How do I define a week start frequency in Pandas?

Question

I am trying to come up with a frequency in Pandas that represents the start of a calendar week (configurable by week start). For example, for all dates from 2025-01-06 (Monday) to 2025-01-13 (Sunday), I want them to be categorized under Monday's date of 2025-01-06.

However, because of how Pandas DateOffset rolls dates forward to the next period, not all the dates in that week are categorized under 2025-01-06.

You can see what I'm describing in this sample example:

>>> import pandas as pd
>>> 
>>> range_dex = pd.date_range(start="2025-01-06", end="2025-01-30")
>>> sample_series = pd.Series(1, index=range_dex)
>>> sample_series
2025-01-06    1
2025-01-07    1
2025-01-08    1
...
2025-01-28    1
2025-01-29    1
2025-01-30    1
Freq: D, dtype: int64
>>> 
>>> week_start = 0  # 0 for Monday, ..., 6 for Sunday
>>> freq_pd = pd.offsets.Week(n=1, weekday=week_start)
>>> 
>>> sample_series.groupby(pd.Grouper(freq=freq_pd)).size()
2025-01-06    1
2025-01-13    7
2025-01-20    7
2025-01-27    7
2025-02-03    3
Freq: W-MON, dtype: int64

The result I am hoping to achieve instead is:

>>> sample_series.groupby(pd.Grouper(freq=freq_pd)).size()
2025-01-06    7
2025-01-13    7
2025-01-20    7
2025-01-27    4
Freq: W-MON, dtype: int64

I've also tried pd.offsets.Week(n=0, weekday=week_start) and pd.offsets.Week(n=-1, weekday=week_start), which both error, and I can't do the typical math I could do on a Series with something like series + pd.offsets.Week(n=1, weekday=week_start) - pd.offsets.Week(n=1, weekday=week_start). I also have a use case for .asfreq so really want this specified in terms of a frequency, if possible. Thanks!

For reference, this is the calendar for January 2025:

LMC · Accepted Answer · 2025-06-06 20:46:32Z

0

The series can be grouped by the weekofyear attribute

grp = sample_series.groupby([x.weekofyear for x in sample_series.index])
# also possible
# grp = sample_series.groupby(by={i:i.weekofyear for i in range_dex})

for g in grp:
    print(g[1].size, g[1].index[0])

7 2025-01-06 00:00:00
7 2025-01-13 00:00:00
7 2025-01-20 00:00:00
4 2025-01-27 00:00:00

Series.groupby docs

If a dict or Series is passed, the Series or dict VALUES will be used to determine the groups

Similar use case using a frequency defined by a string:

range_dex = pd.date_range(start="2025-01-06", freq='D', periods=26)
sample_series = pd.Series(range_dex, index=range_dex)

grp = sample_series.groupby(pd.Grouper(freq='1W-MON', label='left', origin="start"))

print(grp.size())

2024-12-30    1
2025-01-06    7
2025-01-13    7
2025-01-20    7
2025-01-27    4
Freq: W-MON, dtype: int64

edited Jun 6 at 20:46

answered Jun 6 at 15:19

LMC

14.4k3 gold badges34 silver badges62 bronze badges

Sign up to request clarification or add additional context in comments.

Collectives™ on Stack Overflow

How do I define a week start frequency in Pandas?

1 Answer 1

Comments

Your Answer

Hot Network Questions

Sun	Mon	Tue	Wed	Thu	Fri	Sat
29	30	31	1	2	3	4
5	6	7	8	9	10	11
12	13	14	15	16	17	18
19	20	21	22	23	24	25
26	27	28	29	30	31	1

Sun	Mon	Tue	Wed	Thu	Fri	Sat
29	30	31	1	2	3	4
5	6	7	8	9	10	11
12	13	14	15	16	17	18
19	20	21	22	23	24	25
26	27	28	29	30	31	1

Collectives™ on Stack Overflow

1 Answer 1

Comments

Your Answer

Sign up or log in

Post as a guest

Related

Sun	Mon	Tue	Wed	Thu	Fri	Sat
29	30	31	1	2	3	4
5	6	7	8	9	10	11
12	13	14	15	16	17	18
19	20	21	22	23	24	25
26	27	28	29	30	31	1