Converting datetime days to years

So I'm trying to calculate the difference between the date now with the DOB and came up with a column 'Age'

My codes are

df = df.rename(columns={'dob Y':'Year','dob M':'Month','dob D':'Day'})
df = df.head(13)


import datetime
#df['DOB'] = pd.to_datetime(df[['Day','Month','Year']])
df['Age'] = (datetime.datetime.now()-pd.to_datetime(df[['Day','Month','Year']]))

df['Age']
# [i.days for i in df['Age']]

However I've gotten the age in days instead

0    3381 days 01:44:57.149175
1    3513 days 01:44:57.149175
2    3132 days 01:44:57.149175
3    3740 days 01:44:57.149175
4    4551 days 01:44:57.149175
5    3205 days 01:44:57.149175
6    3814 days 01:44:57.149175
7    5820 days 01:44:57.149175
8    4533 days 01:44:57.149175
9    2445 days 01:44:57.149175
10   2665 days 01:44:57.149175
11   5528 days 01:44:57.149175
12   3707 days 01:44:57.149175
Name: Age, dtype: timedelta64[ns]

I've tried to divide by 365.25 but the unit given is still days (as shown below)

0     9 days 06:09:54.223747
1     9 days 14:50:18.864404
2     8 days 13:48:13.197053
3    10 days 05:45:15.784322
4    12 days 11:02:38.084117
5     8 days 18:36:01.369537
6    10 days 10:37:00.507114
7    15 days 22:25:40.424979
8    12 days 09:51:40.178572
9     6 days 16:39:43.135451
10    7 days 07:07:04.203213
11   15 days 03:14:27.735041
12   10 days 03:35:09.624158
Name: Age, dtype: timedelta64[ns]

I'm trying to round off the Age to years in 2 decimal place. Example 12.96 years

1 answer

  • answered 2020-01-14 01:56 Jonhasacat

    Due to issues explained in this post, it is impossible to get the exact number of years from a number of days due to variations in year length.

    If you want to use 365.25 (or 365.2425, as the linked post suggests) as the number of days in a year, you could do something like this:

    df['Age'] = int((datetime.datetime.now()-pd.to_datetime(df[['Day','Month','Year']])).days / 365.2425)
    

    Including the int() cast so that we don't get a long decimal.