numpy array shape change / extend

user_3060835 0
Viewed 177

I'm trying to figure out how to change an array such as:

a = np.array([[1,2,3],[4,5,6],[7,8,9]])
a.shape
(3,3)

into:

a = np.array([[[1,255,255],[2,255,255],[3,255,255]],
[4,255,255],[5,255,255],[6,255,255]],
[7,255,255],[8,255,255],[9,255,255]]])
a.shape
(3,3,3)

essentially turning a single element 1 into a [1, 255, 255]

I've played around with reshape but I can't seem to get the logic to do this without a slow for loop.

numpy arrays python
1 Answers

One solution is to create a new array of the desired shape, filled with 255s using np.full, and then to just populate the index 0 value of each of the inner lists with the values from a.

import numpy as np
a = np.array([[1, 2, 3], [4, 5, 6], [7, 8, 9]])
b = np.full((3, 3, 3), 255)
b[:, :, 0] = a

Output is

array([[[  1, 255, 255],
        [  2, 255, 255],
        [  3, 255, 255]],
       [[  4, 255, 255],
        [  5, 255, 255],
        [  6, 255, 255]],
       [[  7, 255, 255],
        [  8, 255, 255],
        [  9, 255, 255]]])
user_75087000
Related