[R] Dependency errors for package pracma

[Uwe Ligges]

On 09.11.2023 06:42, Hans W wrote:
> I tried to update my package {pracma} on CRAN from 2.4.2 (2022-09-21)
> to version 2.4.4 (2023-11-08). This package reverse depends / imports
> / suggests on 350 packages on CRAN and 25 packages on Bioconductor.
> 
> The only changes are small corrections on some help files, a new
> function for stereographic projection, and `gcd` and `Lcm` require
> integer inputs now (these functions are not used in the packages
> below).
> 
> I received a dependency report saying that
>      *** Changes to worse in reverse dependencies ***
>      celltrackR, geostatsp, gmvjoint, hypr, randnet

Most are from a recent Matrix update, hypr might not, so please check 
the above and tell us whether pracma casuses one of the issues (hypr?) 
and whether the maintainer has been informed in advance.

> 
> Example: geostatsp suggests pracma, but uses only the function 'trapz'
> that has not changed for years and years.
> I cannot check this package, as probably other packages (from
> Bioconductor ?) are needed to install it.
> 
> Example: gmvjoint imports pracma and uses 'grad', 'hessian', and
> 'nearest_spd'; these functions have not changed for years.
> On my system, gmvjoint gets checked without ERRORs !
> 
> What should I do? Frankly, I do not have the time to check these
> packages or to test almost 400 packages before uploading to CRAN.

The CRAN policies point you to functionality so that you can simply run 
these checks autiomatically on your machine. Let me read it for you:

"For a package update, please check that any packages depending on this 
one still pass R CMD check: it is especially expected that you will have 
checked your own packages. Reverse dependencies can conveniently be 
checked using tools::check_packages_in_dir(reverse = list()), and 
changes in check status subsequently be analyzed using 
tools::check_packages_in_dir_changes(). A listing of the reverse 
dependencies of the current version can be found on the CRAN web page 
for the package, or be obtained via tools::package_dependencies(reverse 
= TRUE). If possible, check reverse strong dependencies, reverse 
suggests and the recursive strong dependencies of these (by 
tools::package_dependencies(reverse = TRUE, which = "most", recursive = 
"strong")). "



Best,
Uwe Ligges


> Okay, I can leave it as is and wait until it gets thrown off CRAN
> (because of some new syntax checks, e.g.). I will not mind much. Are
> there better alternatives?
> 
> Thanks, Hans Werner
> 
> ______________________________________________
> R-help using r-project.org mailing list -- To UNSUBSCRIBE and more, see
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.