CODEWITHSUNDEEP
pythonarraysnumpyreshape
5ive
5ive

Reputation: 273

Reshaping Numpy Array from (3, 2, 3) to (3, 3, 2)

I have the follow 3d numpy array in shape (3, 2, 3). 

[
 [[ 0  1  2]
  [ 3  4  5]]

 [[ 6  7  8]
  [ 9 10 11]]

 [[12 13 14]
  [15 16 17]]
             ]

However, I need to reshape into (3, 3, 2) in the follow order:

[
 [[ 0  3]
  [ 6  9]
  [12 15]]

 [[ 1  4]
  [ 7 10]
  [13 16]]

 [[ 2  5]
  [ 8 11]
  [14 17]]
          ]

I am currently using Jupyter with a lot of trial and error.

Thank you for any suggestions!

Upvotes: 1

Views: 1968

Answers (1)

Eric Duminil
Eric Duminil

Reputation: 54213

To define the original array, you can use :

np.arange(18).reshape(3,2,3)

As mentioned by @Divakar in the comments, you can use :

np.arange(18).reshape(3,2,3).transpose(2,0,1)

to get the desired result.

From np.transpose documentation:

axes: By default, reverse the dimensions, otherwise permute the axes according to the values given.

2,0,1 is the permutation needed to go from (3,2,3) shape to (3,3,2). It would also convert a (3400, 7, 100) shape into (100, 3400, 7).

Another method would be to use np.rollaxis (another hint from @Divakar):

np.rollaxis(np.arange(18).reshape(3,2,3),2)

Upvotes: 3

Related Questions