Creating datetime in pandas from year and julian day

问题

                  ad_name    adl_name  year  JD 
0  united_states_of_america  colorado  2000   1 
1  united_states_of_america  colorado  2000   2 
2  united_states_of_america  colorado  2000   3 
3  united_states_of_america  colorado  2000   4 
4  united_states_of_america  colorado  2000   5 

how do I add a datetime column using the year and JD (julian day) columns? I was trying to use:

pd.to_datetime(df, format='%Y_%d'), but that does not work

回答1:

You need add to year column JD converted to_timedelta:

df['date'] = pd.to_datetime(df.year, format='%Y') + pd.to_timedelta(df.JD - 1, unit='d')
print (df)
                    ad_name  adl_name  year  JD       date
0  united_states_of_america  colorado  2000   1 2000-01-01
1  united_states_of_america  colorado  2000   2 2000-01-02
2  united_states_of_america  colorado  2000   3 2000-01-03
3  united_states_of_america  colorado  2000   4 2000-01-04
4  united_states_of_america  colorado  2000   5 2000-01-05

Sample with JD=32 and JD=366:

print (df)
                    ad_name  adl_name  year   JD
0  united_states_of_america  colorado  2000   32
1  united_states_of_america  colorado  2000  366

df['date'] = pd.to_datetime(df.year, format='%Y') + pd.to_timedelta(df.JD - 1, unit='d')
print (df)
                    ad_name  adl_name  year   JD       date
0  united_states_of_america  colorado  2000   32 2000-02-01
1  united_states_of_america  colorado  2000  366 2000-12-31

来源：https://stackoverflow.com/questions/38067348/creating-datetime-in-pandas-from-year-and-julian-day