Deprecate Groupby.collect

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

@@ -2180,7 +2180,13 @@ def func(x):
     @_cudf_nvtx_annotate
     def collect(self):
         """Get a list of all the values for each column in each group."""
-        return self.agg("collect")
+        warnings.warn(
+            "Groupby.collect is deprecated and "
+            "will be removed in a future version. "
+            "Use `.agg(list)` instead.",
+            FutureWarning,
+        )
+        return self.agg(list)
 
     @_cudf_nvtx_annotate
     def unique(self):

python/dask_cudf/dask_cudf/groupby.py

@@ -166,7 +166,7 @@ def max(self, split_every=None, split_out=1):
     def collect(self, split_every=None, split_out=1):
         return _make_groupby_agg_call(
             self,
-            self._make_groupby_method_aggs("collect"),
+            self._make_groupby_method_aggs(list),
             split_every,
             split_out,
         )
@@ -310,7 +310,7 @@ def max(self, split_every=None, split_out=1):
     def collect(self, split_every=None, split_out=1):
         return _make_groupby_agg_call(
             self,
-            {self._slice: "collect"},
+            {self._slice: list},
             split_every,
             split_out,
         )[self._slice]

python/dask_cudf/dask_cudf/tests/test_groupby.py

@@ -9,6 +9,7 @@
 from dask.utils_test import hlg_layer
 
 import cudf
+from cudf.testing._utils import expect_warning_if
 
 import dask_cudf
 from dask_cudf.groupby import OPTIMIZED_AGGS, _aggs_optimized
@@ -62,7 +63,8 @@ def test_groupby_basic(series, aggregation, pdf):
 
     check_dtype = aggregation != "count"
 
-    expect = getattr(gdf_grouped, aggregation)()
+    with expect_warning_if(aggregation == "collect"):
+        expect = getattr(gdf_grouped, aggregation)()
     actual = getattr(ddf_grouped, aggregation)()
 
     if not QUERY_PLANNING_ON: