Logo Questions Linux Laravel Mysql Ubuntu Git Menu
 

How to transpose a 3D matrix?

I have a 3D matrix x_test of size (100, 33, 66) and I want to change its dimensions to (100, 66, 33).

What is the most efficient way to do this using python3.5? I look for something along those lines:

y = x_test.transpose()
like image 210
Ajees Avatar asked Nov 24 '17 09:11

Ajees


1 Answers

You can pass the desired dimensions to the function np.transpose using in your case np.transpose(x_test, (0, 2, 1)).

For example,

import numpy as np

x_test = np.arange(30).reshape(3, 2, 5)

print(x_test)
print(x_test.shape)

This will print

[[[ 0  1  2  3  4]
  [ 5  6  7  8  9]]

 [[10 11 12 13 14]
  [15 16 17 18 19]]

 [[20 21 22 23 24]
  [25 26 27 28 29]]]
(3, 2, 5)

Now, you can transpose the matrix with the command from above

y = np.transpose(x_test, (0, 2, 1))
print(y)
print(y.shape)

which will give

[[[ 0  5]
  [ 1  6]
  [ 2  7]
  [ 3  8]
  [ 4  9]]

 [[10 15]
  [11 16]
  [12 17]
  [13 18]
  [14 19]]

 [[20 25]
  [21 26]
  [22 27]
  [23 28]
  [24 29]]]
(3, 5, 2)
like image 112
Cleb Avatar answered Sep 22 '22 20:09

Cleb