R devel - Oct 2007 - a package depending on other packages does not pass checking on windows

Dear developers,

I am writing a package that depends on some other packages. The
dependencies are stated in the `description' file under "Depends".
They
are installed in my private library, which is pointed to by setting
R_LIBS in .Renviron, and are available if R is started normally.
However, when I try to `R CMD check' my package, R complains about the
dependencies being not available. I believe the reason is the option
--vanilla set in $R_HOME/bin/check that prevents R from reading my
.Renviron.
This problem manifests itself with R 2.5.1 under Windows XP but not with
the same version under Linux. (Yes, I know this is not the latest
version, sorry about that :) )

How do I convince `R CMD check' to find the packages installed in the
private library?

Thanks in advance for your help!

WBR

Mstislav Elagin

I have been bitten by this myself. For that reason I have stopped  
using .Renviron.

I believe the behavior is the same under Linux and Windows, so I  
think you might have a different setup on the two machines.

You can fix it by explicitly setting
   R_LIBS
as an environment variable or by putting the following in your .Rprofile
   .libPaths(SOMEDIR)
But of these are accessed by R CMD ...

Kasper

On Oct 29, 2007, at 3:45 PM, Mstislav Elagin wrote:
> Dear developers,
>
> I am writing a package that depends on some other packages. The
> dependencies are stated in the `description' file under
"Depends".
> They
> are installed in my private library, which is pointed to by setting
> R_LIBS in .Renviron, and are available if R is started normally.
> However, when I try to `R CMD check' my package, R complains about the
> dependencies being not available. I believe the reason is the option
> --vanilla set in $R_HOME/bin/check that prevents R from reading my
> .Renviron.
> This problem manifests itself with R 2.5.1 under Windows XP but not  
> with
> the same version under Linux. (Yes, I know this is not the latest
> version, sorry about that :) )
>
> How do I convince `R CMD check' to find the packages installed in the
> private library?
>
> Thanks in advance for your help!
>
> WBR
>
> Mstislav Elagin
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel

>From the 'Writing R Extensions' manual:
   R CMD check and R CMD build run R with
   --vanilla, so none of the user's startup files are read.  If
   you need R_LIBS set (to find packages in a non-standard library)
   you will need to set it in the environment.

so 'the problem' is your not reading the documentation.

On Mon, 29 Oct 2007, Mstislav Elagin wrote:
> Dear developers,
>
> I am writing a package that depends on some other packages. The
> dependencies are stated in the `description' file under
"Depends". They
> are installed in my private library, which is pointed to by setting
> R_LIBS in .Renviron, and are available if R is started normally.
> However, when I try to `R CMD check' my package, R complains about the
> dependencies being not available. I believe the reason is the option
> --vanilla set in $R_HOME/bin/check that prevents R from reading my
> .Renviron.
> This problem manifests itself with R 2.5.1 under Windows XP but not with
> the same version under Linux. (Yes, I know this is not the latest
> version, sorry about that :) )
>
> How do I convince `R CMD check' to find the packages installed in the
> private library?
>
> Thanks in advance for your help!
>
> WBR
>
> Mstislav Elagin
>
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
>
-- 
Brian D. Ripley,                  ripley at stats.ox.ac.uk
Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
University of Oxford,             Tel:  +44 1865 272861 (self)
1 South Parks Road,                     +44 1865 272866 (PA)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595