It's easy to create a 2D array with filled values:

import numpy as np

np.full((5, 3), [1])
np.full((5, 3), [1, 2, 3])

Then, I wanna create a 3D array with same value for last two dimensions:

import numpy as np

np.full((2, 3, 1), [[1], [2]])

'''
# perferred result
[[[1],
  [1],
  [1]]
 [[2],
  [2],
  [2]]]
'''

However, I got this error:

ValueError: could not broadcast input array from the shape (2,1) into shape (2,3,1)

Does anyone know the correct way to use np.full() for 3D array?

CodePudding user response：

In order to boardcast the value to the desired shape, you require the value in shape (2, 1, 1) to match with the input shape (2, 3, 1)

np.full((2, 3, 1), [[[1]], [[2]]])

output:

array([[[1],
        [1],
        [1]],

       [[2],
        [2],
        [2]]])