Efficiently reshape/reorder numpy array from (a, b, c) to (a, c, b)

Question

I have several Numpy 3D arrays of the shape ( a , b , c ). The values of a , b , and c are unknown. However, I want to reshape each of the arrays to ( a , c , b ) in an efficient way.

Here is what I am doing:

for array in list_of_arrays:
    a, b, c = array.shape
    array = array.reshape(a, c, b)

Is there a more efficient way to do this, possibly in one line of code? Can I use the -1 indexing method to reshape/reorder the arrays?

Thank you.

Answer 1

import numpy as np

# Example array with shape (2, 4, 6)    
array = np.arange(48).reshape((2, 4, 6))

# Swap axis in the 1st and 2nd dimension and print out its shape     
np.swapaxis(array, 1, 2).shape

Output:

(2, 6, 4)

Answer 2

Maybe np.transpose ? It swaps all dimensions to the specified order.

x = np.random.randint(0, 256, (100, 80, 3))

np.transpose(x, (1, 0, 2))

(80, 100, 3)