The pkgstats package is also intended to compare the statistics of an R package to equivalent statistics from all R packages that are, or ever were, on CRAN. For these comparisons, the package also includes developer-only functions to generate pkgstats_summary() results.
These results are uploaded with GitHub release version 0.1.6. A GitHub workflow calls the update function every day.
Warning
This workflow is automatically disabled after 60 days of inactivity on the repository.
There are two main ways the data can be manually updated on a local machine.
1. Direct calls of the update functions
The easiest way is to ensure the latest data have been cached (by calling dl_pkgstats_data()). Then call pkgstats_update().
That function will download any updates from CRAN and rOpenSci which are not present in the latest version, and extract all pkgstats from those. Package tarballs are downloaded in to the temporary directory of the current R session, and so are automatically removed afterwards.
If enormous numbers of packages need to be updated, any bugs in package structure – which do happen! – may cause that function to fail. Even though every effort has been made to ensure that it will recover from every kind of error encountered thus far, R packages can break things in unexpected ways. For this reason, the second approach is recommended if at all possible, and it will always enable complete recovery without having to repeatedly download new packages.
2. Update from a local CRAN mirror
Functions to clone local mirrors of both CRAN and rOpenSci are in the ropensci-review-tools/mirrors repository. The rOpenSci mirror is currently (Oct 2026) around 13GB, while CRAN is 160GB including all archived versions of all packages. These mirrors can be created or updated by calling bash mirror-ropensci.sh or bash mirror-cran.sh.
Data can then be updated by passing those local-mirror paths to pkgstats_from_archive().
Data upload
The pkgstats_update() function will automatically upload data to the GitHub release. The pkgstats_from_archive() returns the data without uploading, so you then have to do that manually – either via piggyyback, or through the GitHub website.