I have an array like this:
[[1 2 3 4]
[9 8 7 6]]
and want to upscale like this:
[[1 0 0 2 0 0 3 0 0 4 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0]
[9 0 0 8 0 0 7 0 0 6 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0]
[0 0 0 0 0 0 0 0 0 0 0 0]]
Try with numpt.pad, numpy.insert and numpy.kron but cant access the answer.
You can just use assignment here. If n might be different for width and height, just index using result[::h, ::w].
n = 3
result = np.zeros(np.array(a.shape) * n, dtype=int)
result[::n, ::n] = a
array([[1, 0, 0, 2, 0, 0, 3, 0, 0, 4, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[9, 0, 0, 8, 0, 0, 7, 0, 0, 6, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0],
[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0]])
If you really want to use kron, a similar principal can be applied, by making a mask with only the desired element set to 1.
kernel = np.zeros((3, 3), dtype=int)
kernel[0, 0] = 1
np.kron(a, kernel)
