通过 numpy.reshape() 交换 3D numpy 数组的列和行

Question

Now I have a 3D numpy array with shape (2,3,4) as follows:

[[[ 0  1  2  3]
  [ 4  5  6  7]
  [ 8  9 10 11]]

 [[12 13 14 15]
  [16 17 18 19]
  [20 21 22 23]]]

Now, I want to reshape the array to (2,4,3) by swapping the last 2 dimensions of the array as follows:

[[[ 0  4  8]
  [ 1  5  9]
  [ 2  6 10]
  [ 3  7 11]]

 [[12 16 20]
  [13 17 21]
  [14 18 22]
  [15 19 23]]]

In the documentation of numpy.reshape, there are 2 types of orders for reshaping numpy arrays, including 'C' and 'F'. I tried to use them and the results are as follows:

1. For order='C':

    [[[ 0  1  2]
      [ 3  4  5]
      [ 6  7  8]
      [ 9 10 11]]
    
     [[12 13 14]
      [15 16 17]
      [18 19 20]
      [21 22 23]]]

2. For order='F':

 [[[ 0  5 10]
  [ 4  9  3]
  [ 8  2  7]
  [ 1  6 11]]

 [[12 17 22]
  [16 21 15]
  [20 14 19]
  [13 18 23]]]

I have also tried to do the reshaping consecutively, but still couldn't get my desired shape. Do anyone know how to reshape the array in the desired order?

Answer 1

It can be done with a transpose on the last two axes:

arr = np.array([[[ 0 , 1,  2,  3],
                 [ 4 , 5,  6,  7],
                 [ 8 , 9, 10, 11]],
                [[12, 13, 14, 15],
                 [16, 17, 18, 19],
                 [20, 21, 22, 23]]])

np.transpose(arr, axes = (0,2,1))

array([[[ 0,  4,  8],
        [ 1,  5,  9],
        [ 2,  6, 10],
        [ 3,  7, 11]],

       [[12, 16, 20],
        [13, 17, 21],
        [14, 18, 22],
        [15, 19, 23]]])