Numpy array reshape element-wise-CodePudding

I have a 3-D array size = (3,2,3)

[
 [[1, 2, 3],[4, 5, 6]],
 [[7, 8, 9],[10,11,12]],
 [[13,14,15],[16,17,19]]
]

How to reshape to (3,3,2):

[
 [[1,4], [2,5], [3,6]],
 [[7,10], [8,11], [9,12],],
 [[13,16],[14,17],[15,19]]
]

CodePudding user response：

x = [[[1, 2, 3],[4, 5, 6]],
     [[7, 8, 9],[10,11,12]],
     [[13,14,15],[16,17,19]]]
res = [[[i,j] for (i,j) in zip(sub[0], sub[1])] for sub in x]

CodePudding user response：

You task is not to reshape the array. You have to swap the last axis (the third dimension of your array) with the second.

import numpy as np

#input
arr = np.array([
 [[1, 2, 3],[4, 5, 6]],
 [[7, 8, 9],[10,11,12]],
 [[13,14,15],[16,17,19]]
])

#output
np.moveaxis(arr, 2, 1)

#an alternative is
np.swapaxes(arr, 1, 2)