Well, in the last chapter you might notice that for the first 10 rows all the trips started before 13 hours (1 p.m.), and ended close to 00 hours. What if it's just AM/PM missing? Let's try to check it.

How can it be checked? Well, in our dataframe we also have trip_duration column representing the trip duration in seconds. What if we just add this duration to pickup_datetime and compare it with dropoff_datetime?

Please note, that you will need to use timedelta objects in this task since you want to 'shift' your datetime object for a certain time. To convert column to timedelta type use pd.to_timedelta() function with two arguments: column you want to convert, and unit - dimension used ('S' for seconds, 'm' for minutes, and so on).