It is somewhat problematic to make the link to the vignette fully portable.
The solution with "../doc/..." may not work if the reference manual is read off
the CRAN page of a package. For a CRAN package, an alternative is to link to
the vignette on its CRAN page (if the vignette is already ther
Dear Thomas,
My FORTRAN code to be used with deSolve contains a module dimmod. During
build a file called dimmod.mod is created in the src directory. I can
avoid this by including it in the .RBuildignore file, but when I submit
the package to CRAN, the package is rejected because it keeps crea
Dear All,
I am seeking direction on how to successfully recreate an CRAN check error (see
email below) for the debian gcc platform (see attached log). I'm pretty
confident that there is still an error associated with my code calling if()
with a vector of length 2 or more. However, the code i
You are missing an underscore. This works for me for a package that I
just added an error to:
rhub::check(env_vars = c("_R_CHECK_LENGTH_1_CONDITION_" = "verbose"),
platform = "debian-gcc-devel",
email = "csardi.ga...@gmail.com")
─ Building package
─ Uploading package
[...]
E checking e
Your documentation files do not match the functions and you have some
files R was not expecting. See inline below for more detail
On 19/11/2018 10:53, Oloyede Isiaq wrote:
I am developing a package in R, but I encountered the below problems,
kindly assist.
Updating relrisks documentation
Lo
Thanks Gabor. Yes, this was my faulty syntax. It appears to be transferring
environmental variables fine and the runs as expected on debian-gcc-devel
replicating the output from the CRAN checks on this system, even though the
environmental variables do not print in the log.
Dear Georgi,
Thank you for this.
None of these solutions seem perfect to me. I think that my suggestion is the
most natural, but as you point out it won’t work in all contexts. Perhaps the
safest approach is to give the vignette() command in the text of the help file,
one of your suggestions.
> None of these solutions seem perfect to me. I think that my suggestion is the
> most natural, but as you point out it won’t work in all contexts. Perhaps the
> safest approach is to give the vignette() command in the text of the help
> file, one of your suggestions.
If you do that, and you us
On Tue, Nov 20, 2018 at 4:45 AM Rampal Etienne wrote:
>
> Dear Thomas,
>
> My FORTRAN code to be used with deSolve contains a module dimmod. During
> build a file called dimmod.mod is created in the src directory. I can
> avoid this by including it in the .RBuildignore file, but when I submit
> th
