modin can use fastpath too

narwhals/pandas_like/utils.py

@@ -234,7 +234,11 @@ def evaluate_simple_aggregation(expr: PandasExpr, grouped: Any, keys: list[str])
     if expr._depth == 0:
         # e.g. agg(pl.len())
         df = getattr(grouped, expr._function_name.replace("len", "size"))()
-        df = df.drop(columns=keys) if len(df.shape) > 1 else df.to_frame("size")
+        df = (
+            df.drop(columns=keys)
+            if len(df.shape) > 1
+            else df.reset_index(drop=True).to_frame("size")
+        )
         return df.rename(columns={"size": expr._output_names[0]})  # type: ignore[index]
     if expr._root_names is None or expr._output_names is None:
         msg = "Expected expr to have root_names and output_names set, but they are None. Please report a bug."