Computer vision: Nearest-neighbour resize

The quickest way to resize is to copy the nearest pixel.

What to build

A program that resizes an image with nearest-neighbour sampling.

Requirements

  • Output pixel (y, x) copies source pixel (y × H // newH, x × W // newW).
  • Print newH rows of newW values.

Input

Line 1: H, W, newH and newW. Then H rows of W values.

Output

newH rows.

Constraints

1 ≤ H, W, newH, newW ≤ 60.

Example

Input:

2 2 4 4
1 2
3 4

Output:

1 1 2 2
1 1 2 2
3 3 4 4
3 3 4 4

Why it works

Each output position is scaled back into the source and rounded down to a whole pixel, so every output pixel copies exactly one source pixel.

Examples

Example 1 · Input
2 2 4 4
1 2
3 4
Expected output
1 1 2 2
1 1 2 2
3 3 4 4
3 3 4 4
Example 2 · Input
4 4 2 2
1 2 3 4
5 6 7 8
9 10 11 12
13 14 15 16
Expected output
1 3
9 11

Prompts inside input() are ignored when checking, so input("n: ") is fine.

Hints

Hint 1

Output row y comes from source row y × H // newH, and output column x from source column x × W // newW.

Hint 2

Build each output row with a comprehension over x.

Solution

Show one correct solution (try it yourself first)
h, w, nh, nw = map(int, input().split())
img = [input().split() for _ in range(h)]
for y in range(nh):
    print(" ".join(img[y * h // nh][x * w // nw] for x in range(nw)))

Lesson: Resizing and interpolation →