List comprehension in python

It’s an interesting simple algorithm for understanding list comprehension in python3.

Here we got a matrix below, how can we get transposed matrix without any API methods or functions?

matrix = [ [1, 2, 3, 4],[5, 6, 7, 8],[9, 10, 11, 12]]

Here is a fantasy solution in one line

[[row[i] for row in matrix] for i in range(len(matrix[0]))]`

[[1, 5, 9], [2, 6, 10], [3, 7, 11], [4, 8, 12]]

How does this work? I think it’s a liitle tricky to understand this command, so before understanding it thoroughly, we can begin it from a simple one.

Problem: how to flatten the origin matrix to [ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12 ]

solution

[row[i] for row in matrix for i in range(len(matrix[0]))]

=> [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]

Maybe you will find something familiar, YES, the solutions of flatten matrix and transpose matrix have striking similarity!

[[row[i] for row in matrix] for i in range(len(matrix[0]))]  ##transpose
[ row[i] for row in matrix  for i in range(len(matrix[0]))]  ##flatten

Well, you now can clearly see the difference, let’s explain the flatten first.

flatten_list = []
for row in matrix:
  for i in range(len(maxtrix[0])):
    flatten_list.append(row(i))

The solution for flatten is equivalent to code above, it’s not hard to understand, there are 2 for in loop.

transpose_list = []
for i in range(len(maxtrix[0])):
  transpose_row = []
  for row in matrix:
    transpose_row.append(row(i))
  transpose_list.append(transpose_row)

Above is equivalent to transposed solution, which is a liitle more complicated comparing with the other, because list comprehension(1st is for i in range(…)) is nested by another list comprehension (2nd is for row in matrix), conversely, solution for flatten is only one list comprehension, please take care of the difference.