Dear Brian,
thanks for the reply.
Indeed, metaGSEA was the package being CHECK'ed & it turns out that the R CMD
CHECK metaGSEA errors were due to a missing 'Depends: methods' from the
DESCRIPTION of a dependent package.
I went back and CHECK'ed under R 2.14.2 and R 2.13.1, and indeed you're right
re the call sequence's having been there for a while, however I did notice that
R 2.15.1's output was somewhat less useful than R 2.14.2's in these 2 sections:
* checking S3 generic/method consistency ... WARNING
* checking Rd \usage sections ... NOTE
which under R 2.14.2 printed the offending functions (see full R CMD CHECK
output for R 2.15, 2.14, 2.13 here: git://gist.github.com/3926486.git)
fixing the DESCRIPTION of the depedent package removed the package/namespace
errors in this package, and all of the Call Sequence dumps. Importantly, fixing
the dependent package also reinstated the usual verbose output under the 2
sections above.
In hindsight, what this specific scenario boiled down to was this section of
the CHECK output:
Error : objects ¡.__T__sort:mjcbase¢, ¡sort¢ are not exported by
¡namespace:mjcbase¢
Error: package/namespace load failed for ¡metaGSEA¢
Execution halted
It looks like this package (or one of its dependent packages) has an
unstated dependence on a standard package. All dependencies must be
declared in DESCRIPTION.
which was not actually that I didn't export the S4 method sort, but that I was
missing the 'Depends: methods' from mjcbase's DESCRIPTION.
It would be *really* helpful if the error produced when checking 'metaGSEA'
looked like this:
Error : objects ¡.__T__sort:mjcbase¢, ¡sort¢ are not exported by
¡namespace:mjcbase¢
Error: package/namespace load failed for ¡mjcbase¢
Error: package/namespace load failed for ¡metaGSEA¢
Execution halted
It looks like this package (or one of its dependent packages) has an
unstated dependence on a standard package. All dependencies must be
declared in DESCRIPTION.
as when developing, I think we can all be guilty of not looking beyond the
spotlight of the current package.
kind regards,
Mark
On 21/10/2012, at 1:50 AM, Prof Brian Ripley <rip...@stats.ox.ac.uk> wrote:
> On Sat, 20 Oct 2012, Mark Cowley wrote:
>
>> Hi guRus,
>
>> i'm running R-2.15.1 and the R CMD CHECK output appears to have changed
>> substantially, in that instead of pointing me to the
>
> Changed since when? This has been the behaviour for years with the error
> quoted.
>
>> offending functions or files, I get lots of 'call sequence' traces as below.
>
> It is an offending _package_, and it tells you which one in
>
>> Error: package/namespace load failed for ¡metaGSEA¢
>
> Package 'metaGSEA' seems not to be installed properly. It also tells you
> that 'library' is involved, so it looks like it is a package load and not a
> namespace load.
>
> You have not told us if ¡metaGSEA¢ is the package under test or a dependent.
> Given that in earlier 'check' output which you excised the loading of the
> package under test was checked, the latter looks more plausible.
>
>> I've been using R CMD CHECK for years & never seen anything like this
>> before. This is on a pretty fresh OSX 10.8 Mountain Lion installation, where
>> I installed R from CRAN via pkg. Is this a problem unique to me?
>
> Maybe not, but unusual. It is a lot harder to get a package installed
> non-functionally these days, but it can happen if it were installed with a
> different version or architecture, or if binary packages are involved.
>
>>
>> cheers,
>> Mark
>>
>> heres' a snippet from testing one of my packages under development:
>>
>> * checking for unstated dependencies in R code ... WARNING
>> Error: package/namespace load failed for ¡metaGSEA¢
>> Call sequence:
>> 2: stop(gettextf("package/namespace load failed for %s", sQuote(package)),
>> call. = FALSE, domain = NA)
>> 1: library(package, lib.loc = lib.loc, character.only = TRUE, verbose =
>> FALSE)
>> Execution halted
>> See the information on DESCRIPTION files in the chapter ¡Creating R
>> packages¢ of the ¡Writing R Extensions¢ manual.
>> * checking S3 generic/method consistency ... WARNING
>> Error: package/namespace load failed for ¡metaGSEA¢
>> Call sequence:
>> 2: stop(gettextf("package/namespace load failed for %s", sQuote(package)),
>> call. = FALSE, domain = NA)
>> 1: library(package, lib.loc = lib.loc, character.only = TRUE, verbose =
>> FALSE)
>> Execution halted
>> See section ¡Generic functions and methods¢ of the ¡Writing R
>> Extensions¢ manual.
>> * checking replacement functions ... WARNING
>> Error: package/namespace load failed for ¡metaGSEA¢
>> Call sequence:
>> 2: stop(gettextf("package/namespace load failed for %s", sQuote(package)),
>> call. = FALSE, domain = NA)
>> 1: library(package, lib.loc = lib.loc, character.only = TRUE, verbose =
>> FALSE)
>> Execution halted
>> The argument of a replacement function which corresponds to the right
>> hand side must be named ¡value¢.
>> * checking foreign function calls ... WARNING
>> Error: package/namespace load failed for ¡metaGSEA¢
>> Call sequence:
>> 2: stop(gettextf("package/namespace load failed for %s", sQuote(package)),
>> call. = FALSE, domain = NA)
>> 1: library(package, lib.loc = lib.loc, character.only = TRUE, verbose =
>> FALSE)
>> Execution halted
>>
>>> sessionInfo()
>> R version 2.15.1 (2012-06-22)
>> Platform: x86_64-apple-darwin9.8.0/x86_64 (64-bit)
>>
>> locale:
>> [1] en_AU.UTF-8/en_AU.UTF-8/en_AU.UTF-8/C/en_AU.UTF-8/en_AU.UTF-8
>>
>> attached base packages:
>> [1] stats graphics grDevices utils datasets methods base
>> ______________________________________________
>> R-devel@r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-devel
>>
>
> --
> Brian D. Ripley, rip...@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
[[alternative HTML version deleted]]
______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
