", "b", "b", "b", "b", "c"],
+        categories=["a", "b", "c", "d"],
+        ordered=True,
+    )
+    args = get_groupby_method_args(groupby_func, df)
+    gb = df.groupby(by=by, observed=observed)
+
+    if groupby_series:
+        gb = gb["d"]
+
+        if groupby_func == "corrwith":
+            assert not hasattr(gb, "corrwith")
+            return
+
+    empty_groups = not observed and any(group.empty for group in gb.groups.values())
+    if (
+        not observed
+        and how != "transform"
+        and isinstance(by, list)
+        and isinstance(by[0], str)
+        and by == ["a", "b"]
+    ):
+        assert not empty_groups
+        # TODO: empty_groups should be true due to unobserved categorical combinations
+        empty_groups = True
+    if how == "transform":
+        # empty groups will be ignored
+        empty_groups = False
+
+    klass, msg = {
+        "all": (None, ""),
+        "any": (None, ""),
+        "bfill": (None, ""),
+        "corrwith": (
+            TypeError,
+            r"unsupported operand type\(s\) for \*: 'Categorical' and 'int'",
+        ),
+        "count": (None, ""),
+        "cumcount": (None, ""),
+        "cummax": (
+            (NotImplementedError, TypeError),
+            "(cummax is not supported for category dtype|"
+            "category dtype not supported|"
+            "category type does not support cummax operations)",
+        ),
+        "cummin": (
+            (NotImplementedError, TypeError),
+            "(cummin is not supported for category dtype|"
+            "category dtype not supported|"
+            "category type does not support cummin operations)",
+        ),
+        "cumprod": (
+            (NotImplementedError, TypeError),
+            "(cumprod is not supported for category dtype|"
+            "category dtype not supported|"
+            "category type does not support cumprod operations)",
+        ),
+        "cumsum": (
+            (NotImplementedError, TypeError),
+            "(cumsum is not supported for category dtype|"
+            "category dtype not supported|"
+            "category type does not support cumsum operations)",
+        ),
+        "diff": (TypeError, "unsupported operand type"),
+        "ffill": (None, ""),
+        "fillna": (
+            TypeError,
+            r"Cannot setitem on a Categorical with a new category \(0\), "
+            "set the categories first",
+        )
+        if not using_copy_on_write
+        else (None, ""),  # no-op with CoW
+        "first": (None, ""),
+        "idxmax": (ValueError, "empty group due to unobserved categories")
+        if empty_groups
+        else (None, ""),
+        "idxmin": (ValueError, "empty group due to unobserved categories")
+        if empty_groups
+        else (None, ""),
+        "last": (None, ""),
+        "max": (None, ""),
+        "mean": (TypeError, "category dtype does not support aggregation 'mean'"),
+        "median": (TypeError, "category dtype does not support aggregation 'median'"),
+        "min": (None, ""),
+        "ngroup": (None, ""),
+        "nunique": (None, ""),
+        "pct_change": (TypeError, "unsupported operand type"),
+        "prod": (TypeError, "category type does not support prod operations"),
+        "quantile": (TypeError, ""),
+        "rank": (None, ""),
+        "sem": (
+            TypeError,
+            "|".join(
+                [
+                    "'Categorical' .* does not support reduction 'sem'",
+                    "category dtype does not support aggregation 'sem'",
+                ]
+            ),
+        ),
+        "shift": (None, ""),
+        "size": (None, ""),
+        "skew": (
+            TypeError,
+            "|".join(
+                [
+                    "category type does not support skew operations",
+                    "dtype category does not support reduction 'skew'",
+                ]
+            ),
+        ),
+        "std": (
+            TypeError,
+            "|".join(
+                [
+                    "'Categorical' .* does not support reduction 'std'",
+                    "category dtype does not support aggregation 'std'",
+                ]
+            ),
+        ),
+        "sum": (TypeError, "category type does not support sum operations"),
+        "var": (
+            TypeError,
+            "|".join(
+                [
+                    "'Categorical' .* does not support reduction 'var'",
+                    "category dtype does not support aggregation 'var'",
+                ]
+            ),
+        ),
+    }[groupby_func]
+
+    if groupby_func == "fillna":
+        kind = "Series" if groupby_series else "DataFrame"
+        warn_msg = f"{kind}GroupBy.fillna is deprecated"
+    else:
+        warn_msg = ""
+    _call_and_check(klass, msg, how, gb, groupby_func, args, warn_msg)
+
+
+def test_subsetting_columns_axis_1_raises():
+    # GH 35443
+    df = DataFrame({"a": [1], "b": [2], "c": [3]})
+    msg = "DataFrame.groupby with axis=1 is deprecated"
+    with tm.assert_produces_warning(FutureWarning, match=msg):
+        gb = df.groupby("a", axis=1)
+    with pytest.raises(ValueError, match="Cannot subset columns when using axis=1"):
+        gb["b"]

venv/lib/python3.12/site-packages/pandas/tests/groupby/test_timegrouper.py

@@ -0,0 +1,963 @@
+"""
+test with the TimeGrouper / grouping with datetimes
+"""
+from datetime import (
+    datetime,
+    timedelta,
+)
+
+import numpy as np
+import pytest
+import pytz
+
+import pandas as pd
+from pandas import (
+    DataFrame,
+    DatetimeIndex,
+    Index,
+    MultiIndex,
+    Series,
+    Timestamp,
+    date_range,
+    offsets,
+)
+import pandas._testing as tm
+from pandas.core.groupby.grouper import Grouper
+from pandas.core.groupby.ops import BinGrouper
+
+
+@pytest.fixture
+def frame_for_truncated_bingrouper():
+    """
+    DataFrame used by groupby_with_truncated_bingrouper, made into
+    a separate fixture for easier reuse in
+    test_groupby_apply_timegrouper_with_nat_apply_squeeze
+    """
+    df = DataFrame(
+        {
+            "Quantity": [18, 3, 5, 1, 9, 3],
+            "Date": [
+                Timestamp(2013, 9, 1, 13, 0),
+                Timestamp(2013, 9, 1, 13, 5),
+                Timestamp(2013, 10, 1, 20, 0),
+                Timestamp(2013, 10, 3, 10, 0),
+                pd.NaT,
+                Timestamp(2013, 9, 2, 14, 0),
+            ],
+        }
+    )
+    return df
+
+
+@pytest.fixture
+def groupby_with_truncated_bingrouper(frame_for_truncated_bingrouper):
+    """
+    GroupBy object such that gb._grouper is a BinGrouper and
+    len(gb._grouper.result_index) < len(gb._grouper.group_keys_seq)
+
+    Aggregations on this groupby should have
+
+        dti = date_range("2013-09-01", "2013-10-01", freq="5D", name="Date")
+
+    As either the index or an index level.
+    """
+    df = frame_for_truncated_bingrouper
+
+    tdg = Grouper(key="Date", freq="5D")
+    gb = df.groupby(tdg)
+
+    # check we're testing the case we're interested in
+    assert len(gb._grouper.result_index) != len(gb._grouper.group_keys_seq)
+
+    return gb
+
+
+class TestGroupBy:
+    def test_groupby_with_timegrouper(self):
+        # GH 4161
+        # TimeGrouper requires a sorted index
+        # also verifies that the resultant index has the correct name
+        df_original = DataFrame(
+            {
+                "Buyer": "Carl Carl Carl Carl Joe Carl".split(),
+                "Quantity": [18, 3, 5, 1, 9, 3],
+                "Date": [
+                    datetime(2013, 9, 1, 13, 0),
+                    datetime(2013, 9, 1, 13, 5),
+                    datetime(2013, 10, 1, 20, 0),
+                    datetime(2013, 10, 3, 10, 0),
+                    datetime(2013, 12, 2, 12, 0),
+                    datetime(2013, 9, 2, 14, 0),
+                ],
+            }
+        )
+
+        # GH 6908 change target column's order
+        df_reordered = df_original.sort_values(by="Quantity")
+
+        for df in [df_original, df_reordered]:
+            df = df.set_index(["Date"])
+
+            exp_dti = date_range(
+                "20130901",
+                "20131205",
+                freq="5D",
+                name="Date",
+                inclusive="left",
+                unit=df.index.unit,
+            )
+            expected = DataFrame(
+                {"Buyer": 0, "Quantity": 0},
+                index=exp_dti,
+            )
+            # Cast to object to avoid implicit cast when setting entry to "CarlCarlCarl"
+            expected = expected.astype({"Buyer": object})
+            expected.iloc[0, 0] = "CarlCarlCarl"
+            expected.iloc[6, 0] = "CarlCarl"
+            expected.iloc[18, 0] = "Joe"
+            expected.iloc[[0, 6, 18], 1] = np.array([24, 6, 9], dtype="int64")
+
+            result1 = df.resample("5D").sum()
+            tm.assert_frame_equal(result1, expected)
+
+            df_sorted = df.sort_index()
+            result2 = df_sorted.groupby(Grouper(freq="5D")).sum()
+            tm.assert_frame_equal(result2, expected)
+
+            result3 = df.groupby(Grouper(freq="5D")).sum()
+            tm.assert_frame_equal(result3, expected)
+
+    @pytest.mark.parametrize("should_sort", [True, False])
+    def test_groupby_with_timegrouper_methods(self, should_sort):
+        # GH 3881
+        # make sure API of timegrouper conforms
+
+        df = DataFrame(
+            {
+                "Branch": "A A A A A B".split(),
+                "Buyer": "Carl Mark Carl Joe Joe Carl".split(),
+                "Quantity": [1, 3, 5, 8, 9, 3],
+                "Date": [
+                    datetime(2013, 1, 1, 13, 0),
+                    datetime(2013, 1, 1, 13, 5),
+                    datetime(2013, 10, 1, 20, 0),
+                    datetime(2013, 10, 2, 10, 0),
+                    datetime(2013, 12, 2, 12, 0),
+                    datetime(2013, 12, 2, 14, 0),
+                ],
+            }
+        )
+
+        if should_sort:
+            df = df.sort_values(by="Quantity", ascending=False)
+
+        df = df.set_index("Date", drop=False)
+        g = df.groupby(Grouper(freq="6ME"))
+        assert g.group_keys
+
+        assert isinstance(g._grouper, BinGrouper)
+        groups = g.groups
+        assert isinstance(groups, dict)
+        assert len(groups) == 3
+
+    def test_timegrouper_with_reg_groups(self):
+        # GH 3794
+        # allow combination of timegrouper/reg groups
+
+        df_original = DataFrame(
+            {
+                "Branch": "A A A A A A A B".split(),
+                "Buyer": "Carl Mark Carl Carl Joe Joe Joe Carl".split(),
+                "Quantity": [1, 3, 5, 1, 8, 1, 9, 3],
+                "Date": [
+                    datetime(2013, 1, 1, 13, 0),
+                    datetime(2013, 1, 1, 13, 5),
+                    datetime(2013, 10, 1, 20, 0),
+                    datetime(2013, 10, 2, 10, 0),
+                    datetime(2013, 10, 1, 20, 0),
+                    datetime(2013, 10, 2, 10, 0),
+                    datetime(2013, 12, 2, 12, 0),
+                    datetime(2013, 12, 2, 14, 0),
+                ],
+            }
+        ).set_index("Date")
+
+        df_sorted = df_original.sort_values(by="Quantity", ascending=False)
+
+        for df in [df_original, df_sorted]:
+            expected = DataFrame(
+                {
+                    "Buyer": "Carl Joe Mark".split(),
+                    "Quantity": [10, 18, 3],
+                    "Date": [
+                        datetime(2013, 12, 31, 0, 0),
+                        datetime(2013, 12, 31, 0, 0),
+                        datetime(2013, 12, 31, 0, 0),
+                    ],
+                }
+            ).set_index(["Date", "Buyer"])
+
+            msg = "The default value of numeric_only"
+            result = df.groupby([Grouper(freq="YE"), "Buyer"]).sum(numeric_only=True)
+            tm.assert_frame_equal(result, expected)
+
+            expected = DataFrame(
+                {
+                    "Buyer": "Carl Mark Carl Joe".split(),
+                    "Quantity": [1, 3, 9, 18],
+                    "Date": [
+                        datetime(2013, 1, 1, 0, 0),
+                        datetime(2013, 1, 1, 0, 0),
+                        datetime(2013, 7, 1, 0, 0),
+                        datetime(2013, 7, 1, 0, 0),
+                    ],
+                }
+            ).set_index(["Date", "Buyer"])
+            result = df.groupby([Grouper(freq="6MS"), "Buyer"]).sum(numeric_only=True)
+            tm.assert_frame_equal(result, expected)
+
+        df_original = DataFrame(
+            {
+                "Branch": "A A A A A A A B".split(),
+                "Buyer": "Carl Mark Carl Carl Joe Joe Joe Carl".split(),
+                "Quantity": [1, 3, 5, 1, 8, 1, 9, 3],
+                "Date": [
+                    datetime(2013, 10, 1, 13, 0),
+                    datetime(2013, 10, 1, 13, 5),
+                    datetime(2013, 10, 1, 20, 0),
+                    datetime(2013, 10, 2, 10, 0),
+                    datetime(2013, 10, 1, 20, 0),
+                    datetime(2013, 10, 2, 10, 0),
+                    datetime(2013, 10, 2, 12, 0),
+                    datetime(2013, 10, 2, 14, 0),
+                ],
+            }
+        ).set_index("Date")
+
+        df_sorted = df_original.sort_values(by="Quantity", ascending=False)
+        for df in [df_original, df_sorted]:
+            expected = DataFrame(
+                {
+                    "Buyer": "Carl Joe Mark Carl Joe".split(),
+                    "Quantity": [6, 8, 3, 4, 10],
+                    "Date": [
+                        datetime(2013, 10, 1, 0, 0),
+                        datetime(2013, 10, 1, 0, 0),
+                        datetime(2013, 10, 1, 0, 0),
+                        datetime(2013, 10, 2, 0, 0),
+                        datetime(2013, 10, 2, 0, 0),
+                    ],
+                }
+            ).set_index(["Date", "Buyer"])
+
+            result = df.groupby([Grouper(freq="1D"), "Buyer"]).sum(numeric_only=True)
+            tm.assert_frame_equal(result, expected)
+
+            result = df.groupby([Grouper(freq="1ME"), "Buyer"]).sum(numeric_only=True)
+            expected = DataFrame(
+                {
+                    "Buyer": "Carl Joe Mark".split(),
+                    "Quantity": [10, 18, 3],
+                    "Date": [
+                        datetime(2013, 10, 31, 0, 0),
+                        datetime(2013, 10, 31, 0, 0),
+                        datetime(2013, 10, 31, 0, 0),
+                    ],
+                }
+            ).set_index(["Date", "Buyer"])
+            tm.assert_frame_equal(result, expected)
+
+            # passing the name
+            df = df.reset_index()
+            result = df.groupby([Grouper(freq="1ME", key="Date"), "Buyer"]).sum(
+                numeric_only=True
+            )
+            tm.assert_frame_equal(result, expected)
+
+            with pytest.raises(KeyError, match="'The grouper name foo is not found'"):
+                df.groupby([Grouper(freq="1ME", key="foo"), "Buyer"]).sum()
+
+            # passing the level
+            df = df.set_index("Date")
+            result = df.groupby([Grouper(freq="1ME", level="Date"), "Buyer"]).sum(
+                numeric_only=True
+            )
+            tm.assert_frame_equal(result, expected)
+            result = df.groupby([Grouper(freq="1ME", level=0), "Buyer"]).sum(
+                numeric_only=True
+            )
+            tm.assert_frame_equal(result, expected)
+
+            with pytest.raises(ValueError, match="The level foo is not valid"):
+                df.groupby([Grouper(freq="1ME", level="foo"), "Buyer"]).sum()
+
+            # multi names
+            df = df.copy()
+            df["Date"] = df.index + offsets.MonthEnd(2)
+            result = df.groupby([Grouper(freq="1ME", key="Date"), "Buyer"]).sum(
+                numeric_only=True
+            )
+            expected = DataFrame(
+                {
+                    "Buyer": "Carl Joe Mark".split(),
+                    "Quantity": [10, 18, 3],
+                    "Date": [
+                        datetime(2013, 11, 30, 0, 0),
+                        datetime(2013, 11, 30, 0, 0),
+                        datetime(2013, 11, 30, 0, 0),
+                    ],
+                }
+            ).set_index(["Date", "Buyer"])
+            tm.assert_frame_equal(result, expected)
+
+            # error as we have both a level and a name!
+            msg = "The Grouper cannot specify both a key and a level!"
+            with pytest.raises(ValueError, match=msg):
+                df.groupby(
+                    [Grouper(freq="1ME", key="Date", level="Date"), "Buyer"]
+                ).sum()
+
+            # single groupers
+            expected = DataFrame(
+                [[31]],
+                columns=["Quantity"],
+                index=DatetimeIndex(
+                    [datetime(2013, 10, 31, 0, 0)], freq=offsets.MonthEnd(), name="Date"
+                ),
+            )
+            result = df.groupby(Grouper(freq="1ME")).sum(numeric_only=True)
+            tm.assert_frame_equal(result, expected)
+
+            result = df.groupby([Grouper(freq="1ME")]).sum(numeric_only=True)
+            tm.assert_frame_equal(result, expected)
+
+            expected.index = expected.index.shift(1)
+            assert expected.index.freq == offsets.MonthEnd()
+            result = df.groupby(Grouper(freq="1ME", key="Date")).sum(numeric_only=True)
+            tm.assert_frame_equal(result, expected)
+
+            result = df.groupby([Grouper(freq="1ME", key="Date")]).sum(
+                numeric_only=True
+            )
+            tm.assert_frame_equal(result, expected)
+
+    @pytest.mark.parametrize("freq", ["D", "ME", "YE", "QE-APR"])
+    def test_timegrouper_with_reg_groups_freq(self, freq):
+        # GH 6764 multiple grouping with/without sort
+        df = DataFrame(
+            {
+                "date": pd.to_datetime(
+                    [
+                        "20121002",
+                        "20121007",
+                        "20130130",
+                        "20130202",
+                        "20130305",
+                        "20121002",
+                        "20121207",
+                        "20130130",
+                        "20130202",
+                        "20130305",
+                        "20130202",
+                        "20130305",
+                    ]
+                ),
+                "user_id": [1, 1, 1, 1, 1, 3, 3, 3, 5, 5, 5, 5],
+                "whole_cost": [
+                    1790,
+                    364,
+                    280,
+                    259,
+                    201,
+                    623,
+                    90,
+                    312,
+                    359,
+                    301,
+                    359,
+                    801,
+                ],
+                "cost1": [12, 15, 10, 24, 39, 1, 0, 90, 45, 34, 1, 12],
+            }
+        ).set_index("date")
+
+        expected = (
+            df.groupby("user_id")["whole_cost"]
+            .resample(freq)
+            .sum(min_count=1)  # XXX
+            .dropna()
+            .reorder_levels(["date", "user_id"])
+            .sort_index()
+            .astype("int64")
+        )
+        expected.name = "whole_cost"
+
+        result1 = (
+            df.sort_index().groupby([Grouper(freq=freq), "user_id"])["whole_cost"].sum()
+        )
+        tm.assert_series_equal(result1, expected)
+
+        result2 = df.groupby([Grouper(freq=freq), "user_id"])["whole_cost"].sum()
+        tm.assert_series_equal(result2, expected)
+
+    def test_timegrouper_get_group(self):
+        # GH 6914
+
+        df_original = DataFrame(
+            {
+                "Buyer": "Carl Joe Joe Carl Joe Carl".split(),
+                "Quantity": [18, 3, 5, 1, 9, 3],
+                "Date": [
+                    datetime(2013, 9, 1, 13, 0),
+                    datetime(2013, 9, 1, 13, 5),
+                    datetime(2013, 10, 1, 20, 0),
+                    datetime(2013, 10, 3, 10, 0),
+                    datetime(2013, 12, 2, 12, 0),
+                    datetime(2013, 9, 2, 14, 0),
+                ],
+            }
+        )
+        df_reordered = df_original.sort_values(by="Quantity")
+
+        # single grouping
+        expected_list = [
+            df_original.iloc[[0, 1, 5]],
+            df_original.iloc[[2, 3]],
+            df_original.iloc[[4]],
+        ]
+        dt_list = ["2013-09-30", "2013-10-31", "2013-12-31"]
+
+        for df in [df_original, df_reordered]:
+            grouped = df.groupby(Grouper(freq="ME", key="Date"))
+            for t, expected in zip(dt_list, expected_list):
+                dt = Timestamp(t)
+                result = grouped.get_group(dt)
+                tm.assert_frame_equal(result, expected)
+
+        # multiple grouping
+        expected_list = [
+            df_original.iloc[[1]],
+            df_original.iloc[[3]],
+            df_original.iloc[[4]],
+        ]
+        g_list = [("Joe", "2013-09-30"), ("Carl", "2013-10-31"), ("Joe", "2013-12-31")]
+
+        for df in [df_original, df_reordered]:
+            grouped = df.groupby(["Buyer", Grouper(freq="ME", key="Date")])
+            for (b, t), expected in zip(g_list, expected_list):
+                dt = Timestamp(t)
+                result = grouped.get_group((b, dt))
+                tm.assert_frame_equal(result, expected)
+
+        # with index
+        df_original = df_original.set_index("Date")
+        df_reordered = df_original.sort_values(by="Quantity")
+
+        expected_list = [
+            df_original.iloc[[0, 1, 5]],
+            df_original.iloc[[2, 3]],
+            df_original.iloc[[4]],
+        ]
+
+        for df in [df_original, df_reordered]:
+            grouped = df.groupby(Grouper(freq="ME"))
+            for t, expected in zip(dt_list, expected_list):
+                dt = Timestamp(t)
+                result = grouped.get_group(dt)
+                tm.assert_frame_equal(result, expected)
+
+    def test_timegrouper_apply_return_type_series(self):
+        # Using `apply` with the `TimeGrouper` should give the
+        # same return type as an `apply` with a `Grouper`.
+        # Issue #11742
+        df = DataFrame({"date": ["10/10/2000", "11/10/2000"], "value": [10, 13]})
+        df_dt = df.copy()
+        df_dt["date"] = pd.to_datetime(df_dt["date"])
+
+        def sumfunc_series(x):
+            return Series([x["value"].sum()], ("sum",))
+
+        msg = "DataFrameGroupBy.apply operated on the grouping columns"
+        with tm.assert_produces_warning(DeprecationWarning, match=msg):
+            expected = df.groupby(Grouper(key="date")).apply(sumfunc_series)
+        msg = "DataFrameGroupBy.apply operated on the grouping columns"
+        with tm.assert_produces_warning(DeprecationWarning, match=msg):
+            result = df_dt.groupby(Grouper(freq="ME", key="date")).apply(sumfunc_series)
+        tm.assert_frame_equal(
+            result.reset_index(drop=True), expected.reset_index(drop=True)
+        )
+
+    def test_timegrouper_apply_return_type_value(self):
+        # Using `apply` with the `TimeGrouper` should give the
+        # same return type as an `apply` with a `Grouper`.
+        # Issue #11742
+        df = DataFrame({"date": ["10/10/2000", "11/10/2000"], "value": [10, 13]})
+        df_dt = df.copy()
+        df_dt["date"] = pd.to_datetime(df_dt["date"])
+
+        def sumfunc_value(x):
+            return x.value.sum()
+
+        msg = "DataFrameGroupBy.apply operated on the grouping columns"
+        with tm.assert_produces_warning(DeprecationWarning, match=msg):
+            expected = df.groupby(Grouper(key="date")).apply(sumfunc_value)
+        with tm.assert_produces_warning(DeprecationWarning, match=msg):
+            result = df_dt.groupby(Grouper(freq="ME", key="date")).apply(sumfunc_value)
+        tm.assert_series_equal(
+            result.reset_index(drop=True), expected.reset_index(drop=True)
+        )
+
+    def test_groupby_groups_datetimeindex(self):
+        # GH#1430
+        periods = 1000
+        ind = date_range(start="2012/1/1", freq="5min", periods=periods)
+        df = DataFrame(
+            {"high": np.arange(periods), "low": np.arange(periods)}, index=ind
+        )
+        grouped = df.groupby(lambda x: datetime(x.year, x.month, x.day))
+
+        # it works!
+        groups = grouped.groups
+        assert isinstance(next(iter(groups.keys())), datetime)
+
+    def test_groupby_groups_datetimeindex2(self):
+        # GH#11442
+        index = date_range("2015/01/01", periods=5, name="date")
+        df = DataFrame({"A": [5, 6, 7, 8, 9], "B": [1, 2, 3, 4, 5]}, index=index)
+        result = df.groupby(level="date").groups
+        dates = ["2015-01-05", "2015-01-04", "2015-01-03", "2015-01-02", "2015-01-01"]
+        expected = {
+            Timestamp(date): DatetimeIndex([date], name="date") for date in dates
+        }
+        tm.assert_dict_equal(result, expected)
+
+        grouped = df.groupby(level="date")
+        for date in dates:
+            result = grouped.get_group(date)
+            data = [[df.loc[date, "A"], df.loc[date, "B"]]]
+            expected_index = DatetimeIndex(
+                [date], name="date", freq="D", dtype=index.dtype
+            )
+            expected = DataFrame(data, columns=list("AB"), index=expected_index)
+            tm.assert_frame_equal(result, expected)
+
+    def test_groupby_groups_datetimeindex_tz(self):
+        # GH 3950
+        dates = [
+            "2011-07-19 07:00:00",
+            "2011-07-19 08:00:00",
+            "2011-07-19 09:00:00",
+            "2011-07-19 07:00:00",
+            "2011-07-19 08:00:00",
+            "2011-07-19 09:00:00",
+        ]
+        df = DataFrame(
+            {
+                "label": ["a", "a", "a", "b", "b", "b"],
+                "datetime": dates,
+                "value1": np.arange(6, dtype="int64"),
+                "value2": [1, 2] * 3,
+            }
+        )
+        df["datetime"] = df["datetime"].apply(lambda d: Timestamp(d, tz="US/Pacific"))
+
+        exp_idx1 = DatetimeIndex(
+            [
+                "2011-07-19 07:00:00",
+                "2011-07-19 07:00:00",
+                "2011-07-19 08:00:00",
+                "2011-07-19 08:00:00",
+                "2011-07-19 09:00:00",
+                "2011-07-19 09:00:00",
+            ],
+            tz="US/Pacific",
+            name="datetime",
+        )
+        exp_idx2 = Index(["a", "b"] * 3, name="label")
+        exp_idx = MultiIndex.from_arrays([exp_idx1, exp_idx2])
+        expected = DataFrame(
+            {"value1": [0, 3, 1, 4, 2, 5], "value2": [1, 2, 2, 1, 1, 2]},
+            index=exp_idx,
+            columns=["value1", "value2"],
+        )
+
+        result = df.groupby(["datetime", "label"]).sum()
+        tm.assert_frame_equal(result, expected)
+
+        # by level
+        didx = DatetimeIndex(dates, tz="Asia/Tokyo")
+        df = DataFrame(
+            {"value1": np.arange(6, dtype="int64"), "value2": [1, 2, 3, 1, 2, 3]},
+            index=didx,
+        )
+
+        exp_idx = DatetimeIndex(
+            ["2011-07-19 07:00:00", "2011-07-19 08:00:00", "2011-07-19 09:00:00"],
+            tz="Asia/Tokyo",
+        )
+        expected = DataFrame(
+            {"value1": [3, 5, 7], "value2": [2, 4, 6]},
+            index=exp_idx,
+            columns=["value1", "value2"],
+        )
+
+        result = df.groupby(level=0).sum()
+        tm.assert_frame_equal(result, expected)
+
+    def test_frame_datetime64_handling_groupby(self):
+        # it works!
+        df = DataFrame(
+            [(3, np.datetime64("2012-07-03")), (3, np.datetime64("2012-07-04"))],
+            columns=["a", "date"],
+        )
+        result = df.groupby("a").first()
+        assert result["date"][3] == Timestamp("2012-07-03")
+
+    def test_groupby_multi_timezone(self):
+        # combining multiple / different timezones yields UTC
+        df = DataFrame(
+            {
+                "value": range(5),
+                "date": [
+                    "2000-01-28 16:47:00",
+                    "2000-01-29 16:48:00",
+                    "2000-01-30 16:49:00",
+                    "2000-01-31 16:50:00",
+                    "2000-01-01 16:50:00",
+                ],
+                "tz": [
+                    "America/Chicago",
+                    "America/Chicago",
+                    "America/Los_Angeles",
+                    "America/Chicago",
+                    "America/New_York",
+                ],
+            }
+        )
+
+        result = df.groupby("tz", group_keys=False).date.apply(
+            lambda x: pd.to_datetime(x).dt.tz_localize(x.name)
+        )
+
+        expected = Series(
+            [
+                Timestamp("2000-01-28 16:47:00-0600", tz="America/Chicago"),
+                Timestamp("2000-01-29 16:48:00-0600", tz="America/Chicago"),
+                Timestamp("2000-01-30 16:49:00-0800", tz="America/Los_Angeles"),
+                Timestamp("2000-01-31 16:50:00-0600", tz="America/Chicago"),
+                Timestamp("2000-01-01 16:50:00-0500", tz="America/New_York"),
+            ],
+            name="date",
+            dtype=object,
+        )
+        tm.assert_series_equal(result, expected)
+
+        tz = "America/Chicago"
+        res_values = df.groupby("tz").date.get_group(tz)
+        result = pd.to_datetime(res_values).dt.tz_localize(tz)
+        exp_values = Series(
+            ["2000-01-28 16:47:00", "2000-01-29 16:48:00", "2000-01-31 16:50:00"],
+            index=[0, 1, 3],
+            name="date",
+        )
+        expected = pd.to_datetime(exp_values).dt.tz_localize(tz)
+        tm.assert_series_equal(result, expected)
+
+    def test_groupby_groups_periods(self):
+        dates = [
+            "2011-07-19 07:00:00",
+            "2011-07-19 08:00:00",
+            "2011-07-19 09:00:00",
+            "2011-07-19 07:00:00",
+            "2011-07-19 08:00:00",
+            "2011-07-19 09:00:00",
+        ]
+        df = DataFrame(
+            {
+                "label": ["a", "a", "a", "b", "b", "b"],
+                "period": [pd.Period(d, freq="h") for d in dates],
+                "value1": np.arange(6, dtype="int64"),
+                "value2": [1, 2] * 3,
+            }
+        )
+
+        exp_idx1 = pd.PeriodIndex(
+            [
+                "2011-07-19 07:00:00",
+                "2011-07-19 07:00:00",
+                "2011-07-19 08:00:00",
+                "2011-07-19 08:00:00",
+                "2011-07-19 09:00:00",
+                "2011-07-19 09:00:00",
+            ],
+            freq="h",
+            name="period",
+        )
+        exp_idx2 = Index(["a", "b"] * 3, name="label")
+        exp_idx = MultiIndex.from_arrays([exp_idx1, exp_idx2])
+        expected = DataFrame(
+            {"value1": [0, 3, 1, 4, 2, 5], "value2": [1, 2, 2, 1, 1, 2]},
+            index=exp_idx,
+            columns=["value1", "value2"],
+        )
+
+        result = df.groupby(["period", "label"]).sum()
+        tm.assert_frame_equal(result, expected)
+
+        # by level
+        didx = pd.PeriodIndex(dates, freq="h")
+        df = DataFrame(
+            {"value1": np.arange(6, dtype="int64"), "value2": [1, 2, 3, 1, 2, 3]},
+            index=didx,
+        )
+
+        exp_idx = pd.PeriodIndex(
+            ["2011-07-19 07:00:00", "2011-07-19 08:00:00", "2011-07-19 09:00:00"],
+            freq="h",
+        )
+        expected = DataFrame(
+            {"value1": [3, 5, 7], "value2": [2, 4, 6]},
+            index=exp_idx,
+            columns=["value1", "value2"],
+        )
+
+        result = df.groupby(level=0).sum()
+        tm.assert_frame_equal(result, expected)
+
+    def test_groupby_first_datetime64(self):
+        df = DataFrame([(1, 1351036800000000000), (2, 1351036800000000000)])
+        df[1] = df[1].astype("M8[ns]")
+
+        assert issubclass(df[1].dtype.type, np.datetime64)
+
+        result = df.groupby(level=0).first()
+        got_dt = result[1].dtype
+        assert issubclass(got_dt.type, np.datetime64)
+
+        result = df[1].groupby(level=0).first()
+        got_dt = result.dtype
+        assert issubclass(got_dt.type, np.datetime64)
+
+    def test_groupby_max_datetime64(self):
+        # GH 5869
+        # datetimelike dtype conversion from int
+        df = DataFrame({"A": Timestamp("20130101"), "B": np.arange(5)})
+        # TODO: can we retain second reso in .apply here?
+        expected = df.groupby("A")["A"].apply(lambda x: x.max()).astype("M8[s]")
+        result = df.groupby("A")["A"].max()
+        tm.assert_series_equal(result, expected)
+
+    def test_groupby_datetime64_32_bit(self):
+        # GH 6410 / numpy 4328
+        # 32-bit under 1.9-dev indexing issue
+
+        df = DataFrame({"A": range(2), "B": [Timestamp("2000-01-1")] * 2})
+        result = df.groupby("A")["B"].transform("min")
+        expected = Series([Timestamp("2000-01-1")] * 2, name="B")
+        tm.assert_series_equal(result, expected)
+
+    def test_groupby_with_timezone_selection(self):
+        # GH 11616
+        # Test that column selection returns output in correct timezone.
+
+        df = DataFrame(
+            {
+                "factor": np.random.default_rng(2).integers(0, 3, size=60),
+                "time": date_range("01/01/2000 00:00", periods=60, freq="s", tz="UTC"),
+            }
+        )
+        df1 = df.groupby("factor").max()["time"]
+        df2 = df.groupby("factor")["time"].max()
+        tm.assert_series_equal(df1, df2)
+
+    def test_timezone_info(self):
+        # see gh-11682: Timezone info lost when broadcasting
+        # scalar datetime to DataFrame
+
+        df = DataFrame({"a": [1], "b": [datetime.now(pytz.utc)]})
+        assert df["b"][0].tzinfo == pytz.utc
+        df = DataFrame({"a": [1, 2, 3]})
+        df["b"] = datetime.now(pytz.utc)
+        assert df["b"][0].tzinfo == pytz.utc
+
+    def test_datetime_count(self):
+        df = DataFrame(
+            {"a": [1, 2, 3] * 2, "dates": date_range("now", periods=6, freq="min")}
+        )
+        result = df.groupby("a").dates.count()
+        expected = Series([2, 2, 2], index=Index([1, 2, 3], name="a"), name="dates")
+        tm.assert_series_equal(result, expected)
+
+    def test_first_last_max_min_on_time_data(self):
+        # GH 10295
+        # Verify that NaT is not in the result of max, min, first and last on
+        # Dataframe with datetime or timedelta values.
+        df_test = DataFrame(
+            {
+                "dt": [
+                    np.nan,
+                    "2015-07-24 10:10",
+                    "2015-07-25 11:11",
+                    "2015-07-23 12:12",
+                    np.nan,
+                ],
+                "td": [
+                    np.nan,
+                    timedelta(days=1),
+                    timedelta(days=2),
+                    timedelta(days=3),
+                    np.nan,
+                ],
+            }
+        )
+        df_test.dt = pd.to_datetime(df_test.dt)
+        df_test["group"] = "A"
+        df_ref = df_test[df_test.dt.notna()]
+
+        grouped_test = df_test.groupby("group")
+        grouped_ref = df_ref.groupby("group")
+
+        tm.assert_frame_equal(grouped_ref.max(), grouped_test.max())
+        tm.assert_frame_equal(grouped_ref.min(), grouped_test.min())
+        tm.assert_frame_equal(grouped_ref.first(), grouped_test.first())
+        tm.assert_frame_equal(grouped_ref.last(), grouped_test.last())
+
+    def test_nunique_with_timegrouper_and_nat(self):
+        # GH 17575
+        test = DataFrame(
+            {
+                "time": [
+                    Timestamp("2016-06-28 09:35:35"),
+                    pd.NaT,
+                    Timestamp("2016-06-28 16:46:28"),
+                ],
+                "data": ["1", "2", "3"],
+            }
+        )
+
+        grouper = Grouper(key="time", freq="h")
+        result = test.groupby(grouper)["data"].nunique()
+        expected = test[test.time.notnull()].groupby(grouper)["data"].nunique()
+        expected.index = expected.index._with_freq(None)
+        tm.assert_series_equal(result, expected)
+
+    def test_scalar_call_versus_list_call(self):
+        # Issue: 17530
+        data_frame = {
+            "location": ["shanghai", "beijing", "shanghai"],
+            "time": Series(
+                ["2017-08-09 13:32:23", "2017-08-11 23:23:15", "2017-08-11 22:23:15"],
+                dtype="datetime64[ns]",
+            ),
+            "value": [1, 2, 3],
+        }
+        data_frame = DataFrame(data_frame).set_index("time")
+        grouper = Grouper(freq="D")
+
+        grouped = data_frame.groupby(grouper)
+        result = grouped.count()
+        grouped = data_frame.groupby([grouper])
+        expected = grouped.count()
+
+        tm.assert_frame_equal(result, expected)
+
+    def test_grouper_period_index(self):
+        # GH 32108
+        periods = 2
+        index = pd.period_range(
+            start="2018-01", periods=periods, freq="M", name="Month"
+        )
+        period_series = Series(range(periods), index=index)
+        result = period_series.groupby(period_series.index.month).sum()
+
+        expected = Series(
+            range(periods), index=Index(range(1, periods + 1), name=index.name)
+        )
+        tm.assert_series_equal(result, expected)
+
+    def test_groupby_apply_timegrouper_with_nat_dict_returns(
+        self, groupby_with_truncated_bingrouper
+    ):
+        # GH#43500 case where gb._grouper.result_index and gb._grouper.group_keys_seq
+        #  have different lengths that goes through the `isinstance(values[0], dict)`
+        #  path
+        gb = groupby_with_truncated_bingrouper
+
+        res = gb["Quantity"].apply(lambda x: {"foo": len(x)})
+
+        df = gb.obj
+        unit = df["Date"]._values.unit
+        dti = date_range("2013-09-01", "2013-10-01", freq="5D", name="Date", unit=unit)
+        mi = MultiIndex.from_arrays([dti, ["foo"] * len(dti)])
+        expected = Series([3, 0, 0, 0, 0, 0, 2], index=mi, name="Quantity")
+        tm.assert_series_equal(res, expected)
+
+    def test_groupby_apply_timegrouper_with_nat_scalar_returns(
+        self, groupby_with_truncated_bingrouper
+    ):
+        # GH#43500 Previously raised ValueError bc used index with incorrect
+        #  length in wrap_applied_result
+        gb = groupby_with_truncated_bingrouper
+
+        res = gb["Quantity"].apply(lambda x: x.iloc[0] if len(x) else np.nan)
+
+        df = gb.obj
+        unit = df["Date"]._values.unit
+        dti = date_range("2013-09-01", "2013-10-01", freq="5D", name="Date", unit=unit)
+        expected = Series(
+            [18, np.nan, np.nan, np.nan, np.nan, np.nan, 5],
+            index=dti._with_freq(None),
+            name="Quantity",
+        )
+
+        tm.assert_series_equal(res, expected)
+
+    def test_groupby_apply_timegrouper_with_nat_apply_squeeze(
+        self, frame_for_truncated_bingrouper
+    ):
+        df = frame_for_truncated_bingrouper
+
+        # We need to create a GroupBy object with only one non-NaT group,
+        #  so use a huge freq so that all non-NaT dates will be grouped together
+        tdg = Grouper(key="Date", freq="100YE")
+        gb = df.groupby(tdg)
+
+        # check that we will go through the singular_series path
+        #  in _wrap_applied_output_series
+        assert gb.ngroups == 1
+        assert gb._selected_obj._get_axis(gb.axis).nlevels == 1
+
+        # function that returns a Series
+        msg = "DataFrameGroupBy.apply operated on the grouping columns"
+        with tm.assert_produces_warning(DeprecationWarning, match=msg):
+            res = gb.apply(lambda x: x["Quantity"] * 2)
+
+        dti = Index([Timestamp("2013-12-31")], dtype=df["Date"].dtype, name="Date")
+        expected = DataFrame(
+            [[36, 6, 6, 10, 2]],
+            index=dti,
+            columns=Index([0, 1, 5, 2, 3], name="Quantity"),
+        )
+        tm.assert_frame_equal(res, expected)
+
+    @pytest.mark.single_cpu
+    def test_groupby_agg_numba_timegrouper_with_nat(
+        self, groupby_with_truncated_bingrouper
+    ):
+        pytest.importorskip("numba")
+
+        # See discussion in GH#43487
+        gb = groupby_with_truncated_bingrouper
+
+        result = gb["Quantity"].aggregate(
+            lambda values, index: np.nanmean(values), engine="numba"
+        )
+
+        expected = gb["Quantity"].aggregate("mean")
+        tm.assert_series_equal(result, expected)
+
+        result_df = gb[["Quantity"]].aggregate(
+            lambda values, index: np.nanmean(values), engine="numba"
+        )
+        expected_df = gb[["Quantity"]].aggregate("mean")
+        tm.assert_frame_equal(result_df, expected_df)

venv/lib/python3.12/site-packages/pandas/tests/groupby/transform/test_numba.py

@@ -0,0 +1,284 @@
+import numpy as np
+import pytest
+
+from pandas.errors import NumbaUtilError
+
+from pandas import (
+    DataFrame,
+    Series,
+    option_context,
+)
+import pandas._testing as tm
+
+pytestmark = pytest.mark.single_cpu
+
+
+def test_correct_function_signature():
+    pytest.importorskip("numba")
+
+    def incorrect_function(x):
+        return x + 1
+
+    data = DataFrame(
+        {"key": ["a", "a", "b", "b", "a"], "data": [1.0, 2.0, 3.0, 4.0, 5.0]},
+        columns=["key", "data"],
+    )
+    with pytest.raises(NumbaUtilError, match="The first 2"):
+        data.groupby("key").transform(incorrect_function, engine="numba")
+
+    with pytest.raises(NumbaUtilError, match="The first 2"):
+        data.groupby("key")["data"].transform(incorrect_function, engine="numba")
+
+
+def test_check_nopython_kwargs():
+    pytest.importorskip("numba")
+
+    def incorrect_function(values, index):
+        return values + 1
+
+    data = DataFrame(
+        {"key": ["a", "a", "b", "b", "a"], "data": [1.0, 2.0, 3.0, 4.0, 5.0]},
+        columns=["key", "data"],
+    )
+    with pytest.raises(NumbaUtilError, match="numba does not support"):
+        data.groupby("key").transform(incorrect_function, engine="numba", a=1)
+
+    with pytest.raises(NumbaUtilError, match="numba does not support"):
+        data.groupby("key")["data"].transform(incorrect_function, engine="numba", a=1)
+
+
+@pytest.mark.filterwarnings("ignore")
+# Filter warnings when parallel=True and the function can't be parallelized by Numba
+@pytest.mark.parametrize("jit", [True, False])
+@pytest.mark.parametrize("pandas_obj", ["Series", "DataFrame"])
+@pytest.mark.parametrize("as_index", [True, False])
+def test_numba_vs_cython(jit, pandas_obj, nogil, parallel, nopython, as_index):
+    pytest.importorskip("numba")
+
+    def func(values, index):
+        return values + 1
+
+    if jit:
+        # Test accepted jitted functions
+        import numba
+
+        func = numba.jit(func)
+
+    data = DataFrame(
+        {0: ["a", "a", "b", "b", "a"], 1: [1.0, 2.0, 3.0, 4.0, 5.0]}, columns=[0, 1]
+    )
+    engine_kwargs = {"nogil": nogil, "parallel": parallel, "nopython": nopython}
+    grouped = data.groupby(0, as_index=as_index)
+    if pandas_obj == "Series":
+        grouped = grouped[1]
+
+    result = grouped.transform(func, engine="numba", engine_kwargs=engine_kwargs)
+    expected = grouped.transform(lambda x: x + 1, engine="cython")
+
+    tm.assert_equal(result, expected)
+
+
+@pytest.mark.filterwarnings("ignore")
+# Filter warnings when parallel=True and the function can't be parallelized by Numba
+@pytest.mark.parametrize("jit", [True, False])
+@pytest.mark.parametrize("pandas_obj", ["Series", "DataFrame"])
+def test_cache(jit, pandas_obj, nogil, parallel, nopython):
+    # Test that the functions are cached correctly if we switch functions
+    pytest.importorskip("numba")
+
+    def func_1(values, index):
+        return values + 1
+
+    def func_2(values, index):
+        return values * 5
+
+    if jit:
+        import numba
+
+        func_1 = numba.jit(func_1)
+        func_2 = numba.jit(func_2)
+
+    data = DataFrame(
+        {0: ["a", "a", "b", "b", "a"], 1: [1.0, 2.0, 3.0, 4.0, 5.0]}, columns=[0, 1]
+    )
+    engine_kwargs = {"nogil": nogil, "parallel": parallel, "nopython": nopython}
+    grouped = data.groupby(0)
+    if pandas_obj == "Series":
+        grouped = grouped[1]
+
+    result = grouped.transform(func_1, engine="numba", engine_kwargs=engine_kwargs)
+    expected = grouped.transform(lambda x: x + 1, engine="cython")
+    tm.assert_equal(result, expected)
+
+    result = grouped.transform(func_2, engine="numba", engine_kwargs=engine_kwargs)
+    expected = grouped.transform(lambda x: x * 5, engine="cython")
+    tm.assert_equal(result, expected)
+
+    # Retest func_1 which should use the cache
+    result = grouped.transform(func_1, engine="numba", engine_kwargs=engine_kwargs)
+    expected = grouped.transform(lambda x: x + 1, engine="cython")
+    tm.assert_equal(result, expected)
+
+
+def test_use_global_config():
+    pytest.importorskip("numba")
+
+    def func_1(values, index):
+        return values + 1
+
+    data = DataFrame(
+        {0: ["a", "a", "b", "b", "a"], 1: [1.0, 2.0, 3.0, 4.0, 5.0]}, columns=[0, 1]
+    )
+    grouped = data.groupby(0)
+    expected = grouped.transform(func_1, engine="numba")
+    with option_context("compute.use_numba", True):
+        result = grouped.transform(func_1, engine=None)
+    tm.assert_frame_equal(expected, result)
+
+
+# TODO: Test more than just reductions (e.g. actually test transformations once we have
+@pytest.mark.parametrize(
+    "agg_func", [["min", "max"], "min", {"B": ["min", "max"], "C": "sum"}]
+)
+def test_string_cython_vs_numba(agg_func, numba_supported_reductions):
+    pytest.importorskip("numba")
+    agg_func, kwargs = numba_supported_reductions
+    data = DataFrame(
+        {0: ["a", "a", "b", "b", "a"], 1: [1.0, 2.0, 3.0, 4.0, 5.0]}, columns=[0, 1]
+    )
+    grouped = data.groupby(0)
+
+    result = grouped.transform(agg_func, engine="numba", **kwargs)
+    expected = grouped.transform(agg_func, engine="cython", **kwargs)
+    tm.assert_frame_equal(result, expected)
+
+    result = grouped[1].transform(agg_func, engine="numba", **kwargs)
+    expected = grouped[1].transform(agg_func, engine="cython", **kwargs)
+    tm.assert_series_equal(result, expected)
+
+
+def test_args_not_cached():
+    # GH 41647
+    pytest.importorskip("numba")
+
+    def sum_last(values, index, n):
+        return values[-n:].sum()
+
+    df = DataFrame({"id": [0, 0, 1, 1], "x": [1, 1, 1, 1]})
+    grouped_x = df.groupby("id")["x"]
+    result = grouped_x.transform(sum_last, 1, engine="numba")
+    expected = Series([1.0] * 4, name="x")
+    tm.assert_series_equal(result, expected)
+
+    result = grouped_x.transform(sum_last, 2, engine="numba")
+    expected = Series([2.0] * 4, name="x")
+    tm.assert_series_equal(result, expected)
+
+
+def test_index_data_correctly_passed():
+    # GH 43133
+    pytest.importorskip("numba")
+
+    def f(values, index):
+        return index - 1
+
+    df = DataFrame({"group": ["A", "A", "B"], "v": [4, 5, 6]}, index=[-1, -2, -3])
+    result = df.groupby("group").transform(f, engine="numba")
+    expected = DataFrame([-4.0, -3.0, -2.0], columns=["v"], index=[-1, -2, -3])
+    tm.assert_frame_equal(result, expected)
+
+
+def test_engine_kwargs_not_cached():
+    # If the user passes a different set of engine_kwargs don't return the same
+    # jitted function
+    pytest.importorskip("numba")
+    nogil = True
+    parallel = False
+    nopython = True
+
+    def func_kwargs(values, index):
+        return nogil + parallel + nopython
+
+    engine_kwargs = {"nopython": nopython, "nogil": nogil, "parallel": parallel}
+    df = DataFrame({"value": [0, 0, 0]})
+    result = df.groupby(level=0).transform(
+        func_kwargs, engine="numba", engine_kwargs=engine_kwargs
+    )
+    expected = DataFrame({"value": [2.0, 2.0, 2.0]})
+    tm.assert_frame_equal(result, expected)
+
+    nogil = False
+    engine_kwargs = {"nopython": nopython, "nogil": nogil, "parallel": parallel}
+    result = df.groupby(level=0).transform(
+        func_kwargs, engine="numba", engine_kwargs=engine_kwargs
+    )
+    expected = DataFrame({"value": [1.0, 1.0, 1.0]})
+    tm.assert_frame_equal(result, expected)
+
+
+@pytest.mark.filterwarnings("ignore")
+def test_multiindex_one_key(nogil, parallel, nopython):
+    pytest.importorskip("numba")
+
+    def numba_func(values, index):
+        return 1
+
+    df = DataFrame([{"A": 1, "B": 2, "C": 3}]).set_index(["A", "B"])
+    engine_kwargs = {"nopython": nopython, "nogil": nogil, "parallel": parallel}
+    result = df.groupby("A").transform(
+        numba_func, engine="numba", engine_kwargs=engine_kwargs
+    )
+    expected = DataFrame([{"A": 1, "B": 2, "C": 1.0}]).set_index(["A", "B"])
+    tm.assert_frame_equal(result, expected)
+
+
+def test_multiindex_multi_key_not_supported(nogil, parallel, nopython):
+    pytest.importorskip("numba")
+
+    def numba_func(values, index):
+        return 1
+
+    df = DataFrame([{"A": 1, "B": 2, "C": 3}]).set_index(["A", "B"])
+    engine_kwargs = {"nopython": nopython, "nogil": nogil, "parallel": parallel}
+    with pytest.raises(NotImplementedError, match="more than 1 grouping labels"):
+        df.groupby(["A", "B"]).transform(
+            numba_func, engine="numba", engine_kwargs=engine_kwargs
+        )
+
+
+def test_multilabel_numba_vs_cython(numba_supported_reductions):
+    pytest.importorskip("numba")
+    reduction, kwargs = numba_supported_reductions
+    df = DataFrame(
+        {
+            "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"],
+            "B": ["one", "one", "two", "three", "two", "two", "one", "three"],
+            "C": np.random.default_rng(2).standard_normal(8),
+            "D": np.random.default_rng(2).standard_normal(8),
+        }
+    )
+    gb = df.groupby(["A", "B"])
+    res_agg = gb.transform(reduction, engine="numba", **kwargs)
+    expected_agg = gb.transform(reduction, engine="cython", **kwargs)
+    tm.assert_frame_equal(res_agg, expected_agg)
+
+
+def test_multilabel_udf_numba_vs_cython():
+    pytest.importorskip("numba")
+    df = DataFrame(
+        {
+            "A": ["foo", "bar", "foo", "bar", "foo", "bar", "foo", "foo"],
+            "B": ["one", "one", "two", "three", "two", "two", "one", "three"],
+            "C": np.random.default_rng(2).standard_normal(8),
+            "D": np.random.default_rng(2).standard_normal(8),
+        }
+    )
+    gb = df.groupby(["A", "B"])
+    result = gb.transform(
+        lambda values, index: (values - values.min()) / (values.max() - values.min()),
+        engine="numba",
+    )
+    expected = gb.transform(
+        lambda x: (x - x.min()) / (x.max() - x.min()), engine="cython"
+    )
+    tm.assert_frame_equal(result, expected)