Using list comprehension for repeated function calls in Python

Question

In my current codebase, I have the following line where to_remove is a set

 [to_remove.update(b) for b in some_set if all(a <= b for a in some_dict)]

Although it works, it bothers me a little bit since it creates a list of None which is not used. Is it considered not standard? Is there a better way to do it?

Update:

Since it has been pointed out that list comprehension solely for side effect is not good, I have changed my code to

to_remove.update(itertools.chain.from_iterable(
    b for b in some_set if all(a <= b for a in some_dict))

Side effects inside a list comprehension are unusual and unexpected, which makes them a very bad idea. Anything that confuses people and has a more straightforward implementation should be avoided. — Mark Ransom, Feb 17 '17 at 19:32
You have to do it this way `to_remove.update([b for b in some_set if all(a <= b for a in some_dict)])` — Elmex80s, Feb 17 '17 at 22:43
@A-B-B Ah you made it into an iterable. Yes that is good, although might confuse (Python) beginners. — Elmex80s, Feb 17 '17 at 23:13

Asclepius · Answer 1 · 2017-02-18T02:48:47.143

2

It's not standard or recommended to use a list comprehension if its output is not assigned to a variable. A static analyzer such as pylint will even flag it.

Use a conventional loop instead:

for b in some_set:
    if all(a <= b for a in some_dict):
        to_remove.update(b)

In your specific case, since to_remove is a set, the following may or may not work correctly:

to_remove.update(b for b in some_set if all(a <= b for a in some_dict))

edited Feb 18 '17 at 02:48

answered Feb 17 '17 at 19:31

Asclepius

57,944
17
167
143

`to_remove.update(b for b in some_set if all(a <= b for a in some_dict))` does not give the same result – nos Feb 18 '17 at 02:34

Using list comprehension for repeated function calls in Python

1 Answers1