PERF: new str.split performance improvement, handles NaN

Author: cgevans

pandas/core/strings.py

@@ -1041,10 +1041,9 @@ def _wrap_result_expand(self, result, expand=False):
         else:
             index = self.series.index
             if expand:
-                cons_row = self.series._constructor
                 cons = self.series._constructor_expanddim
-                data = [cons_row(x) for x in result]
-                return cons(data, index=index)
+                data = [x if (x is not np.nan) else [None] for x in result]
+                return cons(data, index=index).fillna(np.nan)
             else:
                 name = getattr(result, 'name', None)
                 cons = self.series._constructor