Set numpy array values through list comprehension

Question

I can change a set of values of a numpy matrix, passing the indices that I want to change in a list form, e.g. matrix[[some first indices], [some second indices]] = 1

I'm trying to do this but with only one list of indices and then for each element of that list as the first index and the whole list as the other index. Here the example:

import numpy as np

#Matrix
matrix = np.zeros((5,5))

# Indices
elements = [1,2,3]

# Way that works
for i in elements:
    matrix[i, elements] = 1

I would like to do this as a list comprehension, but I cannot figure out how, and, also, I'm not sure if it is a good practice. Something like matrix[[(i,elements) for i in elements]] = 1

Also thinking about to do it with itertools, for example matrix[(itertools.permutation(elements, 2))] = 1. But any of those approaches work, and I'm not sure why.

Are you just trying to set the whole matrix to 1? If so, then just matrix += 1 — ninesalt
– ninesalt, Commented Oct 17, 2018 at 15:12
No, for every element on the list as a first index, and the entire list as a second. — Xbel
– Xbel, Commented Oct 17, 2018 at 15:15
I'm not sure what you mean by that. Can you post your expected output? — ninesalt
– ninesalt, Commented Oct 17, 2018 at 15:16
A list comprehension should be used to create a new list. While possible, it really shouldn't be used for side effects (e.g. modifying some other object). It isn't a total replacement for a loop. — hpaulj
– hpaulj, Commented Oct 17, 2018 at 15:50

Daniel · Accepted Answer · 2018-10-17 15:23:31Z

3

You need to access the 2d-matrix with row and colum arrays as index:

import numpy as np
matrix = np.zeros((5,5))
elements = [1,2,3]
matrix[np.c_[elements], elements] = 1
# array([[0., 0., 0., 0., 0.],
#   [0., 1., 1., 1., 0.],
#   [0., 1., 1., 1., 0.],
#   [0., 1., 1., 1., 0.],
#   [0., 0., 0., 0., 0.]])

answered Oct 17, 2018 at 15:23

Daniel

42.9k4 gold badges57 silver badges82 bronze badges

Sign up to request clarification or add additional context in comments.

1 Comment

Xbel Over a year ago

Thanks, actually is a much better way than with list comprehension or itertools.

DavidG · Accepted Answer · 2018-10-17 15:21:24Z

2

You can create a slice from the first to the last element (+1) of elements (provided elements is ordered):

s = slice(elements[0], elements[-1]+1)

matrix[s, elements]= 1 
# [[0. 0. 0. 0. 0.]
# [0. 1. 1. 1. 0.]
# [0. 1. 1. 1. 0.]
# [0. 1. 1. 1. 0.]
# [0. 0. 0. 0. 0.]]

answered Oct 17, 2018 at 15:21

DavidG

25.6k14 gold badges101 silver badges87 bronze badges

Collectives™ on Stack Overflow

Set numpy array values through list comprehension

2 Answers 2

1 Comment

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

1 Comment

Comments

Your Answer

Sign up or log in

Post as a guest

Related