Importing Functions

[jonthegeek]

I'd love to see advice on when to import (vs use_standalone?) vs write your own, and any related advice around them. For example, Clean Code recommends always wrapping imported functions to make it easier to deal with changes. Would you recommend that in R? Only in some situations, maybe?

[olivroy]

On that, I would be curious about adding a dependency vs using a re-export

If importing dplyr in a package but only need tibble::tibble(), is it better to use dplyr::tibble(), same for tidyselect. I haven't found something on that in r-pkgs.org

I know these are "free" dependencies, but cran may complain if there are too many deps.