Unable to load image
Reported by:

python 3.11 just dropped, claims to be 10-60% faster than 3.10

https://docs.python.org/3.11/whatsnew/3.11.html

inshallah this shall shave off like 2ms from my AOC time tonight :marseyinshallah:

69
Jump in the discussion.

No email address required.

Doesn't affect me because my reseach work depends on a bunch of specific numerical calculations performed by a half dozen packages written by some neurodivergent that break if you use anything beyond Python 2.7.

Jump in the discussion.

No email address required.

surely it would be worth port them over to 3 in 2022

Jump in the discussion.

No email address required.

sciencecels cant actually write software they just make the worlds worst glueshit, call it "codes" in their paper and then i guess they lose it when their laptop breaks

Jump in the discussion.

No email address required.

God the r packages that aren’t maintained by legit cs academics make me want to:marseyrope:

Jump in the discussion.

No email address required.

I’ve heard that’s very common with jupyter

Jump in the discussion.

No email address required.

What actually is the use case for jupyter? I know what it does and what the notebooks look like but what's an actual use case?

Also I'm still not entirely convinced that 'data science' is actually a thing separate from some weird rebranding of statistics. I get a weird sort of crypto-bro vibe coming from the whole 'data scientist' set.

Jump in the discussion.

No email address required.

Couldn’t tell you :marseyshrug#:

Not an academiacel (sadly?)

Jump in the discussion.

No email address required.

>academic research work

>dogshit and outdated practices and no maintenance done

Checks out

Jump in the discussion.

No email address required.

Which are all implemented in C++ anyways

Jump in the discussion.

No email address required.

Autogenerated C++ with heaps of adhoc manual tweaks. :marseychefkiss:

Jump in the discussion.

No email address required.

just move them to numpy you numpy

Jump in the discussion.

No email address required.

The data is in numpy arrays and pandas dfs. The issues that break everything arise from the specific packages deciding to implement everything with numba, and I have no time nor desire to refactor everything to not use that.

Jump in the discussion.

No email address required.

lol. lmao even. imagine not writing your own padLeft

Jump in the discussion.

No email address required.

Link copied to clipboard
Action successful!
Error, please refresh the page and try again.