如何通过索引列表访问多维数组的元素

Question

I would like to access a specific element of a multidimensional array (ie a tensor) via a defined list of indices. Let's say I have an array "P" in 4 dimension, and a list of indices "ind" defined as

 ind = [0,4,1,3]

which characterized one of its element. I would like to assign "something" to the element P[0,4,1,3]. However, when I do

 P[ind] = "something"

Python understands P[[0,4,1,3]] instead of P[0,4,1,3].

So, how could I access the element P[0,4,1,3] via "ind" ?

Thank you very much for your help.

Answer 1

Updated answer :

To set a value in an N-dimensional list by a given list of indices the following setter function can be used:

def set_val(ndim, ind, val):
    arr = ndim
    for i in ind[:-1]:
        arr = arr[i]
    arr[ind[-1]] = val

# usage:
P = [[[42]]]
set_val(P, [0, 0, 0,], 9001)

Original answer :

To set elements corresponding to given indices of a plain Python list you need to run a loop, something like this:

arr = [1, 2, 3, 4, 5, 6, 7, 8]
ind = [0, 4, 1, 3]
val = 42
for i in ind:
    arr[i] = val

Plain Python does not provide vector operations on lists. You might be confusing Python with Pandas.