Hey #rstats friends, if {quantreg} v5.87 is incompatible with R versions below 4.1¹, why is update.packages() happily trying – and failing without a meaningful error – to install it on R v4.0.5?
I’m so confused.
¹: community.rstudio.com/t/gith…
1
2
1
The package does not have an explicit dependency declared as R (>= 4.1.0) so `update.packages()` does not know. So this may be a subtle bug in the `quantreg` package and we could let its maintainer (/cc @rkoenker) know.
cloud.r-project.org/web/pack…
2
2
Yeah, there was one use of the new \(x) convention that I've reverted to the usual function(x) in my development version, pending resubmitting to CRAN I suppose users will have to upgrade R...
1
1
Thanks, @rkoenker. Upgrading R isn’t an option for me at the moment, but since I’m in no hurry to upgrade {quantreg} I’m happy to wait until CRAN gets the fixed version. :)
And thanks for the package!
1
Not to forget that all of this is of course still open source so it takes about the time it takes to tweet to download, edit, and then install a variant that will work under the slightly-outdated and by April even more outdated R 4.0.*.
1
1
Aaaaah, I hadn’t updated my `/etc/apt/sources.list` to get the latest R version that @eddelbuettel provides. 🤦
Now fixed following the instructions here: cloud.r-project.org/bin/linu…
Thanks, Dirk! :)
1
1
Next: Add the c2d4u.team PPA for ~ 5k CRAN packages pre-made as binary (thanks to @marutterstat)
Extra bonus: @enchufa2 and my @arXiv paper on using `bspm` so #Rstats gives you those binaries via `install.packages()` + `update.packages()`
arxiv.org/abs/2103.08069
Jan 30, 2022 · 3:30 PM UTC
1
2
4