using
import pandas as pd
import numpy as np
Given this Dataframe,
df = pd.DataFrame(np.array([[1, 2, 3],
[4, 5, 6],
[7, 8, 9],
[10, 11, 12],
[13, 14, 15],
[16, 17, 18],
[19, 20, 21]
]),
columns=['a', 'b', 'c'])
Out[1]:
a b c
0 1 2 3
1 4 5 6
2 7 8 9
3 10 11 12
4 13 14 15
5 16 17 18
6 19 20 21
I want to re order and put back in place the rows 2 to 5,
2 7 8 9
3 10 11 12
4 13 14 15
5 16 17 18
If the order within the subset is [2,0,1,3] the desired result is,
Out[2]:
a b c
0 1 2 3
1 4 5 6
4 13 14 15
2 7 8 9
3 10 11 12
5 16 17 18
6 19 20 21
(I need to do this for different subsets in different orders. This is just an example.)
My attempt,
My subset,
idx = [2,3,4,5]
idx2 = np.array(idx)
The new order
i = [2,0,1,3]
if I do,
df.iloc[idx].reindex(idx2[i])
I do get the subset in the right order, then, I thought the following should work,
df.iloc[idx] = df.iloc[idx].reindex(idx2[i]).reset_index(drop=True)
but it does not, because on both sides they need to match the indices. So, I would need either to set an offset on the index, which would be a bit nasty. Or make this operation to ignore the indices on the right hand side. Any idea?
As pandas index are not mutable, you can make it as an array, modify the part of the array you want and reindex:
idx = [2,3,4,5]
i = [2,0,1,3]
# pandas index to array
arr_idx = df.index.to_numpy()
# modify the order of the array
arr_idx[idx] = arr_idx[idx][i]
# reindex
df = df.reindex(arr_idx)
print (df)
a b c
0 1 2 3
1 4 5 6
4 7 8 9
2 10 11 12
3 13 14 15
5 16 17 18
6 19 20 21