@masalmon.eu
https://masalmon.eu/ 🧰 #Rstats / research software engineer. 🗒️ Blogger. 📦 Software review editor for @ropensci. 💜 #RLadies. 📈 PhD in statistics. 🍋 Nancy, France (let's say this emoji is a bergamot orange).
The @rconsortium.bsky.social has ROR! 💫
ror.org/01z833950
So, if you have received a grant for your package, you can now acknowledge it by adding it to the DESCRIPTION file.
Here are more details:
ropensci.org/blog/2025/05...
#RStats
Aww this is so nice to say, thank you 😊
📝 [blog] How do we keep software peer review human-centered in the age of generative AI?
That’s the question behind our latest post at rOpenSci, where we describe:
• new expectations around AI use
• what this means for authors, reviewers, and editors
• why transparency matters more than ever […]
Table showing whether 3 Git commands can change files only, history only, both files and history. git restore can only change files, git revert can change files only (--no-commit) and both files and history (default), git reset can change the history only (--mixed, --soft) and both files and history (--hard)
Today, I gave a talk about what to replace git checkout with:
🌳 git switch to change branches
↩️ git restore to restore a file to a given state
Including demos of the commands with {saperlipopette}
masalmon.eu/talks/2026-0...
rOpenSci Coworking and Office Hours coming up in an hour!
Today's Theme: rOpenSci Editors with cohost @mpadge
Tuesday March 3rd 14:00 Europe Central (13:00 UTC)
https://ropensci.org/events/coworking-2026-03
#RStats
@rstats@a.gup.pe
folks, please don't submit LLM-generated PRs to open source projects. It makes no sense.
If the maintainers want to use an LLM to fix an issue, they can use Claude or whatnot directly. They don't need you as intermediary, that's just silly.
If they don't want to use LLMs, they have reasons.
Grumpy cat "Before I knew how to extend roxygen2". Happy cat "Once I got how to extend roxygen2".
On the right you see the steps of roxygenize(), on the left different files and objects. First roxygenize() loads packages. Then it parses R scripts, recognizing normal and custom tags. This creates blocks which are sorts of lists of tags. Then for any registered roclet: Rd roclet, collate roclet, NAMESPACE roclet, your custom roclet, roxygen2 will run the preprocess methods then the process methods then the output methods. The Rd roclet work will lead to the man/ folder, the NAMESPACE roclet work generates the NAMESPACE file, your custom roclet could have done anything like generating test files or a report on statistical software standards.
Meme "Doctors hate him!". plumber2 extends roxygen2 without roclets. With this one weird trick! Learn the truth now.
Table of uses and extensions of roxygen2 with these two dimensions: parsing code decorators, generating stuff when calling roxygenize(). The most populated cell has the two aspects, it contains every example we saw except plumber2. Plumber2 does use code decorator parsing but not through roxygenize(). If you do neither of those things you’re clearly not in roxygen2-land. Now there’s a weird thing where you can use roxygenize() without using information from the code decorator parsing. Roxygen2 has an old roclet that is no longer recommended, the vignette roclet that would update the built versions of your vignettes using Make. It’s no longer recommended to have those files in your package source. In any case, you could make roxygenize() run any code from a roclet: maybe formatting your code? Updating your changelog? I can’t think of actual good ideas but you can do it.
Yesterday I got to present what I painstakingly learnt about
✨ extending roxygen2 ✨
masalmon.eu/talks/2026-0...
#RStats
A minimalist isometric 3D illustration on a dark background showing a central glowing cube labeled ".dotfiles" with a git icon. Glowing neon lines radiate from the cube to icons representing "Shell Config," "Editor & Tooling," and "AI Agents (Claude & OpenCode)." A sidebar labeled "Skills Marketplace" displays specialized icons for R, Shiny, Quarto, and writing voices, illustrating a unified development environment.
My dotfiles now manage my AI coding agents 🤖 Same instructions for Claude Code and OpenCode, symlinked from one repo. Here's how I set it up: https://drmo.site/CCrf37
¡Miércoles, Git! 💩
Taller práctico para aprender a resolver 7 errores comunes en Git y ganar confianza desde R,
Enseñado en LatinR 2025 con @yabellini.bsky.social
¡Ahora con vídeo!
masalmon.eu/talks/2025-1...
[blog] Guia de desenvolvimento da rOpenSci 1.0.0: trilíngue e aprimorado 🥳
A orientação da revisão por pares de Software rOpenSci está reunida em um livro on-line que continua melhorando! Esta postagem do blog resume o que há de novo em nosso Dev Guide 1.0.0 […]
[blog] rOpenSci Dev Guide 1.0.0: Trilingüe y mejorada 🥳
Las orientaciones de rOpenSci para realizar la revisión de software por pares se recopilan en un libro en línea ¡que sigue mejorando! Este artículo resume las novedades de nuestra Guía de Desarrollo 1.0.0 […]
[blog] rOpenSci Dev Guide 1.0.0: Trilingual and Improved 🥳
rOpenSci Software Peer Review’s guidance is gathered in an online book that keeps improving! This blog post summarises what’s new in our Dev Guide 1.0.0,
https://ropensci.org/blog/2026/03/02/devguide-1.0.0/
The future of #DataScience is at #positconf 2026. Check out our keynote lineup:
Wes McKinney: Future of high-performance computing
Christine Zhang: Data storytelling at NY Times
Emily Riederer: Scaling open-source at Capital One
Sara Altman & Simon Couch: AI-driven workflows
Register: pos.it/conf
New post: Zut, Git ! Porting {saperlipopette} to the terminal with Claude
masalmon.eu/2026/02/27/z...
😂🙏
Délai de soumission aux Rencontres R étendu jusqu'au vendredi 6 mars ! Profitez-en !
💩 Awesome post by @heltweg.org: www.heltweg.org/posts/who-wr...
💩 Slide in my useR! 2024 keynote: user-maelle.netlify.app#/12 (recording linked from masalmon.eu/talks/2024-0...)
🥰 and your post is such a great one!!
February @ropensci.org news! 🗞️ #RStats
💙 Strong Engagement in the Champions Program Call
🤔 Policy Updates for use of Generative AI Tools
🚀 Dashboard for your own R-Universe
📆 Coworking
💓 New packages {Athlytics}, {orgmetrics}
🗒️ Blog posts!
🧰 Package Dev Corner
ropensci.org/blog/2026/02...
Rereading this excellent post by @vickiboykis.com, "I want to see the claw", in a world where "OpenClaw" exists 😭
vickiboykis.com/2025/10/20/i-want-to-see-the-claw/
Why do I have to pretend that I'm going to print something in order to save it as a PDF. Why do I have to engage in a little ruse.
Nature is healing
#rstats 4.5.3 "Reassured Reassurer" scheduled for March 11. Full schedule on developer.r-project.org (or the svn if you're impatient.) This should be the wrap-up release for the 4.5 series.
Table showing whether 3 Git commands can change files only, history only, both files and history. git restore can only change files, git revert can change files only (--no-commit) and both files and history (default), git reset can change the history only (--mixed, --soft) and both files and history (--hard)
Preparing a talk featuring Git restore, and got thinking about reset/revert/restore jvns.ca/blog/2023/11...
Quarto c'est transversal à R, on peut s'y lancer sans R et ne mettre que peu ou pas du tout de code dans ses documents/présentations/livres/etc :-)
Il faut bien commencer quelque part !
Aux Rencontres R il y aura un tutoriel de Typst avec Quarto (pour faire des PDFs avec Quarto, sans LaTeX) rr2026.sciencesconf.org/resource/pag...
In case you were curious what I was sewing the other night when I shared a photo of my serger at 1am 😂 HOW CUTE IS THIS?!
Where are my fellow #rstats #databs sewists & makers?? 🧵🪡🧶
(I loved my @posit.co sweatshirt, but WOO DANG do crew neck collars bother me when they touch my neck 🥹)
Photo of a pair of handmade drop earrings with ggplot2 hex stickers
A recent #RStats craft project of mine 🎨
ggseg 2.0: One atlas object for both 2D and 3D brain visualization.
In v1, you loaded `dk` for 2D and `dk_3d` for 3D.
That's gone. One call — dk() — feeds both renderers.
ggsegverse.github.io/news/ecosyst...
#rstats #neuroscience #datavisualization
Celle-ci ça semble être du Quarto : github.com/kantiles/rr_... (quarto.org/docs/present...)
