The answers to this question: Determine if R package is available on Linux mention installing R packages from source, specifically in the context of R running on linux. For instance, pages such as this and this give specific instructions for installing packages on various linux systems, and StackOverflow questions such as: How do I install an R package from source? address the issue more generally. Finally, the answer in this SO question mentions that unix users tend to install packages from source.
I've used R on my university's CentOS linux server a decent bit and have always just added new packages using install.packages()
from within R and specifying a CRAN repository. This generally works fine, though I do notice that it often takes far longer than when installing packages on my personal computer.
What would be the reason to:
(a) Download the .tar.gz file and specify its path via the path
argument in install.packages()
and/or
(b) Install the package directly from a command line call, as specified, for instance, here, here, and here
as opposed to just using install.package()
from within R and specifying a given online repository?