TST: pytest.mark.single for test_unique_bad_unicode (#43526)

Author: jbrockmendel

pandas/tests/base/test_unique.py

@@ -115,12 +115,12 @@ def test_nunique_null(null_obj, index_or_series_obj):
         assert obj.nunique(dropna=False) == max(0, num_unique_values)
 
 
-@pytest.mark.parametrize(
-    "idx_or_series_w_bad_unicode", [pd.Index(["\ud83d"] * 2), pd.Series(["\ud83d"] * 2)]
-)
-def test_unique_bad_unicode(idx_or_series_w_bad_unicode):
+@pytest.mark.single
+def test_unique_bad_unicode(index_or_series):
     # regression test for #34550
-    obj = idx_or_series_w_bad_unicode
+    uval = "\ud83d"  # smiley emoji
+
+    obj = index_or_series([uval] * 2)
     result = obj.unique()
 
     if isinstance(obj, pd.Index):