[FIXED] What is the best way to perform an anti-transpose in python?

Issue

Lets say I have an array

a = np.arange(16).reshape((4,4))

0   1  2   3
4   5  6   7
8   9  10  11
12  13 14  15

But I want

15  11  7  3
14  10  6  2
13  9   5  1
12  8   4  0

which is a flip across the secondary diagonal, or a kind of anti-transpose.

How can I do this in numpy?

Solution

One could do one of the following:

rot90(a,2).T

rot90(flipud(a),1)

rot90(fliplr(a), -1)

or as hpaulj suggested in the comments (thanks hpaulj)

a[::-1,::-1].T

Here are the speed rankings as ratios of the slowest method after anti-transposing 1000 random 10000x10000 arrays.

1. 63.5% - a[::-1,::-1].T
2. 85.6% - rot90(a,2).T
3. 97.8% - rot90(flipud(a),1)
4. 100% -rot90(fliplr(a),-1)

Answered By - Jaden Travnik