Assign a TimeZone to Dataframe Datetime Column-CodePudding

I cannot figure out how to assign a timezone to a dataframe. I have a df called 'df' that looks like this:

df
Out[67]: 
           date  daily_flow
0    2002-02-13    144000.0
1    2002-02-14    184000.0
2    2002-02-15    159000.0
3    2002-02-16    126000.0
4    2002-02-17    114000.0
        ...         ...
7277 2022-02-02    152000.0
7278 2022-02-03    159000.0
7279 2022-02-04    150000.0
7280 2022-02-05    165000.0
7281 2022-02-06    148000.0

[7282 rows x 2 columns]

df.dtypes
Out[68]: 
date          datetime64[ns]
daily_flow           float64
dtype: object

I have read the documentation and other posts and it is not clear how to assign the "date" column to a timezone such as 'US/Pacific'. thank you!

Here is an example of an error I keep getting when I try and assign a timezone (UTC) to the datetime column in the index position ('date').

df.date.tz_localize('UTC')
Traceback (most recent call last):

  Input In [10] in <module>
    df.tz_localize('UTC')

  File ~\Anaconda3\envs\ARIMA\lib\site-packages\pandas\core\generic.py:9977 in tz_localize
    ax = _tz_localize(ax, tz, ambiguous, nonexistent)

  File ~\Anaconda3\envs\ARIMA\lib\site-packages\pandas\core\generic.py:9959 in _tz_localize
    raise TypeError(

TypeError: index is not a valid DatetimeIndex or PeriodIndex

CodePudding user response：

To set a time zone for a column, parse to_datetime and use the dt accessor of the Series to tz_localize. EX:

df
Out[3]: 
         date  daily_flow
0  2002-02-13    144000.0
1  2002-02-14    184000.0
2  2002-02-15    159000.0

df['date'] = pd.to_datetime(df['date']).dt.tz_localize('UTC')

df['date']
Out[5]: 
0   2002-02-13 00:00:00 00:00
1   2002-02-14 00:00:00 00:00
2   2002-02-15 00:00:00 00:00
Name: date, dtype: datetime64[ns, UTC]

In the example, you can replace 'UTC' with the appropriate time zone. You can also convert to another time zone using the same approach (.dt.tz_convert('your-time-zone')).