CODEWITHSUNDEEP
pythonpandasdatetimedataframe
Ray_G
Ray_G

Reputation: 13

Adding minutes to timestamp in python

I have a DataFrame df with 2000 rows. The first few rows look like the following: +--------------+----------+--+ | Timestamp | Duration | | +--------------+----------+--+ | 1/5/18 0:23 | 22 | | | 1/5/18 3:34 | 32 | | | 1/5/18 23:48 | 24 | | | 1/6/18 2:26 | 15 | | | 1/6/18 14:13 | 23 | | +--------------+----------+--+

I would like to create a new column C based on the following criteria: if Duration >= 20: then add 20 mins to the timestamp. if Duration < 20: then null.

So, given the above information and data, the new table should look like: +--------------+----------+--------------+--+--+ | Timestamp | Duration | C | | | +--------------+----------+--------------+--+--+ | 1/5/18 0:23 | 22 | 1/5/18 0:43 | | | | 1/5/18 3:34 | 32 | 1/5/18 3:54 | | | | 1/5/18 23:48 | 24 | 1/6/18 0:08 | | | | 1/6/18 2:26 | 15 | Null | | | | 1/6/18 14:13 | 23 | 1/6/18 14:33 | | | +--------------+----------+--------------+--+--+

Any ideas on how to do that?

Upvotes: 1

Views: 2101

Answers (1)

cs95
cs95

Reputation: 402263

A little datetime arithmetic ought to do it.

df['C'] = pd.to_datetime(df.Timestamp) + pd.to_timedelta(
              np.where(df.Duration > 20, 20, np.nan), unit='m')


df['C']

0   2018-01-05 00:43:00
1   2018-01-05 03:54:00
2   2018-01-06 00:08:00
3                   NaT
4   2018-01-06 14:33:00
Name: C, dtype: datetime64[ns]

Feel free to format the result as needed:

df['C'] = df['C'].dt.strftime('%m/%d/%y %H:%M')

Note: if your dates start with the dayfirst, change the Timestamp conversion code to 

pd.to_datetime(df.Timestamp, dayfirst=True, errors='coerce')

So dates are handled appropriately as having the day first.

Upvotes: 1

Related Questions