is there any methode while doing resampling() to ffill() or bfill() a object column?
Suppose we have:
| Date | Sort | Value |
|---|---|---|
| 2022-10-23 15:40:41 | A | 1 |
| 2022-10-23 18:43:13 | B | 2 |
| 2022-10-24 15:40:41 | C | 3 |
| 2022-10-24 18:43:13 | D | 4 |
i would like to have following results with:
df.resample("15min").mean()
| Date | Sort | Value |
|---|---|---|
| 2022-10-23 15:45:00 | A | 1 |
| 2022-10-23 16:00:00 | A | 1 |
| 2022-10-23 16:15:00 | A | 1 |
| 2022-10-23 16:35:00 | A | 1 |
| ... | ... | ... |
| 2022-10-23 18:00:00 | D | 1 |
| 2022-10-23 18:15:00 | D | 1 |
| 2022-10-23 18:30:00 | D | 1 |
| 2022-10-23 18:45:00 | D | 1 |
but it always kick out the "sort column". would be nice if anyone here can help!
best M.
CodePudding user response：
You can specify the aggregation functions for your columns separately, for example:
df = df.resample("15min").agg({"Sort": min, "Value": np.mean}).ffill()
Output:
Sort Value
Date
2022-10-23 15:30:00 A 1.0
2022-10-23 15:45:00 A 1.0
2022-10-23 16:00:00 A 1.0
2022-10-23 16:15:00 A 1.0
2022-10-23 16:30:00 A 1.0
... ... ...
2022-10-24 17:30:00 C 3.0
2022-10-24 17:45:00 C 3.0
2022-10-24 18:00:00 C 3.0
2022-10-24 18:15:00 C 3.0
2022-10-24 18:30:00 D 4.0
CodePudding user response：
If need forward filling first valeus per Sort and mean per Value use:
df = df.resample("15min").agg({'Sort':'first', 'Value':'mean'}).ffill()
print (df)
Sort Value
Date
2022-10-23 15:30:00 A 1.0
2022-10-23 15:45:00 A 1.0
2022-10-23 16:00:00 A 1.0
2022-10-23 16:15:00 A 1.0
2022-10-23 16:30:00 A 1.0
... ...
2022-10-24 17:30:00 C 3.0
2022-10-24 17:45:00 C 3.0
2022-10-24 18:00:00 C 3.0
2022-10-24 18:15:00 C 3.0
2022-10-24 18:30:00 D 4.0
[109 rows x 2 columns]