python pandas time series year extraction

Question

I have a DF containing timestamps:

0     2005-08-31 16:39:40
1     2005-12-28 16:00:34
2     2005-10-21 17:52:10
3     2014-01-28 12:23:15
4     2014-01-28 12:23:15
5     2011-02-04 18:32:34
6     2011-02-04 18:32:34
7     2011-02-04 18:32:34

I would like to extract the year from each timestamp, creating additional column in the DF that would look like:

0     2005-08-31 16:39:40 2005
1     2005-12-28 16:00:34 2005
2     2005-10-21 17:52:10 2005
3     2014-01-28 12:23:15 2014
4     2014-01-28 12:23:15 2014
5     2011-02-04 18:32:34 2011
6     2011-02-04 18:32:34 2011
7     2011-02-04 18:32:34 2011

Obviously I can go over all DF entries stripping off the first 4 characters of the date. Which is very slow. I wonder if there is a fast python-way to do this. I saw that it's possible to convert the column into the datetime format by DF = pd.to_datetime(DF,'%Y-%m-%d %H:%M:%S') but when I try to then apply datetime.datetime.year(DF) it doesn't work. I will also need to parse the timestamps to months and combinations of years-months and so on... Help please. Thanks.

Answer 1

No need to apply a function for each row there is a new datetime accessor you can call to access the year property:

In [35]:

df1['year'] = df1['timestamp'].dt.year
df1
Out[35]:
            timestamp  year
0 2005-08-31 16:39:40  2005
1 2005-12-28 16:00:34  2005
2 2005-10-21 17:52:10  2005
3 2014-01-28 12:23:15  2014
4 2014-01-28 12:23:15  2014
5 2011-02-04 18:32:34  2011
6 2011-02-04 18:32:34  2011
7 2011-02-04 18:32:34  2011

If your timestamps are str then you can convert to datetime64 using pd.to_dateime:

df['timestamp'] = pd.to_datetime(df['timestamp'])

You can access the months and other attributes using dt like the above.

For version prior to 0.15.0 you can perform the following:

df1['year'] = df1['timestamp'].apply(lambda x: x.year)