example: df
value date_time 1 2020.12.1 1 2020.12.2 <--need to be dropped 1 2020.12.3 <--need to be dropped 2