CODEWITHSUNDEEP
pythonpandasdataframedatetime
AlSub
AlSub

Reputation: 1155

Calculate time difference between a pandas dataframe column and a datetime object

I am trying to get the difference between a pandas dataframe column and a datetime object by using a customized function (years_between), here's how pandas dataframe looks like:

input_1['dataadmissao'].head(5)

0   2018-02-10
1   2009-08-23
2   2015-05-21
3   2016-12-17
4   2019-02-01
Name: dataadmissao, dtype: datetime64[ns]

And here's my code:

###################### function to return difference in years ####################

def years_between(start_year, end_year):
    start_year = datetime.strptime(start_year, "%d/%m/%Y")
    end_year = datetime.strptime(end_year, "%d/%m/%Y")
    return abs(end_year.year - start_year.year)

input_1['difference_in_years'] = np.vectorize(years_between(input_1['dataadmissao'], datetime.now()))

Which returns:

TypeError: strptime() argument 1 must be str, not Series

How could I adjust the function to return a integer which represents the difference in years between pandas dataframe column and datetime.now()?

Upvotes: 1

Views: 173

Answers (2)

DYZ
DYZ

Reputation: 57033

Simply subtract the series from datetime.datetime.now(), divide by the duration of one year, and convert to an integer:

import numpy as np
((datetime.now() - input_1['dataadmissao'])/np.timedelta64(1, 'Y')).astype(int)

Upvotes: 1

Sayandip Dutta
Sayandip Dutta

Reputation: 15872

Use pandas.Timestamp.now:

>>> df
0   2018-02-10
1   2009-08-23
2   2015-05-21
3   2016-12-17
4   2019-02-01
Name: 1, dtype: datetime64[ns]

>>> pd.Timestamp.now() - df

0   1089 days 02:41:50.467993
1   4182 days 02:41:50.467993
2   2085 days 02:41:50.467993
3   1509 days 02:41:50.467993
4    733 days 02:41:50.467993
Name: 1, dtype: timedelta64[ns]

# If you want days
>>> (pd.Timestamp.now() - df).dt.days
0    1089
1    4182
2    2085
3    1509
4     733
Name: 1, dtype: int64

# If you want years
>>> (pd.Timestamp.now().year - df.dt.year)
0     3
1    12
2     6
3     5
4     2
Name: 1, dtype: int64

Upvotes: 1

Related Questions