Pandas：根据 DataFrame 中的其他列在 DataFrame 中创建新列

Question

I have a pandas DataFrame called df that has the following data:

Index    SourceDate
0        AUG_2013
1        SEP_2013
2        JAN_2012

I need to add an additional column that turns each of these dates into the following ConvertedDate Column. This column will have the date in YYYY-MM-DD format with the day always being 01.

Index    SourceDate    ConvertedDate
0        AUG_2013      2013-08-01
1        SEP_2013      2013-09-01
2        JAN_2012      2012-01-01

I attempted doing this with:

df['ConvertedDate'] = time.strptime(str.replace(str.rsplit(df.SourceDate,'_',1)[0],'_','-01-'),'%b-%d-%Y')

Unfortunately this does not work since df.SourceDate is a Series, and string functions won't work on a Series.

Answer 1

Use to_datetime and pass a format string:

In [64]:
df['ConvertedDate'] =pd.to_datetime(df['SourceDate'], format='%b_%Y')
df

Out[64]:
   Index SourceDate ConvertedDate
0      0   AUG_2013    2013-08-01
1      1   SEP_2013    2013-09-01
2      2   JAN_2012    2012-01-01

The python datetime format string specifiers can be found here