CODEWITHSUNDEEP
pandasdatetimetime
CollegePythonNoob
CollegePythonNoob

Reputation: 37

Calculating Start Time in Pandas from End Time and Run Time

I've just started my first internship and already been given a task that is giving me trouble. I have a dataframe with a "complete_time" and a "cycle_time" column where the cycle column is the run time in %H:%M:%S.%f.

I need to calculate the Start Time with date by subtracting the run time from the end time. I have tried using the datetime lib but that has proven unsuccessful as that wants both columns in datetime format, but the run time doesn't have a date, nor should it.

The cycle column was converted from a decimal string that originally looked like 25.2 (in seconds) using:

df['cycle_time'] = df['Cycle Time'].astype('float64')

df['cycle_time'] = pd.to_datetime(df['cycle_time'],unit='s')

df['cycle_time'] = pd.Series([val.time() for val in df['cycle_time']])

Here is the dataframe:

    complete_time           cycle_time
0   2018-05-07 17:12:34.220 00:00:25.200000
1   2018-05-07 17:12:37.807 00:00:00
2   2018-05-07 17:12:43.453 00:00:25.200000
3   2018-05-07 17:12:51.193 00:00:25.100000
4   2018-05-07 17:12:52.223 00:00:25.300000
5   2018-05-07 17:12:54.297 00:00:00
6   2018-05-07 17:12:59.430 00:00:25.200000
7   2018-05-07 17:13:03.047 00:00:00
8   2018-05-07 17:13:08.697 00:00:25.200000
9   2018-05-07 17:13:16.417 00:00:25.200000

I want to add the start_time as a new column in the dataframe.

Thanks in advance for any direction.

Upvotes: 1

Views: 481

Answers (1)

Quang Hoang
Quang Hoang

Reputation: 150735

Use pd.to_timedelta() instead of pd.to_datetime:

df['cycle_time'] = pd.to_timedelta(df['Cycle Time'].astype(float), unit='s')

df['start_time'] = df['complete_time'] - df['cycle_time']

Upvotes: 1

Related Questions