I am currently working on grouping/aggregating data based on date range for a weekly plot.
Below is how my dataframe looks like for Daily data:
| daily_dates | registered | attended |
|---|---|---|
| 02/10/2022 | 0 | 0 |
| 02/09/2022 | 0 | 0 |
| 02/08/2022 | 1 | 0 |
| 02/07/2022 | 1 | 0 |
| 02/06/2022 | 20 | 06 |
| 02/05/2022 | 05 | 03 |
| 02/04/2022 | 15 | 12 |
| 02/03/2022 | 10 | 08 |
| 02/02/2022 | 10 | 05 |
The first day of the week I'd want is Sunday.
My current code to perform weekly group is:
weekly_df = weekly_df.resample('w').sum().reset_index()
The output I am desiring is:
| weekly_dates | registered | attended |
|---|---|---|
| 02/06/2022 | 22 | 06 |
| 01/30/2022 | 40 | 28 |
A bit of explanation about desired output - the reason for 02/06/2022 & 01/30/2022 is that both these dates are start date of that respective week which is a sunday. And for the week of 01/30/2022 only 02/05/2022|05|03, 02/04/2022, 02/03/2022, 02/02/2022 dates are considered as those are the one's in the daily dataframe.
My current implementation follows the instructions provided here.
I am looking for any suggestion to achieve my Desired Output
CodePudding user response：
Try:
df.resample('W-SUN', label='left', closed='left').sum().reset_index()
Output:
daily_dates registered attended
0 2022-01-30 40 28
1 2022-02-06 22 6