Ramiro Magno

@ramiro_magno
47 Followers
139 Following
346 Posts
scientist and software engineer
Affiliationhttps://pattern.institute
githubhttps://www.github.com/ramiromagno
githubhttps://www.github.com/patterninstitute
Ramiro MagnoDec 30
#rstats and #emacs fediverse: anyone using emacs as an IDE for R/Python data science? I'm considering giving it a try as a replacement for RStudio/Positron/VSCode. A quick search led me to doom emacs and ess package... I do understand that the investment is significant initially but it seems it can pay off in the long run. So I'm wondering if others have gone down this path and how you deal with graphics, inspecting environments, literate programming (I'm thinking Quarto), etc..
Ramiro MagnoSep 11, 2025
@rstats #rstats: what are you using to parse INI files?
Ramiro MagnoJun 22, 2025
#rstats: Has anyone ever created a bash script to sanitize file permissions on an R project?
Ramiro MagnoJun 22, 2025
#rstats: Just noticed ggplot2 online docs support AI prompting! AI, AI everywhere! :D
Ramiro MagnoApr 22, 2025
US friends: A banner with the text "This repository is under review for potential modification in compliance with Administration directives." is all over NIH websites... what?
Show threadRamiro MagnoApr 3, 2025

@yjunechoe @gaborcsardi (continue) I was under the impression I could not go beyond the three components with CRAN but that might have been some influence from reading tidyverse guidelines in the past: https://r-pkgs.org/lifecycle.html#sec-lifecycle-version-number-tidyverse.

Do you guys actually know of a CRAN package using calendar versioning?

21  Lifecycle – R Packages (2e)

Learn how to create a package, the fundamental unit of shareable, reusable, and reproducible R code.

Show threadRamiro MagnoApr 3, 2025

@yjunechoe Thanks for the pointer! I don't have control over the release dates (upstream), although in my case that is actually quite predictable, like every quarter, so it is not too often, and should be fine with CRAN.

Now that I was looking into it more closely, it seems that you're not even forced to limit yourself to x.y.z on CRAN, and calendar versions seem to be possible too: https://forum.posit.co/t/cran-package-version-guidelines/156330 (thanks @gaborcsardi ) ;)

CRAN package version guidelines

From what I understand, CRAN packages are ideally supposed to follow semantic versioning (X.Y.Z) similar to R itself. However, other version schemes exist. For example, RStudio uses calendar-based versioning (YYYY.MM.patch). A long time ago, I tried to submit a package to CRAN and there was a problem with the version formatting (2022.01 did not work, but 22.1 was okay). Now I cannot seem to find these rules. Are they explicitly stated somewhere? And more specifically, which variants of calendar ...

Posit Community
Ramiro MagnoApr 3, 2025

#rstats hivemind: would it be too funky to define a package version major.minor.patch.dev as YYYY.MM.DD.VERSION, i.e. map major to year, minor to month, patch to day, and leave the dev component for the actual version..? I'm thinking of a data package whose upstream data releases are versioned based on the date... anyone ever tried such heretic approach? Would CRAN maintainers be okay with this?! ;)

asking for a friend.

Ramiro MagnoMar 31, 2025
#rstats: Has anyone ever tried to implement hyperreal infinitesimals in R vectors (e.g., with the vctrs package)?
Show threadRamiro MagnoMar 31, 2025
@josi @trevorld @mdsumner @klmr For example, I guess r-lib packages are not under a meta-package like tidyverse but are effectively developed in a coordinated fashion, so that they do not overlap in scope, use the tidy design patterns, and so on... But conceptually, and effectively, they are being developed in a coordinated manner. Perhaps the lack of any formal bindings was what prompted Hadley to start writing https://design.tidyverse.org/.
Tidy design principles