Get week numbers on multiple year that is ready for plotting in pandas

Question

Here's my data

I was doing sheet2['device_create_week'] = sheet2['device_create_at'].dt.week here's the result

        device_create_at        device_create_week
136     2014-08-27 17:29:23     35
245     2015-09-06 15:46:00     39
257     2014-09-29 22:26:34     40
258     2014-11-05 13:02:18     40
480     2020-02-02 17:59:54     6
481     2020-02-02 17:59:54     6
482     2020-02-02 17:59:54     6

But I want

        device_create_at        device_create_week
136     2014-08-27 17:29:23     2014-35
245     2015-09-06 15:46:00     2015-39
257     2014-09-29 22:26:34     2014-40
258     2014-11-05 13:02:18     2014-40
480     2020-02-02 17:59:54     2020-6
481     2020-02-02 17:59:54     2020-6
482     2020-02-02 17:59:54     2020-6

Other solution is fine as long as it is plotable, and I can see growth every week

Use `df.device_create_at.dt.strftime('%Y-%U')` or `%Y-%W`? – Zero Oct 12 '17 at 05:20 — Zero, Oct 12 '17 at 05:20

jezrael · Accepted Answer · 2017-10-12T05:28:08.490

2

Use year + week:

sheet2['device_create_week'] = sheet2['device_create_at'].dt.year.astype(str) + '-' + 
                               sheet2['device_create_at'].dt.week.astype(str)

print (sheet2)

       device_create_at device_create_week
136 2014-08-27 17:29:23            2014-35
245 2015-09-06 15:46:00            2015-36
257 2014-09-29 22:26:34            2014-40
258 2014-11-05 13:02:18            2014-45
480 2020-02-02 17:59:54             2020-5
481 2020-02-02 17:59:54             2020-5
482 2020-02-02 17:59:54             2020-5

Another solution with strftime with V:

sheet2['device_create_week'] = sheet2['device_create_at'].dt.strftime('%Y-%V')

print (sheet2)
       device_create_at device_create_week
136 2014-08-27 17:29:23            2014-35
245 2015-09-06 15:46:00            2015-36
257 2014-09-29 22:26:34            2014-40
258 2014-11-05 13:02:18            2014-45
480 2020-02-02 17:59:54            2020-05
481 2020-02-02 17:59:54            2020-05
482 2020-02-02 17:59:54            2020-05

edited Oct 12 '17 at 05:28

answered Oct 12 '17 at 05:17

jezrael

822,522
95
1,334
1,252

Oh my god, I not think to convert to string – Nabih Ibrahim Bawazir Oct 12 '17 at 05:18
Yes, but datetime `2014-35` does not exist :( – jezrael Oct 12 '17 at 05:32
1

The 2nd one works great for `sheet2.groupby(['device_create_week']).size().reset_index(name='device created count weekly')` – Nabih Ibrahim Bawazir Oct 12 '17 at 05:32
datetime `2014-35` do not exist? I don't get it – Nabih Ibrahim Bawazir Oct 12 '17 at 05:34

Get week numbers on multiple year that is ready for plotting in pandas

1 Answers1

Linked