What is proper convention for repeating tuples?

Question

For lack of a better title, I want to find the most proper way to write (0, 0, 0, 0).

I am simply asking if (0,) * 4 is acceptable, and should it be done?

Are there any specific cases for this?

What if it could be either RGB, or RGBA? Should I do (0,) * n?

If this were in a loop, how significant would the overhead be?

Thanks.

Disclaimer: I am not asking for opinions, I am asking for a general consensus, or any written specification. Do not give your sole opinion without any support, or this post runs the risk of being closed due to going off-topic.

Answer 1

If you had 100 zeroes to write in a tuple or list would you ask yourself this question?

Since the contents of the tuple are immutable elements, it's equivalent (it's faster to parse - see the end of the post - and the intent is clearer) to do (0,) * 4 instead of (0,0,0,0). It's beginning to become ridiculous for 2 elements but that's just my opinion :).

If the number increases it may save some precious debugging time because you missed one count/paste.

For mutable types don't do that!, but you can do a similar thing: ex:

• [[] for _ in range(n)] for list of lists
• tuple([] for _ in range(n)) for tuple of lists

Note that the (0,)*n form is definitely faster to parse. It's tricky to time such constructs because doing this naively skips the parsing part. But using evil eval helps in that case:

import time

n=100000

start_time = time.time()
for _ in range(n):
    eval("(0)*50")

print(time.time()-start_time)

fifty_zeros_tuple = "({})".format("0,"*50)  # generate (0,0,0,...) expression to pass to eval, not clocked to avoid bias

start_time = time.time()
for _ in range(n):
    eval(fifty_zeros_tuple)

print(time.time()-start_time)

no photo-finish on the results, (0,)*50 is many times faster:

0.764380931854248
5.553457021713257