Skip to content

Dateint

It's very common to store date/datetimes as integers or strings using formats such as YYYYmmdd or YYYYmm. In python, to perform date/datetime arithmetic on those values, one needs to:

  1. convert the original value to date or datetime
  2. perform the date/datetime operation
  3. convert the result back to the original format

With dateint, we abstract all convertion operations so you can focus on the arithmetic step:

  • single value:

    import dateint as di

di.add(20220510, days=15)
# 20220525

  • pandas:

    import dateint as di
import pandas as pd

dates = pd.Series([202201, 202202, 202203])

di.add(dates, months=2)
'''
0    202203
1    202204
2    202205
dtype: int64
'''

Installation

pip install dateint

Examples

Add a time interval to a date

import dateint as di
import pandas as pd

di.add(20220510, days=15)
# 20220525

dates = pd.Series([202201, 202202, 202203])
di.add(dates, months=2)
'''
0    202203
1    202204
2    202205
dtype: int64
'''

Subtract a time interval from a date

import dateint as di
import pandas as pd

di.sub(20220510, days=8)
# 20220502

dates = pd.Series([202201, 202202, 202203])
di.sub(dates, years=1)
'''
0    202101
1    202102
2    202103
dtype: int64
'''

Day of week

import dateint as di
import pandas as pd

di.sub(20220510, days=8)
# 20220502

df = pd.DataFrame(
    {
        'date': [
            20220103,
            20220104,
            20220105,
            20220106,
            20220107,
            20220108,
            20220109,
        ]
    }
)

df['weekday'] = di.weekday(df['date'])
df
'''
       date  weekday
0  20220103        0
1  20220104        1
2  20220105        2
3  20220106        3
4  20220107        4
5  20220108        5
6  20220109        6
'''

License

This project is licensed under the terms of the MIT license.