Implement Groupby pct_change

python/cudf/cudf/core/groupby/groupby.py

@@ -1416,6 +1416,48 @@ def shift(self, periods=1, freq=None, axis=0, fill_value=None):
         result = self._mimic_pandas_order(result)
         return result._copy_type_metadata(values)
 
+    def pct_change(
+        self, periods=1, fill_method="ffill", axis=0, limit=None, freq=None
+    ):
+        """
+        Calculates the percent change between sequential elements
+        in the group.
+
+        Parameters
+        ----------
+        periods : int, default 1
+            Periods to shift for forming percent change.
+        fill_method : str, default 'ffill'
+            How to handle NAs before computing percent changes.
+        limit : int, optional
+            The number of consecutive NAs to fill before stopping.
+            Not yet implemented.
+        freq : str, optional
+            Increment to use from time series API.
+            Not yet implemented.
+
+        Returns
+        -------
+        Series or DataFrame
+            Percentage changes within each group
+        """
+        if not axis == 0:
+            raise NotImplementedError("Only axis=0 is supported.")
+        if limit is not None:
+            raise NotImplementedError("limit parameter not supported yet.")
+        if freq is not None:
+            raise NotImplementedError("freq parameter not supported yet.")
+        elif fill_method not in {"ffill", "pad", "bfill", "backfill"}:
+            raise ValueError(
+                "fill_method must be one of 'ffill', 'pad', "
+                "'bfill', or 'backfill'."
+            )
+
+        filled = self.fillna(method=fill_method, limit=limit)
+        fill_grp = filled.groupby(self.grouping)
+        shifted = fill_grp.shift(periods=periods, freq=freq)
+        return (filled / shifted) - 1
+
     def _mimic_pandas_order(
         self, result: DataFrameOrSeries
     ) -> DataFrameOrSeries:

python/cudf/cudf/tests/test_groupby.py

@@ -2604,3 +2604,76 @@ def test_groupby_transform_maintain_index(by):
     assert_groupby_results_equal(
         pdf.groupby(by).transform("max"), gdf.groupby(by).transform("max")
     )
+
+
+@pytest.mark.parametrize(
+    "data, gkey",
+    [
+        (
+            {
+                "id": ["a", "a", "a", "b", "b", "b", "c", "c", "c"],
+                "val1": [5, 4, 6, 4, 8, 7, 4, 5, 2],
+                "val2": [4, 5, 6, 1, 2, 9, 8, 5, 1],
+                "val3": [4, 5, 6, 1, 2, 9, 8, 5, 1],
+            },
+            ["id"],
+        ),
+        (
+            {
+                "id": [0, 0, 0, 0, 1, 1, 1],
+                "a": [1, 3, 4, 2.0, -3.0, 9.0, 10.0],
+                "b": [10.0, 23, -4.0, 2, -3.0, None, 19.0],
+            },
+            ["id", "a"],
+        ),
+        (
+            {
+                "id": ["a", "a", "b", "b", "c", "c"],
+                "val1": [None, None, None, None, None, None],
+            },
+            ["id"],
+        ),
+    ],
+)
+@pytest.mark.parametrize("periods", [-5, -2, 0, 2, 5])
+@pytest.mark.parametrize("fill_method", ["ffill", "bfill", "pad", "backfill"])
+def test_groupby_pct_change(data, gkey, periods, fill_method):
+    gdf = cudf.DataFrame(data)
+    pdf = gdf.to_pandas()
+
+    actual = gdf.groupby(gkey).pct_change(
+        periods=periods, fill_method=fill_method
+    )
+    expected = pdf.groupby(gkey).pct_change(
+        periods=periods, fill_method=fill_method
+    )
+
+    assert_eq(expected, actual)
+
+
+@pytest.mark.xfail(reason="https://github.com/rapidsai/cudf/issues/11259")
+@pytest.mark.parametrize("periods", [-5, 5])
+def test_groupby_pct_change_multiindex_dataframe(periods):
+    gdf = cudf.DataFrame(
+        {
+            "a": [1, 1, 2, 2],
+            "b": [1, 1, 2, 3],
+            "c": [2, 3, 4, 5],
+            "d": [6, 8, 9, 1],
+        }
+    ).set_index(["a", "b"])
+
+    actual = gdf.groupby(level=["a", "b"]).pct_change(periods)
+    expected = gdf.to_pandas().groupby(level=["a", "b"]).pct_change(periods)
+
+    assert_eq(expected, actual)
+
+
+def test_groupby_pct_change_empty_columns():
+    gdf = cudf.DataFrame(columns=["id", "val1", "val2"])
+    pdf = gdf.to_pandas()
+
+    actual = gdf.groupby("id").pct_change()
+    expected = pdf.groupby("id").pct_change()
+
+    assert_eq(expected, actual)