Live, laugh, lapply
07.03.2026 15:25
👍 1
🔁 0
💬 0
📌 0
Live, laugh, lapply
Until they decide that giving you admin permissions, API keys, cloud infrastructure etc. is too much for your role
A wild `separate_wider_delim()` appears 🙌
I agree, not to mention English not being one's own L1 (and I'm willing to bet there are even differences among native speakers)
Normally load_all() is enough to load the dev help (very rarely it doesn't refresh). Is the .Rd there? Maybe try reloading the session?
The different time scales are really misleading.
Cycling seems to be most protective at 2 MET·hours per week.
At a light commuting pace (MET=4), this is 30 minutes per week (5 minutes per day). You do not even warm up! And zero benefits at 10 MET·hours per week (2.5 hours per week).
Yes, I had the habit to click on the active icon without using the keyboard
But I like the new position, I'll get used to it!
But then you can't hide the activity bar by clicking on the active tab :(
Honestly I've never had issues using R + RStudio on Windows: this as long as you have RTools installed, and you're not installing non-CRAN packages that require compilation
What's different? The backend (knitr/pandoc) is the same, and you should also be able to change the HTML template in case it lacks metadata. Or maybe also writing a Lua filter to preprocess your Qmd for those extreme cases...
Not that new, but I'd say ggiraph: it's magic! It takes a ggplot2 and adds SVG + Shiny interactivity to the grobs almost for free. If only there were some extension mechanism...
Will it just become `verse`?
Yes officially with the recent {quarto} package:
cran.r-project.org/web/packages...
And unofficially you can use that syntax under knitr. (And the !expr syntax in code chunk options.)
Wow! Yet another awesome tool with Rust-awesomeness ✨ it's SO pleasant to work in a snappy environment!
I have a dream: template strings in base R!
Just explicitly call utils::data(), do not prefix yours
I can move the .venv outside (UV_PROJECT_ENVIRONMENT) but this is not standard and not well recognized by Positron.
uv is generally fine (new releases take a while to get whitelisted — same as Positron, ark, ...). But most other packages are dead: dbt, airflow, flask, csvkit, pip, sometimes Python. (Plus Onedrive doesn't support symlinks, the .venv gets actually filled by the library.)
Still, Python ships lots of executables, and Windows Defender is not happy. So uv is still unusable since I can't add an exception for each project. And the maintainer refuses to use a centralized library: {renv} does it better!
I'm increasingly proud to have written and published things before LLMs existed.
4. New argument labs(dictionary) to label based on variable name rather than based on aesthetic (@teunbrand, #5178)
AKA: more awesomeness just like scales::label_dictionary()!
github.com/tidyverse/gg...
3. New layer(layout) argument to interact with facets (@teunbrand, #3062)
AKA: repeat a layer in every facet without messing with data
github.com/tidyverse/gg...
2. New stat: stat_manual() for arbitrary computations (@teunbrand, #3501)
github.com/tidyverse/gg...
1. coord_*(expand) can now take a logical vector to control expansion at any side of the panel (top, right, bottom, left)
github.com/tidyverse/gg...
My top 5 ggplot2 4.0.0 updates (well hidden in the NEWS) and their usage: #rstats (there's so much to unpack!) 🙌
Luckily it's getting less and less necessary in R, empty arguments (trailing commas) are often supported within the tidyverse 🍻
And also in DuckDB SQL FYI
Without leaving the tidyverse, just extract the numbers with readr::parse_number(), then use order() on the string:
Treatment <- c("Control", "Asprin 10mg/day", "Asprin 50mg/day", "Asprin 100mg/day", "Acetomycin 100mg/day" "Acetomycin 1000mg/day")
Treatment[order(readr::parse_number(Treatment))]
ggplot2 and now plumber2! What a year! 🙌
It is like saying that I know how to fix a faucet because I can call a plumber
