[Rd] Bitmap device available for R CMD check?

Friedrich Leisch friedrich.leisch at stat.uni-muenchen.de
Thu Jun 1 09:52:57 CEST 2006

>>>>> On Wed, 31 May 2006 11:11:08 -0400,
>>>>> Gabor Grothendieck (GG) wrote:

  > Others may have better ideas but if you have a problem getting
  > your vignette to pass R CMD check then you could process it
  > yourself using Sweave to create a .tex file and then have a
  > mypkg.Rnw file that just contains the SweaveOpt, %Vignette
  > statements and an \input statement to input the .tex file that you
  > generated yourself.  That way you can generate it in a known
  > environment.

Please do not give advice on the R lists on how to cheat the R check
process (unless there is really good reason for it), it's not like we
implement the tests just for the fun of bothering developers. Simply
compare installing and automatically updating a couple of hundred R
packages on all major computing platforms to installing the same
number (OK, there aren't nearly as many) S toolboxes from Statlib in
Splus and you may get an idea why enforced standards may not be so bad
after all ...

We have already discussed several times on this list that

a) "vignette" is a reserved name for documents where

   - the user can both read a version in PDF format and get a handle
   on the code in electronic form

   - R CMD check can test all R code contained in the document, such
   that the user can rely on the code being up to date

b) we currently have only infrastructure for doing a) for documents in
   Sweave format

c) you can put documents in other formats into inst/doc, but these may
   not call themselves vignettes.

d) I have already offered to provide a better indexing infrastructure
   for documents from c) for R 2.4, but that is a summer project.

Most of the above is explained in "Writing R Extensions".


Prof. Dr. Friedrich Leisch 

Institut für Statistik                          Tel: (+49 89) 2180 3165
Ludwig-Maximilians-Universität                  Fax: (+49 89) 2180 5308
Ludwigstraße 33
D-80539 München                 http://www.stat.uni-muenchen.de/~leisch

More information about the R-devel mailing list