use cfg context

modin/tests/pandas/test_groupby.py

@@ -22,11 +22,11 @@
 
 import modin.pandas as pd
 from modin.config import (
-    DynamicPartitioning,
     IsRayCluster,
     NPartitions,
     RangePartitioning,
     StorageFormat,
+    context,
 )
 from modin.core.dataframe.algebra.default2pandas.groupby import GroupBy
 from modin.core.dataframe.pandas.partitioning.axis_partition import (
@@ -2438,8 +2438,6 @@ def test_multi_column_groupby_different_partitions(
 
 
 def test_empty_partitions_after_groupby():
-    DynamicPartitioning.put(True)
-
     def func_to_apply(grp):
         return grp.agg(
             {
@@ -2452,15 +2450,16 @@ def func_to_apply(grp):
     md_df, pd_df = create_test_dfs(data)
     by = pd_df.columns[0]
 
-    md_grp, pd_grp = (
-        md_df.groupby(by),
-        pd_df.groupby(by),
-    )
-    eval_general(
-        md_grp,
-        pd_grp,
-        func_to_apply,
-    )
+    with context(DynamicPartitioning=True):
+        md_grp, pd_grp = (
+            md_df.groupby(by),
+            pd_df.groupby(by),
+        )
+        eval_general(
+            md_grp,
+            pd_grp,
+            func_to_apply,
+        )
 
 
 @pytest.mark.parametrize(