How to get the values of NumPy array that are NOT in my list of indices

Question

I am trying to find values in an np.array corresponding to an index and then those that are not in the index. In R, I would do

> x <- c(1,2,3,4,5)
> ind = c(1,4)
> x[ind]
[1] 1 4
> x[-ind]
[1] 2 3 5

In Python, I can index and get the values in the index in the following manner

import numpy as np
x = np.array([1,2,3,4,5])
ind = [0,3]
x[ind]

However, in Python how can I get the values NOT in my index as I show in the R code using x[-ind]?

I found the SO post below and have replicated this, but is this indeed the "right" or best approach? NumPy array, change the values that are NOT in a list of indices

What do you mean by "right or best"? Best in terms of what? That answer works, so what else exactly do you want to know? — Georgy, Apr 27 '20 at 23:01

Björn · Accepted Answer · 2020-04-27T22:13:42.613

3

I think the most intuitive way would be with np.delete

np.delete(x, ind)

As it gives you back the remainig elements

array([2, 3, 5])

edited Apr 27 '20 at 22:13

answered Apr 27 '20 at 22:12

Björn

I like that - but it is inplace. – modesitt Apr 27 '20 at 22:13
1

@modesitt from the docs: A copy of arr with the elements specified by obj removed. Note that delete does not occur in-place. If axis is None, out is a flattened array. – hilberts_drinking_problem Apr 27 '20 at 22:23
Nice - my bad. Don't know why i thought was in place – modesitt Apr 27 '20 at 22:26

score 2 · Answer 2 · answered Apr 27 '20 at 22:11

2

I do not think there is a solution quite as pretty as Rs, but how about

x[~np.isin(np.arange(len(x)), ind)]

answered Apr 27 '20 at 22:11

modesitt

2 Answers2