attempt to mimic previous count_values behavior by reversing before sort

pandas/core/algorithms.py

@@ -819,22 +819,19 @@ def value_counts_arraylike(values, dropna: bool):
         # TODO: handle uint8
         f = getattr(htable, f"value_count_{ndtype}")
         keys, counts = f(values, dropna)
+        # GH 35922. Mimic previous value_counts behavior now that
+        # Series.sort_values is stable
+        keys = keys[::-1]
+        counts = counts[::-1]
 
         mask = isna(values)
         if not dropna and mask.any():
             # GH 35922. Series.sort_values is stable now, so need to
-            # append NaN counts or move to the end to make sure they are
+            # append NaN counts  to make sure they are
             # sorted toward the end when calling value_counts
             if not isna(keys).any():
                 keys = np.append(keys, np.NaN)
                 counts = np.append(counts, mask.sum())
-            else:
-                nan_pos = np.where(np.isnan(keys))
-                keys[nan_pos] = keys[-1]
-                keys[-1] = np.NaN
-                tmp = counts[nan_pos]
-                counts[nan_pos] = counts[-1]
-                counts[-1] = tmp
 
     keys = _reconstruct_data(keys, original.dtype, original)