Raphaël Simon

Bon, je fais quoi maintenant que j’ai fini le jeu ? Je sens que ce truc de récap quotidien ça ne va pas suffire pour continuer la progression

Quad digits 🤙

sparcopen.org/our-work/big-d… Inspiring read about MIT’s success in ditching Elsevier

What did I do wrong for my twitter feed to suddenly only show me crazy conspiracy theories / chemtrails and all the usual insane crowd ?! How do I back :-(

@JosiahParry Come back in August for conf ! I’d love to buy you a coffee as thank you for all the documentation you share about R + Rust :-)

I’ve been writing small utils functions to work with {duckdb} functions in R (pivot, insert, enums …), and the collection is growing large enough I feel like a {duckdbextra} pkg is creeping up on me. Idk if I should make it or contribute them to the R {duckdb} repo ? #rstats

@JosiahParry Will you do a write up of the full process ? Or are willing to talk about it maybe ?

@JosiahParry I was on my way to update again my comments after testing : indeed i can distribute the pkg to colleagues if I build it as a binary ! Thank you for all the ressources you publish !!

@JosiahParry I believe I understand my own question after further reading : it’s because gifski has no rust code, it’s a C header to call Rust compiled code from C, while uwu for example needs run to run the rust code. Sorry for the spam

@JosiahParry Hello, thank you for all the Rust-R ressources. I’m strugling to understand why this needs Rust installed, while a pkg like « gifski » (r-rust GH) doesn’t ? I’m trying to convince my IT to install Rust on my machine but if my colleagues (aka users) also need Rust i’m doomed :-(

@antoine_fabri @JosiahParry Thank you so much @antoine_fabri for this gist, I didnt know about this and it completely solves my current issue with dispatching based on smthg else than the first arg of caller … I didn’t know it was possible after reading advR and my google-fu was weak apparently. Cheers !!

@JosiahParry If you're as obsessed with robustness then performance you must have a fat stack of tests! this looks dangerous! That's a very cool design problem so I was nerd sniped, here's an idea.

Going strong 💪

jamanetwork.com/journals/jama/… Intéressant, et je serais curieux de voir la même étude en France !

Not bad !

Dites les pédiatres il se passe quoi avec le Zymad pédia pour que ça cause autant de discussion enflammées entre parents sur le net ? Entre les discours sur les huiles essentielles, les désordres intestinaux, les compléments alimentaires … il y a de la littérature ? #doctoctoc

I heavily dislike that my timeline keeps resting to « for you » instead of « following ». Actually, I heavily dislike the « for you » timeline.

✅ That’s it for me at #positconf2023 ! What a great conf, I loved it! See you next year in Seattle ?

Last day for me at #positconf2023 today. ✈️ Quick run along the riverside walk to start the day, straight into @kara_woo’s keynote. I wish I could stay longer, but I also can’t wait to come back & dive deep into the things I learned here 🤓 I’ll def come back to Chicago !!

The first keynote of the day at #positconf2023 was excellent, discussing the challenges of setting up a robust and scalable data science env in an org. I’m a tad disappointed by the 2nd keynote, which is a prerecorded video about LLMs that feels more like a Youtube tutorial :/

✅ Day 2 @ #positconf2023 🛠️ Workshop : pkg dev masterclass by @hadleywickham and @JennyBryan, focusing on the news in {testthat} 3.0. Testing, mocking & defensive programming ! 🛡️ 👀 Pharma is moving away from SAS to R, Pharmaverse is 🔥 ⭐️ Chicago pizza 🍕 !