Skip to content

Access to cran packages

3 messages · Elisabeth Slooten, Jim Lemon, Ivan Krylov

#
I am able to access R packages on my laptop, but not on my desktop computer. Both are Macs, both running OS10.13.6 High Sierra. When I open package installer on the laptop and click"Get list" it provides the usual list of packages. On the desktop computer (Mac Pro Tower, 2012) the following message pops up in the R console window:
"Warning: unable to access index for repository https://cran.stat.auckland.ac.nz/bin/macosx/contrib/4.2:
  cannot open URL 'https://cran.stat.auckland.ac.nz/bin/macosx/contrib/4.2/PACKAGES'

The same happens if I type >install.packages() into the R console window. A list of packages appears on the laptop, but on the desktop computer this happens:
Warning: unable to access index for repository https://cran.stat.auckland.ac.nz/src/contrib:
  cannot open URL 'https://cran.stat.auckland.ac.nz/src/contrib/PACKAGES'
Error in install.packages() : argument "pkgs" is missing, with no default
Error in install.packages(sf) : object 'sf' not found

Both computers are online, and on both of them I can see the list of packages here: https://cran.stat.auckland.ac.nz/bin/macosx/contrib/4.2/PACKAGES

Is there a way to directly download the packages?

Many thanks!

Liz Slooten
Professor Emeritus, University of Otago, New Zealand
#
Hi Liz,
Try:

install.packages("sf")

Jim

On Mon, Jul 18, 2022 at 8:37 PM Elisabeth Slooten
<liz.slooten at otago.ac.nz> wrote:
#
On Sun, 17 Jul 2022 22:35:23 +0000
Elisabeth Slooten <liz.slooten at otago.ac.nz> wrote:

            
There seems to be something wrong with R's ability to download files
from the Internet on one of your machines. Does
download.file('https://cran.stat.auckland.ac.nz/bin/macosx/contrib/4.2/PACKAGES',
'PACKAGES') provide a useful error message? Is the R version the same
on both computers? Any differences in the sessionInfo() output?
Anything suspect in .Rprofile?

Try asking at r-sig-mac at r-project.org.
You can always visit the package web page (e.g.
<https://CRAN.R-project.org/package=sf>) to get the download link, and
since you're on a macOS machine, you can download the binary.
Unfortunately, it's up to you to resolve the dependencies manually.

To do that semi-manually, you can download
https://cran.r-project.org/web/packages/packages.rds, read it using
readRDS and feed it as the db argument to
tools::package_dependencies('sf', db, recursive = TRUE) in order to
know which packages to download.