bips-hb.r-universe.dev

This contains the index for the r-universe repository.
The R-universe project provides an alternative package repository ecosystem including automated package builds and checks outside of the familiar CRAN or BioConductor infrastructure.
Aside from the discoverability bonus (see e.g. the R-universe landing page), the additional build system also allows package installation via install.packages():

You can continue to install e.g. the CRAN version of arf like this:

install.packages("arf")

Or you can install the current development version from GitHub and built by R-universe like this:

install.packages("arf", repos = "https://bips-hb.r-universe.dev")

Note that

1. You do not need to use remotes::instal_github() or similar solutions
2. This will install a binary package that does not require compilation on Windows, macOS, and as of November '23, also WASM binaries.

The WASM binaries are for use with webR to run R in the browser.

Enable it globally

You can also choose to always draw from this repository when using install.packages() by adding its URL to your repos option like this (e.g. in your ~/.Rprofile):

options(repos = c(
  "bips-hb" = "https://bips-hb.r-universe.dev",
  CRAN = "https://cloud.r-project.org/"
))

The next time you try to install one of the packages available at the BIPS R-universe, this repository will now be preferred over the CRAN repository. That means that you can install a binary (on Windows and macOS) version of the latest GitHub version of each package the same way you would normally install CRAN packages:

install.packages("arf")

Please be warned however that if you explicitly want to install the latest CRAN release, you will need to specify this accordingly, i.e. by using

install.packages("arf", repos = "https://cloud.r-project.org/")

...which is why it may not be desirable to add the repository to your global repos.

Maintainign this universe

See the r-universe docs for the general setup procedure.

To add new packages, edit packages.json accordingly (make sure it's syntacically valid JSON).
For CRAN packages, this should not be necessary, but for completeness it is likely a good idea to keep track of packages anyway.