ENH: Add numba engine to groupby.min/max (pandas-dev#45428)

Author: mroeschke

doc/source/whatsnew/v1.5.0.rst

@@ -35,6 +35,7 @@ Other enhancements
 - :class:`StringArray` now accepts array-likes containing nan-likes (``None``, ``np.nan``) for the ``values`` parameter in its constructor in addition to strings and :attr:`pandas.NA`. (:issue:`40839`)
 - Improved the rendering of ``categories`` in :class:`CategoricalIndex` (:issue:`45218`)
 - :meth:`to_numeric` now preserves float64 arrays when downcasting would generate values not representable in float32 (:issue:`43693`)
+- :meth:`.GroupBy.min` and :meth:`.GroupBy.max` now supports `Numba <https://numba.pydata.org/>`_ execution with the ``engine`` keyword (:issue:`45428`)
 -
 
 .. ---------------------------------------------------------------------------

pandas/core/groupby/groupby.py

@@ -2208,17 +2208,49 @@ def prod(
 
     @final
     @doc(_groupby_agg_method_template, fname="min", no=False, mc=-1)
-    def min(self, numeric_only: bool = False, min_count: int = -1):
-        return self._agg_general(
-            numeric_only=numeric_only, min_count=min_count, alias="min", npfunc=np.min
-        )
+    def min(
+        self,
+        numeric_only: bool = False,
+        min_count: int = -1,
+        engine: str | None = None,
+        engine_kwargs: dict[str, bool] | None = None,
+    ):
+        if maybe_use_numba(engine):
+            from pandas.core._numba.kernels import sliding_min_max
+
+            return self._numba_agg_general(
+                sliding_min_max, engine_kwargs, "groupby_min", False
+            )
+        else:
+            return self._agg_general(
+                numeric_only=numeric_only,
+                min_count=min_count,
+                alias="min",
+                npfunc=np.min,
+            )
 
     @final
     @doc(_groupby_agg_method_template, fname="max", no=False, mc=-1)
-    def max(self, numeric_only: bool = False, min_count: int = -1):
-        return self._agg_general(
-            numeric_only=numeric_only, min_count=min_count, alias="max", npfunc=np.max
-        )
+    def max(
+        self,
+        numeric_only: bool = False,
+        min_count: int = -1,
+        engine: str | None = None,
+        engine_kwargs: dict[str, bool] | None = None,
+    ):
+        if maybe_use_numba(engine):
+            from pandas.core._numba.kernels import sliding_min_max
+
+            return self._numba_agg_general(
+                sliding_min_max, engine_kwargs, "groupby_max", True
+            )
+        else:
+            return self._agg_general(
+                numeric_only=numeric_only,
+                min_count=min_count,
+                alias="max",
+                npfunc=np.max,
+            )
 
     @final
     @doc(_groupby_agg_method_template, fname="first", no=False, mc=-1)

pandas/tests/groupby/conftest.py

@@ -184,7 +184,10 @@ def nopython(request):
         ("std", {"ddof": 1}),
         ("std", {"ddof": 0}),
         ("sum", {}),
-    ]
+        ("min", {}),
+        ("max", {}),
+    ],
+    ids=["mean", "var_1", "var_0", "std_1", "std_0", "sum", "min", "max"],
 )
 def numba_supported_reductions(request):
     """reductions supported with engine='numba'"""

pandas/tests/groupby/test_numba.py

@@ -25,7 +25,7 @@ def test_cython_vs_numba_frame(
         )
         expected = getattr(gb, func)(**kwargs)
         # check_dtype can be removed if GH 44952 is addressed
-        check_dtype = func != "sum"
+        check_dtype = func not in ("sum", "min", "max")
         tm.assert_frame_equal(result, expected, check_dtype=check_dtype)
 
     def test_cython_vs_numba_getitem(
@@ -40,7 +40,7 @@ def test_cython_vs_numba_getitem(
         )
         expected = getattr(gb, func)(**kwargs)
         # check_dtype can be removed if GH 44952 is addressed
-        check_dtype = func != "sum"
+        check_dtype = func not in ("sum", "min", "max")
         tm.assert_series_equal(result, expected, check_dtype=check_dtype)
 
     def test_cython_vs_numba_series(
@@ -55,7 +55,7 @@ def test_cython_vs_numba_series(
         )
         expected = getattr(gb, func)(**kwargs)
         # check_dtype can be removed if GH 44952 is addressed
-        check_dtype = func != "sum"
+        check_dtype = func not in ("sum", "min", "max")
         tm.assert_series_equal(result, expected, check_dtype=check_dtype)
 
     def test_as_index_false_unsupported(self, numba_supported_reductions):