Python, numpy sort array

Question

I'am using numpy and have an array (ndarray type) which contain some values. Shape of this array 1000x1500. I reshaped it

brr = np.reshape(arr, arr.shape[0]*arr.shape[1])

when I try

brr.reverse()

I get the following error:

AttributeError: ‘numpy.ndarray’ object has no attribute ‘reverse’

How I can sort this array ?

Doesn't solve your problem, but to make your code a little neater and more efficient you can use arr.size instead of arr.shape[0]*arr.shape[1]. — Junuxx
– Junuxx, Commented Feb 14, 2013 at 12:59
I want to sort it descending. Btw sort() give the same error — user2046488
– user2046488, Commented Feb 14, 2013 at 13:12

Thorsten Kranz · Accepted Answer · 2013-02-14 13:38:33Z

30

If you just want to reverse it:

brr[:] = brr[::-1]

Actually, this reverses along axis 0. You could also revert on any other axis, if the array has more than one.

To sort in reverse order:

>>> arr = np.random.random((1000,1500))
>>> brr = np.reshape(arr, arr.shape[0]*arr.shape[1])
>>> brr.sort()
>>> brr = brr[::-1]
>>> brr
array([  9.99999960e-01,   9.99998167e-01,   9.99998114e-01, ...,
     3.79672182e-07,   3.23871190e-07,   8.34517810e-08])

or, using argsort:

>>> arr = np.random.random((1000,1500))
>>> brr = np.reshape(arr, arr.shape[0]*arr.shape[1])
>>> sort_indices = np.argsort(brr)[::-1]
>>> brr[:] = brr[sort_indices]
>>> brr
array([  9.99999849e-01,   9.99998950e-01,   9.99998762e-01, ...,
         1.16993050e-06,   1.68760770e-07,   6.58422260e-08])

edited Feb 14, 2013 at 13:38

answered Feb 14, 2013 at 12:57

Thorsten Kranz

12.8k2 gold badges45 silver badges57 bronze badges

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

3 Comments

user2046488 Over a year ago

No, it is unsorted, I have to sort it descending

user2046488 Over a year ago

Thanks! Tried argsort() with no luck. Did not knew about [::-1]

Nirmal Over a year ago

brr = np.sort(brr)[::-1]

Hima · Accepted Answer · 2014-07-28 13:48:02Z

16

Try this for sorting in descending order ,

import numpy as np
a = np.array([1,3,4,5,6])
print -np.sort(-a)

answered Jul 28, 2014 at 13:48

Hima

12.2k3 gold badges24 silver badges33 bronze badges

Comments

rafaelvalle · Accepted Answer · 2017-06-02 02:19:16Z

4

To sort a 1d array in descending order, pass reverse=True to sorted. As @Erik pointed out, sorted will first make a copy of the list and then sort it in reverse.

import numpy as np
import random
x = np.arange(0, 10)
x_sorted_reverse = sorted(x, reverse=True)

edited Jun 2, 2017 at 2:19

answered Jan 11, 2015 at 17:36

rafaelvalle

7,1233 gold badges36 silver badges39 bronze badges

2 Comments

Erik Over a year ago

sorted uses python's iterable sort, which will first copy all of the data into a list and then do the sort in reverse. While this does indeed result in a reverse sorting of the data, it doesn't do so in numpy, and generally removes any benefit of using numpy.

rafaelvalle Over a year ago

Thanks for sharing this, Erik. I added it to the answer.

Collectives™ on Stack Overflow

Python, numpy sort array

3 Answers 3

3 Comments

Comments

2 Comments

Your Answer

Linked

Hot Network Questions

Collectives™ on Stack Overflow

3 Answers 3

3 Comments

Comments

2 Comments

Your Answer

Sign up or log in

Post as a guest

Linked

Related