What do #rstats folks do as a protection against CRAN package removals, in terms of scripts that require a fresh installation each run (like for a script on databricks)? Use a different site for package installs? Install important packages straight from GitHub?
2
2
Maybe some kind of internal company repository?
1
Replying to @kevin_m_kent
a) Nothing ever disappears from CRAN, there is an Archive section b) I have used R and CRAN for decades and I have gotten by very well by not relying on packages that are or were removed c) You can also create local / personal repositories, see my `drat` package

Apr 4, 2023 · 6:03 PM UTC

2
1
Replying to @eddelbuettel
Thank you! This is helpful. I need to figure out how to use one of these options for my current use cases
Replying to @eddelbuettel
How do you assess a package up front to determine what is likely to be maintained well enough to rely on (as to not get removed from CRAN)? Recent Commit activity, number of maintainers?