Clean _check_for_collision

[gran4]

Fixed unpythonic code and made clearer variable names(some are up for debate).

[einarf]

This will affect the performance profile of the function. It was optimized to extract the the positions and size into new varaibles. @cspotcode should know more about this.

It's another example that we should have brifely discussed this first 😄

[cspotcode]

Yeah, I benchmarked the code before and after, and this code made it about 20% faster. Attribute lookups and subscripting in python are relatively slow, and this code runs a huge number of times per frame. I should have added code comments to explain why we are intentionally using un-pythonic code.

I used the benchmarking script in /benchmarks/bench.sh to run the collision stress-test in /benchmarks/collisions/bench.py. I shared benchmarking results in #1524.

However, your improved variable names, like radius_sum_sq, look good to me.

[gran4]

So if I decrease the attributes look ups it should be fine.

[cspotcode]

I forget, but I do remember doing my benchmarking on 3.11. We could bench again and find out. `min()` is better than `math.min()` because the latter has an extra dictionary lookup.

…

On Sat, Apr 15, 2023, 4:27 PM Paul ***@***.***> wrote: ***@***.**** commented on this pull request. ------------------------------ In arcade/sprite_list/collision.py <#1702 (comment)>: > (sprite1_width if sprite1_width > sprite1_height else sprite1_height) + (sprite2_width if sprite2_width > sprite2_height else sprite2_height) @cspotcode <https://github.com/cspotcode> what's the overhead of this vs max on recent Python versions? — Reply to this email directly, view it on GitHub <#1702 (review)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/AAC35OAVJCMP3FH73RNZ5ZLXBMADHANCNFSM6AAAAAAW7CVA74> . You are receiving this because you were mentioned.Message ID: ***@***.***>