Converting a column in pandas with timestamps into local time knowing the timezone

Asked Oct 24 '17 at 19:52

Active Oct 24 '17 at 19:52

Viewed 385 times

Suppose we have a pandas dataframe that have tow columns, one contains timestamps which are in GMT timezone but that is not known by the data and another column that contains timezones as follows

df = pd.DataFrame({'Time': [2016-02-16 13:59:12, 2015-07-22 15:07:06], 'Time_zone': ['Asia/Bangkok', 'America/Mexico_City']})

i want to convert the timestamps in the first column into local time using the timestamp in the second column and store that into a new column named 'local_time'

I tried the following

import datetime
import pytz

df['Time'].apply(lambda x: x.tz_localize('GMT'))  

df['local_time'] = df['Time'].apply(lambda x: x.astimezone(pytz.timezone(df['Time_zone'].apply(lambda x: x))))

but it is giving the follwing error

'Series' object has no attribute 'upper'

any help is very much appreciated

asked Oct 24 '17 at 19:52

Martin Heusen

Have you tried converting the df elements to `datetime` objects? – pstatix Oct 24 '17 at 19:55
Yes, the data in the 'Time' column is in the 'pandas._libs.tslib.Timestamp' format. I know that it is in GMT but it doesn't know that about itself – Martin Heusen Oct 24 '17 at 19:57
can you please right down the whole answer with all the apply functions and so on, if you advice on using apply() – Martin Heusen Oct 24 '17 at 20:02
I don't have a whole answer, hence the comments. – pstatix Oct 24 '17 at 20:02
module 'datetime' has no attribute 'fromtimestamp' got this error – Martin Heusen Oct 24 '17 at 20:04
That was the class with method, you must include the module: `datetime.datetime.fromtimestamp(aTimeStamp, tz=aTimeZone)` – pstatix Oct 24 '17 at 20:06

Converting a column in pandas with timestamps into local time knowing the timezone

0 Answers0