I feel like these types of plots are so easy to get wrong, especially the more observations there are, great article!
12.01.2026 21:20 — 👍 2 🔁 0 💬 0 📌 0@simonherrera.bsky.social
Political science student (& research assistant) at Universidad de la República (Uruguay) | Exchange student at Universidad de Salamanca (Spain) | Researcher & data scientist at Focus Asesoramiento
I feel like these types of plots are so easy to get wrong, especially the more observations there are, great article!
12.01.2026 21:20 — 👍 2 🔁 0 💬 0 📌 0now?
04.01.2026 21:47 — 👍 2 🔁 0 💬 0 📌 0#Rstats slides back into the top 10 programming languages. R is far from dead. www.tiobe.com/tiobe-index/
04.01.2026 10:03 — 👍 56 🔁 17 💬 8 📌 3problem areas in current week over the past few months, tracked across survey rounds (panel) #rstats
23.12.2025 23:57 — 👍 0 🔁 0 💬 0 📌 0El son cubano se convierte en Patrimonio Inmaterial de la Unesco. El género que mezcla elementos españoles y africanos es reconocido como “un pilar histórico” en la música de la isla
10.12.2025 19:22 — 👍 54 🔁 12 💬 1 📌 1Amazing work! Just a heads up: add_geom_usage_to_files() has a bug where it returns a character instead of a dataframe (missing return statement). After some debugging, what I did was create a fix that stores the result before the cli message and returns it explicitly (it worked).
01.12.2025 16:56 — 👍 2 🔁 0 💬 1 📌 0hell naw
21.11.2025 19:10 — 👍 1 🔁 0 💬 0 📌 0deal with*
18.11.2025 23:56 — 👍 0 🔁 0 💬 0 📌 0we do qualitative social research surveys (panels) for my job (via WhatsApp) and we've had to deal a couple of people answering using AI, they've been clumsy about it so far and the sample size has been small enough for us to realize and do something about it, but it is indeed an emerging threat
18.11.2025 22:48 — 👍 2 🔁 0 💬 1 📌 1new paper by Sean Westwood:
With current technology, it is impossible to tell whether survey respondents are real or bots. Among other things, makes it easy for bad actors to manipulate outcomes. No good news here for the future of online-based survey research
I'm sure it's a pretty unorthodox way of doing it but I got Gemini CLI to run R code by using {mcptools} (I did this way back)
17.11.2025 21:47 — 👍 2 🔁 0 💬 0 📌 0Democrats LOVE losing elections
17.11.2025 02:31 — 👍 7 🔁 0 💬 0 📌 0Amazing! Do you publish any code for doing these kinds of maps? My city does a census of the trees here that I would love to see plotted this way!
07.11.2025 18:03 — 👍 1 🔁 0 💬 0 📌 02 years ago at the UN Datathon with some of my favorite people #rstats
04.11.2025 22:25 — 👍 0 🔁 0 💬 0 📌 0Government has published one of those quiet but important documents that might get overlooked as it is not 'newsy'. The headline finding is that £1 of public R&D investment generates £8 in net economic benefits for the UK over the long term
www.gov.uk/government/p...
ranking coverage of news stories surveyed people have said they remembered over the past two months in Uruguay #rstats
29.10.2025 23:40 — 👍 1 🔁 1 💬 0 📌 0started working as a research assistant at my university, where I'll be developing a Shiny app to visualize the results of a recent survey about the welfare state! #rstats
29.10.2025 23:34 — 👍 3 🔁 1 💬 0 📌 0what happened to building great software? enshittification is killing me
15.10.2025 13:32 — 👍 0 🔁 0 💬 0 📌 0unsurprising
14.10.2025 23:16 — 👍 0 🔁 0 💬 0 📌 0this is a complete fiction and the sooner that foreign policy people recognize american imperial history is real the better
14.10.2025 14:48 — 👍 24 🔁 7 💬 1 📌 0Self-inflicted wound
12.10.2025 02:51 — 👍 101 🔁 27 💬 5 📌 1ok so this is how i want my culinary experiments to go from now on
12.10.2025 01:39 — 👍 1 🔁 0 💬 0 📌 0for those of us out there with a love-hate relationship with pdfs, this model is very small and very good! www.ibm.com/new/announce...
29.09.2025 01:24 — 👍 0 🔁 0 💬 0 📌 0{gm} - Create music with R: flujoo.github.io/gm/
#RStats bookmark from old twitter
afaik you need rmd or qmd to do that since pdf needs a latex engine, that's what you get for it being an extremely proprietary format
25.09.2025 20:58 — 👍 0 🔁 0 💬 0 📌 0finally, a worthy oponent for Xaringan
25.09.2025 01:39 — 👍 1 🔁 0 💬 0 📌 0that's a really cool idea!
25.09.2025 00:48 — 👍 1 🔁 0 💬 0 📌 0I mostly find this useful since we have a lot of mechanic workflows at my job, and some scripts that are ran once every couple of months need changes in their logic to work with new data, and AI is usually pretty useful and one-shots it. It's like checking the diff but with less steps!
25.09.2025 00:47 — 👍 1 🔁 0 💬 0 📌 0genuinely curious as to what’s supposed to be used instead
24.09.2025 22:33 — 👍 1 🔁 0 💬 1 📌 0my stylistic choice is to have all my comments be in lowercase, but it's more a feature than a bug, and helps me distinguish AI generated code from my own
24.09.2025 21:51 — 👍 1 🔁 0 💬 1 📌 0