A very visual and great introduction to Monte Carlo simulation by @MarbleScience
youtube.com/watch?v=7ESK5SaP…
Looks like I'm not the only one who gets +50Gb of binaries when compiling #Rust projects... 😅
Replying to @marktenenholtz
I think there is something that would be more impactful for AI libraries than removing the GIL. Allowing libraries to overwrite `and`, `or` and `is` operators (see PEP-335):
Used #pixi by @prefix_dev for the first time, and loved the experience.
Way simpler than #conda, and seems fast (my environment was quite small). I guess we're at the beginning of another great migration to a #rustlang tool, like with #ruff.
prefix.dev/
Last day of the #pandas sprint. Today's discussions:
- Funding
- More NaN vs NULL
- Data types API
- User survey
- Feedback from @Novartis users (our sprint host)
- Governance
- Roadmap and sprint recap
Tomorrow: #DataFrameSummit2023
df-summit-23.pydata.org/
First results of the tests to bring Rust extensions to @pandas_dev / @ApacheArrow are very promising.
Computing the average distance of a Series to a scalar is an order of magnitude faster when implementing the operation as a whole instead of using generic #dataframe functions.











