I haven’t read the article. But I’d assume this is for the same reason that not not string is faster than bool(string). Which is to say that it has to do with having to look up a global function rather than a known keyword.
Python Performance: Why 'if not list' is 2x Faster Than Using len()
Submitted 11 months ago by abhi9u@lemmy.world to technology@lemmy.world
https://blog.codingconfessions.com/p/python-performance-why-if-not-list
Comments
Archr@lemmy.world 11 months ago
Harvey656@lemmy.world 11 months ago
I could have tripped, knocked over my keyboard, cried for 13 straight minutes on the floor, picked my keyboard back up, accidentally hit the enter key making a graph and it would have made more sense than this thing.
-2x faster. What does that even mean?
AnUnusualRelic@lemmy.world 11 months ago
There’s probably an “import * from relativity” in there somewhere.
ne0n@lemmy.world 11 months ago
Isn’t “-2x faster” 2x slower?
sugar_in_your_tea@sh.itjust.works 11 months ago
I think it was supposed to be a
~, since they use that in the paragraph below the image.Randelung@lemmy.world 11 months ago
Maybe they mean up to?
ChaoticNeutralCzech@feddit.org 11 months ago
That woulb be 0.5x. −2x implies negative duration, which makes no sense. Neither does the layout of anything else in the image.
gigachad@sh.itjust.works 11 months ago
I don’t like it very much, my variable could also be
Nonehereiknowitwheniseeit@lemmynsfw.com 11 months ago
You’d need to explicitly check for None if using the len() construct as well, so this doesn’t change the point of the article.
gigachad@sh.itjust.works 11 months ago
But
Nonehas nolenif not foo:-> foo could be an empty list or
None, it is ambiguous.len(foo)will lead to an exceptionTypeError, I can cleanly catch that.It suggests I deal with a boolean when that is not the case. Explicit is better than inplicit, and
if not footo check for an empty list may be pythonic, but it’s still implicit af
Opisek@lemmy.world 11 months ago
The graph makes no sense. Did a generative AI make it.
pyre@lemmy.world 11 months ago
yeah I got angry just looking at it
gerryflap@feddit.nl 11 months ago
Looks like it. It’s a complete fever dream graph. I really don’t get how someone can use an image like that. Personally I don’t really like AI art anyways, but I could somewhat understand it as a sort of “filler” image to make your article a bit more interesting. But a graph that is supposed to convey actual information? No idea why anyone would AI gen that without checking
sugar_in_your_tea@sh.itjust.works 11 months ago
I think there’s a good chance of that:
-2xinstead of~2x- a human is unlikely to make that mistake- no space here:
==0- there’s a space every other time it’s done, including the screenshot - the numbers are wrong - the screenshot has different data than the image
- why are there three bars? A naive approach would have two.
iknowitwheniseeit@lemmynsfw.com 11 months ago
My ad blocker has blocked all pictures on this article, so I can’t say. 😄
nutsack@lemmy.dbzer0.com 11 months ago
thanks I appreciate it
borokov@lemmy.world 11 months ago
Isn’t it because list is linked list, so to get the Len it has to iterate over the whole list whereas to get emptyness it just have to check if there is a 1st element ?
I’ too lazy to read the article BTW.
sugar_in_your_tea@sh.itjust.works 11 months ago
Then please be less lazy next time.
dreugeworst@lemmy.ml 11 months ago
why comment if you don’t even want to read the article? python lists are not linked lists, they’re contiguous with a smart growth strategy.
borokov@lemmy.world 11 months ago
I comment because this is how a social network works, and this is how you keep lemmy alive. My comment has generated a dozen of other comments, so he achieved his goal.
There is not a single question that’s already have been answered on internet, so there no point on asking anything on social plateforms except just for the sake of interacting with other peoples.
Lemmy is not stackoverflow 😉
sugar_in_your_tea@sh.itjust.works 11 months ago
Like in most reasonable languages. Linked lists would be a terrible implementation for a list where grabbing arbitrary indices is explicitly supported.
riodoro1@lemmy.world 11 months ago
So… it has to iterate over the whole empty list is what you’re saying? like once for every of the zero items in the lost?
ChaoticNeutralCzech@feddit.org 11 months ago
The list is not necessarily empty. If you were sure it was, why check?
borokov@lemmy.world 11 months ago
Don’t know how list are implemented in Python. But in the dumb linked list implementation (like C++ std::list), each element has a “next” member that point the the next element. So, to have list length, you have to do (pseudo code, not actual python code):
len = 0 elt = list.fisrt while exist(elt): elt = elt.next len++ return len
Whereas to test if list is empty, you just have to:
return exist(list.first)
Kacarott@aussie.zone 11 months ago
No,
lenis a constant time operation, at least in most cases I believe.
uis@lemm.ee 11 months ago
There are decades of articles on c++ optimizations, that say “use empty() instead of size()”, which is same as here.
dreugeworst@lemmy.ml 11 months ago
except for c++ it was just to avoid a single function call, not extra indirection. also on modern compilers size() will get inlined and ultimate instructions generated by the compiler will likely be the same
PattyMcB@lemmy.world 11 months ago
I know I’m gonna get downvoted to oblivion for this, but… Serious question: why use Python if you’re concerned about performance?
WolfLink@sh.itjust.works 11 months ago
I’ve worked on a library that’s Python because the users of said library are used to Python.
The original version of the project made heavy use of numpy, so the actual performance sensitive code was effectively C++ and fourtran, which is what numpy is under the hood.
We eventually replaced the performance sensitive part of the code with Rust (and still some fourtran because BLAS) which ended up being about 10x faster.
sugar_in_your_tea@sh.itjust.works 11 months ago
Yes, Python is the wrong choice if performance is your top priority.
But here’s another perspective: why leave easy performance wins on the table? Especially if the cost is simpler code that works as you probably wanted anyway with both
Noneand[]?Python is great if you want a really fast development cycle, because the code is generally quite simple and it’s “fast enough.” Any wins for “fast enough” is appreciated, because it delays me needing to actually look into little performance issues. It’s pretty easy for me to write a simple regex to fix this cose (
s/if len\((\w+)\) == 0:/if not \1:/), and my codebase will be slightly faster. That’s awesome! I could even write up a quickpylintorruffrule to catch these cases for developers going forward (if there isn’t one already).If I’m actively tweaking things in my Python code to get a little better performance, you’re right, I should probably just use something else (writing a native module is probably a better use of time). But the author isn’t arguing that you should do that, just that, in this case,
if not foois preferred overif len(foo) == 0for technical reasons, and I’ll add that it makes a ton of sense for readability reasons as well.Randelung@lemmy.world 11 months ago
It comes down to the question “Is YOUR C++ code faster than Python?” (and of course the reverse).
I’ve built a SCADA from scratch and performance requirements are low to begin with, seeing as it’s all network bound and real world objects take time to react, but I’m finding everything is very timely.
A colleague used SQLAlchemy for a similar task and got abysmal performance. No wonder, it’s constantly querying the DB for single results.
sugar_in_your_tea@sh.itjust.works 11 months ago
Exactly!
We rewrote some Fortran code (known for fast perf) into Python and the net result was faster. Why? They used
bubble sortin a hot loop, whereas we used Python’s built-in sort (probably qsort or similar). So despite Python being “slower” on average, good architecture matters a lot more.And your Python code doesn’t have to be 100% Python, you can write performance-critical code in something else, like C++ or Rust. This is very common, and it’s why popular Python libraries like numpy and scipy are written in a more performant language with a Python wrapper.
pastermil@sh.itjust.works 11 months ago
Because, while you don’t want to nitpick on each instruction cycle, sometimes the code runs millions of times and each microsecond adds up.
Keep in mind that people use this kind of things for work, serving real world customers who are doing their work.
Yes, the language itself is not optimal even by design, but its easy to work with, so they are making it worth a while. There’s no shortage of people who can work with it. It is easy to develop and maintain stuff with it, cutting development cost. Yes, we’re talking real businesses with real resource constraints.
sugar_in_your_tea@sh.itjust.works 11 months ago
Exactly. We picked it for the reasons you mentioned, and I still think it’s a good choice.
That said, some of our heavier logic is in a lower-level language. We had some Fortran code until recently (rewrote in Python and just ate the perf cost to lower barrier to other devs fixing stuff), and we’re introducing some C++ code in the next month or two. But the bulk of our code is in Python, because that’s what glues everything together, and the code is fast enough for our needs.
JustAnotherKay@lemmy.world 11 months ago
Honestly most people use Python because it has fantastic libraries. They optimize it because the language is middling, but the libraries are gorgeous
ThirdConsul@lemmy.ml 11 months ago
Honestly most people use Python because it has fantastic libraries
In C++ if I remember correctly…
jerkface@lemmy.ca 11 months ago
Alternatively, why wait twice as long for your python code to execute as you have to?
lengau@midwest.social 11 months ago
It’s all about trade-offs. Here are a few reasons why one might care about performance in their Python code:
- Performance is often more tied to the code than to the interpreter - an O(n³) algorithm in blazing fast C won’t necessarily perform any better than an O(nlogn) algorithm in Python.
- Just because this particular Python code isn’t particularly performance constrained doesn’t mean you’re okay with it taking twice as long.
- Rewriting a large code base can be very expensive and error-prone. Converting small, very performance-sensitive parts of the code to a compiled language while keeping the bulk of the business logic in Python is often a much better value proposition.
These are also performance benefits one can get essentially for free with linter rules.
Anecdotally: in my final year of university I took a computational physics class. Many of my classmates wrote their simulations in C or C++. I would rotate between Matlab, Octave and Python. During one of our labs where we wrote particle simulations, I wrote and ran Octave and Python simulations in the time it took my classmates to write their C/C++ versions, and the two fastest simulations in the class were my Octave and Python ones, respectively. (The professor’s own sim came in third place). The overhead my classmates had dealing with poorly optimised code that caused constant cache misses was far greater than the interpreter overhead in my code (though at the time I don’t think I could have explained why their code was so slow compared to mine).
uis@lemm.ee 11 months ago
- Performance is often more tied to the code than to the interpreter - an O(n³) algorithm in blazing fast C won’t necessarily perform any better than an O(nlogn) algorithm in Python.
An O(n³) algorithm in Python won’t necessarily perform any better than an O(nlogn) algorithm in C. Ever heard of galactic algorithms?
PattyMcB@lemmy.world 11 months ago
I appreciate the large amount of info. Great answer. It just doesn’t make sense to me, all things being equal (including performant algorithms), why choose Python and then make a small performance tweak like in the article? I understand preferring the faster implementation, but it seems to me like waxing your car to reduce wind resistance to make it go faster, when installing a turbo-charger would be much more effective.
Takapapatapaka@lemmy.world 11 months ago
You may want to beneficiate from little performance boost even though you mostly don’t need it and still need python’s advantages. Being interested in performance isnt always looking for the very best performance there is out of any language, it can also be using little tips to go a tiny bit faster when you can.
Reptorian@lemmy.zip 11 months ago
I have the same question. I prefer other languages. I use G’MIC for image processing over Python and C++.
antlion@lemmy.dbzer0.com 11 months ago
Could also compare against:
if not len(mylist)That way this version isn’t evaluating two functions. The bool evaluation of an integer is false when zero, otherwise true.
sugar_in_your_tea@sh.itjust.works 11 months ago
That’s worse. IMO, solve this problem with two things:
- type hint
mylistaslist | Noneor justlist - use
if not mylist:
The first documents intent and gives you static analysis tools some context to check for type consistency/compatibility, and the second shows that
Nonevs empty isn’t an important distinction here.- type hint
FooBarrington@lemmy.world 11 months ago
This is honestly the worst version regarding readability. Don’t rely on implicit coercion, people.
antlion@lemmy.dbzer0.com 11 months ago
But the first example does the same thing for an empty list. I guess the lesson is that if you’re measuring the speed of arbitrary stylistic syntax choices, maybe Python isn’t the best language for you.
knighthawk0811@lemmy.ml 11 months ago
so these are the only 2 ways then? huge if true
sugar_in_your_tea@sh.itjust.works 11 months ago
Oh, there are plenty of other terrible ways:
for _ in mylist: break else: # whatever you'd do if mylist was empty
if not any(True for _ in mylist):try: def do_raise(): raise ValueError _ = [do_raise() for _ in mylist] except ValueError: pass else: # whatever you'd do i mylist was empty
I could probably come up with a few others as well.
iAvicenna@lemmy.world 11 months ago
Yea and then you use “not” with a variable name that does not make it obvious that it is a list and another person who reads the code thinks it is a bool. Hell a couple of months later you yourself wont even understand that it is a list. You should not sacrifice code readability for over optimization, this is phyton after all I don’t think list lengths will be your bottle neck.
LegoBrickOnFire@lemmy.world 11 months ago
I really dislike using boolean operators on anything that is not a boolean. I recently made an esception to my rule and got punished… Yeah it is skill issue on my part that I tried to check that a variable equal to 0 was not None using “if variable…”. But many programming rules are there to avoid bugs caused by this kind of inattention.
sugar_in_your_tea@sh.itjust.works 11 months ago
That’s why we use type-hinting at my company:
def do_work(foo: list | None): if not foo: return ...
Boom, self-documenting, faster, and very simple.
LegoBrickOnFire@lemmy.world 11 months ago
Well, in your case it is not clear whether you intended to branch in the variable foo being None, or on the list being empty which is semantically very different…
Thats why it’s better to explicitelly express whether you want an empty collection (len = 0) or a None value.
Artyom@lemm.ee 11 months ago
In my experience, if you didn’t write the function that creates the list, there’s a solid chance it could be
Nonetoo, and if you try to check the length ofNone, you get an error. This is also why returningNonewhen a function fails is bad practice IMO, but that doesn’t seem to stop my coworkers.LegoBrickOnFire@lemmy.world 11 months ago
Passing None to a function expecting a list is the error…
iAvicenna@lemmy.world 11 months ago
good point I try to initialize None collections to empty collections in the beginning but not guaranteed and len would catch it
acosmichippo@lemmy.world 11 months ago
if you’re worried about readability you can leave a comment.
sugar_in_your_tea@sh.itjust.works 11 months ago
Better yet, a type hint.
list | Nonecan be checked by static analysis,# foo is a listisn’t.chunkystyles@sopuli.xyz 11 months ago
Comments shouldn’t explain code. Code should explain code by being readable.
Comments are for whys. Why is the code doing the things it’s doing. Why is the code doing this strange thing here. Why does a thing need to be in this order. Why do I need to store this value here.
Stuff like that.
thebestaquaman@lemmy.world 11 months ago
There is no guarantee that the comment is kept up to date with the code. “Self documenting code” is a meme, but clearly written code is pretty much always preferable to unclear code with a comment, largely because you can actually be sure that the code does what it says it does.
Note: You still need to comment your code kids.
iAvicenna@lemmy.world 11 months ago
If there is an alternative through which I can achieve the same intended effect and is a bit more safer (because it will verify that it has len implemented) I would prefer that to commenting. Also if I have to comment every len use of not that sounds quite redundant as len checks are very common
jerkface@lemmy.ca 11 months ago
Strongly disagree that
not ximplies to programmers thatxis a bool.jj4211@lemmy.world 11 months ago
In context, one can consider it a bool.
Besides, I see c code all the time that treats pointers as bool for the purposes of an if statement. !pointer is very common and no one thinks that means pointer it’s exclusively a Boolean concept.
sugar_in_your_tea@sh.itjust.works 11 months ago
Maybe, but that serves as a very valuable teaching opportunity about the concept of “empty” is in Python. It’s pretty intuitive IMO, and it can make a lot of things more clear once you understand that.
That said, larger projects should be using type hints everywhere, and that should make the intention here painfully obvious:
def do_work(foo: list | None): if not foo: ... handle empty list ... ...
That’s obviously not a boolean, but it’s being treated as one. If the meaning there isn’t obvious, then look it up/ask someone about Python semantics.
I’m generally not a fan of learning a ton of jargon/big frameworks to get the benefits of more productivity (e.g. many design patterns are a bit obtuse IMO), but learning language semantics that are used pretty much everywhere seems pretty reasonable to me. And it’s a lot nicer than doing something like this everywhere:
if foo is None or len(foo) == 0:JustAnotherKay@lemmy.world 11 months ago
Doesn’t matter what it implies. The entire purpose of programming is to make it so a human doesn’t have to go do something manually.
not xtells me I need to go manually check what typexis in Python.len(x) == 0tells me that it’s being type-checked automaticallyacosmichippo@lemmy.world 11 months ago
i haven’t programmed since college 15 years ago and even i know that 0 == false.
taladar@sh.itjust.works 11 months ago
It does if you are used to sane languages instead of the implicit conversion nonsense C and the “dynamic” languages are doing
iAvicenna@lemmy.world 11 months ago
well it does not imply directly per se since you can “not” many things but I feel like my first assumption would be it is used in a bool context
sirber@lemmy.ca 11 months ago
How does Python know of it’s my list or not?
thebestaquaman@lemmy.world 11 months ago
I write a lot of Python. I hate it when people use “X is more pythonic” as some kind of argument for what is a better solution to a problem. I also have a hang up with people acting like python has any form of type safety, instead of just embracing duck typing.This lands us at the following:
The article states that “you can check a list for emptiness in two ways:
if not mylistorif len(mylist) == 0”. Already here, a fundamental mistake has been made: You don’t know (and shouldn’t care) whethermylistis a list. These two checks are not different ways of doing the same thing, but two different checks altogether. The first checks whether the object is “falsey” and the second checks whether the object has a well defined length that is zero. These are two completely different checks, which often (but far from always) overlap. Embrace the duck type- type safe python is a myth.
AnUnusualRelic@lemmy.world 11 months ago
From that little image, they’re happy it takes a tenth of a second to check if a list is empty?
What kind of dorito chip is that code even running on?