转换熊猫系列和日期时间对象

Question

I have a series of dates with format:

df['myDateTimes']
0                       NaT
1       2017-07-23 00:26:50
2                       NaT
3       2017-07-31 04:07:24

(where the first number is just the pandas dataframe index) and I'd like to convert these to Modified Julian Dates

from datetime import datetime, timedelta
import julian
import datetime
dtime  = julian.to_jd(df['myDateTimes'], fmt='jd')

just gives a

AttributeError: 'Series' object has no attribute 'month'

df['myDateTimes'] is a pandas series and (I think) julian.to_jd needs a datetime.datetime object.

Answer 1

Using apply , since you have NaT julian will not expected that , so we filter before we transfer the datetime to julian

s[s.notna()].apply(lambda x : julian.to_jd(x,fmt='jd')).reindex(s.index)
Out[139]: 
0             NaN
1    2.457958e+06
2             NaN
3    2.457966e+06
dtype: float64