Subtract a month from a date in Python?

Question

I need to get the first day of last month from any date. I know that I can use monthdelta(datetime(2010,3,30), -1) to get last month, but it doesn't return the first day.

Because I have a lot of date `2016-05-13 19:05:38, 2016-06-18 19:05:46, 2016-06-12 19:06:46, 2016-03-13 19:07:04, 2016-02-13 19:09:59, 2016-09-13 19:11:11 ` and to all date I want to get first day of last month — NineWasps, Jul 05 '16 at 10:09

mitghi · Answer 1 · 2016-07-05T10:22:57.817

8

This can be done by first calculating the first day of current month ( or any given date ), then subtracting it with datetime.timedelta(days=1) which gives you the last day of previous month.

For demonstration, here is a sample code:

import datetime

def get_lastday(current):
    _first_day = current.replace(day=1)
    prev_month_lastday = _first_day - datetime.timedelta(days=1)
    return prev_month_lastday.replace(day=1)

edited Jul 05 '16 at 10:22

answered Jul 05 '16 at 10:11

mitghi

889
7
20

2

down voted as the current answer gives you the *last* day of last month, not the *first* – donkopotamus Jul 05 '16 at 10:21
@donkopotamus Thanks for pointing out, I did a mistake. I edited the post. – mitghi Jul 05 '16 at 10:23

score 5 · Answer 2 · answered Jul 05 '16 at 10:58

5

Try like this. With using datetime and datetutil.

(if datetutil not available for you install pip install python-dateutil)

In [1]: from datetime import datetime
In [2]: import dateutil.relativedelta
In [3]: today_date = datetime.now().date()
In [4]: today_date
Out[1]: datetime.date(2016, 7, 5)
In [5]: last_month = today_date - dateutil.relativedelta.relativedelta(months=1)
In [6]: last_mont_first_date = last_month.replace(day=1)
In [7]: last_mont_first_date
Out[2]: datetime.date(2016, 6, 1)

Input:

datetime.date(2016, 7, 5)

Output

datetime.date(2016, 6, 1)

answered Jul 05 '16 at 10:58

Rahul K P

15,740
4
35
52

I try `df['used_at'] > (dates[i] - dateutil.relativedelta.relativedelta(months=1))).replace(day=1)` but it return `TypeError: replace() got an unexpected keyword argument 'day'` – NineWasps Jul 05 '16 at 11:08
Before that I try `df['used_at'] > dates[i].replace(month=date.month - 1, day=1)` but it return only one day – NineWasps Jul 05 '16 at 11:09
I think you are trying a different things. – Rahul K P Jul 05 '16 at 11:11
`df['used_at'] > (dates[i] - dateutil.relativedelta.relativedelta(months=1)).replace(day=1)` try this working for me. – Rahul K P Jul 05 '16 at 11:15
I write about it earlier, it return `TypeError: replace() got an unexpected keyword argument 'day'` – NineWasps Jul 05 '16 at 11:22
Check the documentation https://docs.python.org/2/library/datetime.html . It'a a valid argument. – Rahul K P Jul 05 '16 at 11:27
it returns only day, when `buy=1` – NineWasps Jul 05 '16 at 11:27

JoseKilo · Answer 3 · 2016-07-05T11:37:55.887

4

Try this :

from datetime import date
d = date.today()
d.replace(
    year=d.year if d.month > 1 else d.year - 1,
    month=d.month - 1 if d.month > 1 else 12,
    day=1
)

edited Jul 05 '16 at 11:37

answered Jul 05 '16 at 10:09

JoseKilo

2,343
1
16
28

Does this work if the current month is january? – RemcoGerlich Jul 05 '16 at 11:01
i try `df['used_at'] > dates[i].replace(month=date.month - 1, day=1)` but in return only 1 day – NineWasps Jul 05 '16 at 11:09
So what about year ? – Rahul K P Jul 05 '16 at 11:16
Yeah, good point ;) – JoseKilo Jul 05 '16 at 11:38
Note that the year issue was fixed. After 3 years I wasn't sure if I had done it or not. – JoseKilo Sep 17 '19 at 14:08

Subtract a month from a date in Python?

3 Answers3