Replacing above diagonal elements in a numpy array

Question

If I want to create a 5x5 zero matrix with values of 10, 20, 30, 40 just above the diagonal I can do the following:

import numpy as np
np.diag((1+np.arange(4))*10,k=1)

but how can i replace the elements above the diagonal in a 5x5 random matrix with the same array 10, 20, 30, 40 ? I have tried to use the numpy where function which works with 1D arrays like:

import numpy as np
array1 = np.array([2, 2, 2, 0, 2, 0, 2])
print np.where(array1==0, 1, array1)

but I cannot make it working in higher dimensions. I can manually assign the values, but i am looking for a better solution.

Quang Hoang · Accepted Answer · 2021-02-11 16:40:48Z

1

You can try advance indexing:

a = np.arange(25).reshape(5,5)

s = np.arange(len(a))
a[s[:-1], s[1:]] = [10,20,30,40]

Output:

array([[ 0, 10,  2,  3,  4],
       [ 5,  6, 20,  8,  9],
       [10, 11, 12, 30, 14],
       [15, 16, 17, 18, 40],
       [20, 21, 22, 23, 24]])

answered Feb 11, 2021 at 16:40

Quang Hoang

151k11 gold badges64 silver badges86 bronze badges

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

Comments

danuzco · Accepted Answer · 2021-02-11 19:12:18Z

Maybe this works. For example, for this array:

arr = np.random.rand(5,5)
print(arr)

[[0.63267449 0.81436882 0.49014052 0.85241815 0.39175126]
 [0.79926876 0.46784356 0.64146423 0.24392249 0.70449611]
 [0.28667995 0.58503395 0.80665148 0.84331471 0.10687276]
 [0.59349235 0.23448985 0.25971096 0.60335227 0.31760505]
 [0.10723313 0.44694671 0.99660858 0.31529209 0.42713487]]

with np.diag(arr, k=1) you get the diagonal above the main diagonal.

diag = np.diag(arr, k=1)

you can get the indexes of the elements in diag using np.isin(...) and then replace those entries with [10, 20, 30, 40 ].

idxs = np.isin(arr, diag).nonzero()
arr[idxs] = np.array([10, 20, 30, 40 ], dtype = np.float)
arr

array([[ 0.63267449, 10.        ,  0.49014052,  0.85241815,  0.39175126],
       [ 0.79926876,  0.46784356, 20.        ,  0.24392249,  0.70449611],
       [ 0.28667995,  0.58503395,  0.80665148, 30.        ,  0.10687276],
       [ 0.59349235,  0.23448985,  0.25971096,  0.60335227, 40.        ],
       [ 0.10723313,  0.44694671,  0.99660858,  0.31529209,  0.42713487]])

Collectives™ on Stack Overflow

Replacing above diagonal elements in a numpy array

2 Answers 2

Comments

Comments

Your Answer

Hot Network Questions

Collectives™ on Stack Overflow

2 Answers 2

Comments

Comments

Your Answer

Sign up or log in

Post as a guest

Related