Re: [Rd] R-2.15.1 CMD CHECK output: "Call Sequence"

From: Mark Cowley <m.cowley_at_garvan.org.au>
Date: Sun, 21 Oct 2012 21:17:02 +1100

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 ˇ.__T__sort:mjcbase˘, ˇsort˘ are not exported by ˇnamespace:mjcbase˘ Error: package/namespace load failed for ˇmetaGSEA˘ 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 ˇ.__T__sort:mjcbase˘, ˇsort˘ are not exported by ˇnamespace:mjcbase˘ Error: package/namespace load failed for ˇmjcbase˘ Error: package/namespace load failed for ˇmetaGSEA˘ 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 <ripley_at_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 ˇmetaGSEA˘
> 
> 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 ˇmetaGSEA˘ 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 ˇmetaGSEA˘
>> 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 ˇCreating R
>> packages˘ of the ˇWriting R Extensions˘ manual.
>> * checking S3 generic/method consistency ... WARNING
>> Error: package/namespace load failed for ˇmetaGSEA˘
>> 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 ˇGeneric functions and methods˘ of the ˇWriting R
>> Extensions˘ manual.
>> * checking replacement functions ... WARNING
>> Error: package/namespace load failed for ˇmetaGSEA˘
>> 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 ˇvalue˘.
>> * checking foreign function calls ... WARNING
>> Error: package/namespace load failed for ˇmetaGSEA˘
>> 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_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


	[[alternative HTML version deleted]]


______________________________________________

R-devel_at_r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel Received on Sun 21 Oct 2012 - 10:20:16 GMT

This quarter's messages: by month, or sorted: [ by date ] [ by thread ] [ by subject ] [ by author ]

All messages

Archive maintained by Robert King, hosted by the discipline of statistics at the University of Newcastle, Australia.
Archive generated by hypermail 2.2.0, at Mon 22 Oct 2012 - 01:20:51 GMT.

Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-devel. Please read the posting guide before posting to the list.

list of date sections of archive