I am learning Python 3 and I saw that you can zip 2 lists together, like this:
l1 = [1,2,3]
l2 = ['a','b','c']
l3 = zip(l1,l2)
print(list(zip_L1L2))
# Output: [(1, 'a'), (2, 'b'), (3, 'c')]
And then reverse the process like that:
l4, l5 = zip(*l3)
but I cannot wrap my head around why it works...
While I was trying to write my question clearly, I stumbled upon this answer Why does x,y = zip(*zip(a,b)) work in Python?, which gave me a hint that I still did not get before trying it in a REPL. The key is unpacking the result of zip:
print(*zip(l1,l2))
# Output: (1, 'a') (2, 'b') (3, 'c')
We get 3 tuples of 2 elements, which are treated like 3 lists of 2 elements passed as arguments to zip:
l1 = [ 1 , 2 ]
l2 = ['a','b']
l3 = ['+','-']
print(list(zip(l1,l2,l3)))
# Output: [(1, 'a', '+'), (2, 'b', '-')]
※ Note that if you use 3 lists of 4 elements, zip will output 4 lists of 3 elements
l1 = [ 1 , 2 , 3 , 4 ]
l2 = ['a','b','c','d']
l3 = ['+','-','*','/']
print(list(zip(l1,l2,l3)))
# Output: [(1, 'a', '+'), (2, 'b', '-'), (3, 'c', '*'), (4, 'd', '/')]
I now think of it as rotating values in a matrix 90° to make rows into columns and columns into rows.
It is like rotating a matrix by providing the rows as arguments instead of the matrix itself.
The last piece of the puzzle is to unpack the result and assign each tuple in its own variable using multiple assignment!
l1 = [1,2,3]
l2 = ['a','b','c']
l3,l4= zip(*zip(l1,l2))
print('l3= ',l3)
print('l4= ', l4)
# Output: l3= (1, 2, 3)
# l4= ('a', 'b', 'c')
I decided to keep writing this Q&A to learn by explaining what I had discovered by experimenting, after not immediately getting the meaning of the answer I found while writing.
I also hope that my analogy with matrix rotation will help someone else get it faster in the future.