有效枚举 DataFrame 中每个组的 bin 中的行

Question

I am trying to find a way of flexibly changing the number of rows that I bin for each group in a pandas data frame.

Each group ID has ~700 rows and I would like to add a column called bin_number so that starting at 0 it repeats 0 for the length of the bin I desire, and then the bin_number would be 1 and repeat n times and so on.

So, say I want bin_length of 10, I would have 70 bins and the bin number would span from 0-69 repeating 10 times starting over for each ID group. The column would look something like the following:

0
0
0 (repeating bin_length number of times)
.
.
1
1
1 

Plus would be if it could be flexible to different number of rows in each group.

This is what I have been working with but it doesn't seem like the right approach.

df.groupby("ID").apply(lambda x: np.arange(len(df)) // 10)

Any pointers appreciated! Thanks!

Answer 1

Try groupby cumcout + // :

df['bins'] = df.groupby('ID').cumcount() // bin_len

Sample DF bin length of 2:

    ID  bins
0    1     0
1    1     0
2    1     1
3    1     1
4    1     2
5    1     2
6    1     3
7    1     3
8    2     0
9    2     0
10   2     1
11   2     1
12   2     2
13   2     2
14   2     3
15   2     3

---

Complete Working Example:

import numpy as np
import pandas as pd

df = pd.DataFrame({
    'ID': np.repeat(np.arange(1, 3), 8)
})

bin_len = 2

df['bins'] = df.groupby('ID').cumcount() // bin_len

print(df)