Flipping with nested for loops#

Let’s use nested for loops to implement flip. An initial framework we can think of is as follows:

for y in range(len(img)):
    # flip the y-th row

The for loop lets the variable y take values 0,1,...,len(img)-1 one by one, which denotes row indices. The body part is to be filled in so that it flips the y-th row left to right for each iteration. Now, we can just focus on how to flip a single row. Consider the following row of pixels as a fun example.

../../../_images/flip_rows.png

Here are the steps in details to flip the row image above.

  • Swap Pixel 0 and Pixel 6

  • Swap Pixel 1 and Pixel 5

  • Swap Pixel 2 and Pixel 4

Then, it results in the following image.

../../../_images/inverse.png

How can this be written in Python? We can interpret the steps above as follows. (Remember we are temporarily at the y-th row now.)

  • Swap img[y][0] and img[y][len(img[y]) - 1]

  • Swap img[y][1] and img[y][len(img[y]) - 2]

⋮

  • Swap img[y][len(img[y])//2 - 1] and img[y][len(img[y]) - len(img[y])//2]

However, note that every step can be represented in a generalized form by introducing x, which indicates column indices.

  • x = 0: Swap img[y][x] and img[y][len(img) - 1 - x]

  • x = 1: Swap img[y][x] and img[y][len(img) - 1 - x]

⋮

  • x=len(img[y])//2 - 1: Swap img[y][x] and img[y][len(img) - 1 - x]

Notice that the pixels to be swapped are img[y][x] and img[y][len(img) - 1 - x] for all steps, regardless of what the value of x is. Therefore, we can use a for loop with the variable x taking values in range(len(img[y])//2). Here is the advanced framework incorporating the for loop with the variable x.

for y in range(len(img)):
    for x in range(len(img[y])//2):
        # Swap img[y][x] and img[y][len(img[y])-x-1]

Now, how do we swap the two pixel values img[y][x] and img[y][len(img[y])-x-1]? What happens if we just run the following?

img[y][x] = img[y][len(img[y])-x-1]
img[y][len(img[y])-x-1] = img[y][x]

This leads to a wrong outcome because the first line already deletes the original value of img[y][x], in which case it is impossible to assign the original value of img[y][x] to img[y][len(img[y])-x-1]. In order to handle this issue, create a temporary variable temp and save the value of img[y][x] first.

temp = img[y][x]

Now that temp is keeping record of what the original value of img[y][x] is, we are ready to reassign img[y][x] to a new value.

img[y][x] = img[y][len(img[y])-x-1]

Then, by assigning img[y][len(img[y])-x-1] to temp, which denotes the original value of img[y][x], we complete swapping two pixels.

img[y][len(img[y])-x-1] = temp

Here is our final code of nested for loops that flips img left to right.

for y in range(len(img)):
    for x in range(len(img[y])//2):
        temp = img[y][x]
        img[y][x] = img[y][len(img[y])-x-1]
        img[y][len(img[y])-x-1] = temp