1>>> df.reset_index(level=['tick', 'obs'])
2 tick obs val
3tag
4C 2016-02-26 2 0.0139
5A 2016-02-27 2 0.5577
6C 2016-02-28 6 0.0303
7
1>>> df
2 val
3tick tag obs
42016-02-26 C 2 0.0139
52016-02-27 A 2 0.5577
62016-02-28 C 6 0.0303
7