[Rd] How to list package dependency on a Bioconductor package?
Martin Morgan
mtmorgan at fhcrc.org
Tue May 8 17:56:20 CEST 2012
On 5/8/2012 8:44 AM, Simon Urbanek wrote:
> Steven,
>
> On May 7, 2012, at 1:24 PM, Steven Lund wrote:
>
>> I recently submit a package called QuasiSeq. The package source and
>> Windows binary worked fine, but the MacOS X binary failed. The bottom of
>> the CRAN page for QuasiSeq shows
>>
>> Downloads:
>> Package source: QuasiSeq_1.0-1.tar.gz
>> MacOS X binary: not available, see check log.
>> Windows binary: QuasiSeq-1.0-1.zip
>> Reference manual: QuasiSeq.pdf
>>
>>
>> The log for the MacOS X binary shows
>> - using R version 2.15.0 (2012-03-30)
>> - using platform: i386-apple-darwin9.8.0 (32-bit)
>> - using session charset: ASCII
>> - checking for file 'QuasiSeq/DESCRIPTION' ... OK
>> - this is package 'QuasiSeq' version '1.0-1'
>> - checking package namespace information ... OK
>> - checking package dependencies ... ERROR
>> Package required but not available: 'edgeR'
>>
>> See the information on DESCRIPTION files in the chapter
>> 'Creating R
>> packages' of the 'Writing R Extensions' manual.
>> - elapsed time (check, wall clock): 0:01
>>
>> Is there a simple remedy to the problems caused by the dependency on edgeR,
>> or would you advise that I put a work-around into my source code using
>> 'require' so I can list edgeR under Suggests instead of Depends or Includes?
>>
> Well, if your package doesn't depend on it, it should certainly not be in Depends. But that is a design decision (depending on packages outside CRAN is somewhat tricky because in general you cannot guarantee availability, but it is legal ;)).
>
> If you need any additional non-CRAN dependencies on Mac OS X, they can be installed if there are working self-sufficient binaries, you just need to tell me.
We're talking about a Bioconductor package here; the Mac binaries are
available for R-2-15
http://bioconductor.org/packages/release/bioc/html/edgeR.html in the
usual way; the Windows binaries are apparently discovered by the CRAN
build system, so it sounds like CRAN's Mac builds are mis-configured, or
Bioconductor's binary builds are inadequate?
Martin
>
> Cheers,
> Simon
>
>
>> On Tue, Jan 3, 2012 at 7:42 AM, Uwe Ligges
>> <ligges at statistik.tu-dortmund.de>wrote:
>>
>>>
>>> On 03.01.2012 00:16, Steven Lund wrote:
>>>
>>>> I know others have asked similar questions to the R developers, but I
>>>> could not find the solution to this question. Please forgive me if I
>>>> have missed a crucial point in a previous post.
>>>>
>>>> I would like to submit a package to CRAN that depends on the
>>>> bioconductor package "edgeR". Listing "edgeR" under the Depends or
>>>> Imports lines in the DESCRIPTION file for my package causes an error
>>>> when running the `R CMD check' command on my package's 'tar.gz' file:
>>>>
>>>> * checking package dependencies ... ERROR
>>>> Package required but not available: āedgeRā
>>>>
>>>>
>>>> Is there documentation or an example anywhere of how to handle
>>>> dependencies on Bioconductor packages when creating a package for
>>>> CRAN?
>>>>
>>>
>>> Same as for CRAN packages. The CRAN check farm will automatically install
>>> the dependency from BioConductor if available for the corresponding
>>> platform / R version.
>>>
>>> Uwe Ligges
>>>
>>>
>>>
>>>
>>>
>>>> Thank you!
>>>>
>>>> -Steve
>>>>
>>>> ______________________________**________________
>>>> R-devel at r-project.org mailing list
>>>> https://stat.ethz.ch/mailman/**listinfo/r-devel<https://stat.ethz.ch/mailman/listinfo/r-devel>
>>>>
>> [[alternative HTML version deleted]]
>>
>> ______________________________________________
>> R-devel at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-devel
> ______________________________________________
> R-devel at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
--
Dr. Martin Morgan, PhD
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N.
PO Box 19024 Seattle, WA 98109
More information about the R-devel
mailing list