I would like to do list comprehension only on the second and third element of each sub-list. I've been able to do this, but I lose the first element. I know I could do this pretty easily with a for loop, but I'd like to be pythonic (idiomatic python) and use list comprehension.
test = [[3,6,9],[3,6,9]]
v = [[x/3.0 for x in y[1:3]] for y in test]
print v
[[2.0, 3.0], [2.0, 3.0]]
[[3, 2.0, 3.0], [3, 2.0, 3.0]]
Just include y[:1] + in the result:
>>> test = [[3,6,9],[3,6,9]]
>>> [y[:1] + [x/3.0 for x in y[1:3]] for y in test]
[[3, 2.0, 3.0], [3, 2.0, 3.0]]
It looks like you fundamentally misunderstand comprehensions. They don't modify the original in-place, they build a new one. That's why, if you only take a part of the original, you only get a part of the original.
If you want to modify certain parts, then a loop might serve you better and there's nothing unpythonic about it.
An alternative solution is to check whether you are dealing with the first element or not using a ternary operator:
v = [[elem/3.0 if idx!=0 else elem for idx, elem in enumerate(k)] for k in test]
This case would serve for all values of the list except for the first.
You could use the following list comprehension
>>> [[i[0]] + [j/i[0] for j in i[1:]] for i in test]
[[3, 2.0, 3.0], [3, 2.0, 3.0]]
Note this does not prevent against divide-by-zero if you are expecting the first element of each sublist to possibly be zero.
