[R-SIG-Mac] Problems installing packages on Lion

[Prof Brian Ripley]

On 10/03/2012 00:53, R. Michael Weylandt wrote:
> Oh -- perhaps I was ambiguous and I apologize for any confusion that
> caused: I meant to just build the package himself from source so that
> it would match his architecture, hence why gfortran wouldn't be needed
> as it would to build R.
>
> Incidentally, doesn't CRAN usually support one minor version back? It
> seems that it checks for "R-oldrel" (which appears to currently be
> 2.13.2) -- I don't know if the patch releases are ever incompatible,
> though I doubt so since things seem to divide into R x.y folders,
> rather than R x.y.z in the CRAN servers.

Yes, that's pretty much the policy.  But with 2.15.0 alpha, 2.14.x is 
now 'one minor release back', and building for 2.13.x has been switched 
off (it seems to have been for Macs on Feb 10, a little earlier than I 
would have expected -- the Windows last updates were on Feb 28).

> Michael Weylandt
>
> On Fri, Mar 9, 2012 at 7:25 PM, David Winsemius<dwinsemius at comcast.net>  wrote:
>>
>> On Mar 9, 2012, at 6:42 PM, R. Michael Weylandt wrote:
>>
>>> Following on what David said,
>>>
>>> Are you equipped to build the package locally? That is, do you have
>>> gcc and g++ installed?
>>>
>>> I'm almost certain that a local build (or an upgrade to 2.14 or 2.15
>>> in just over two weeks time) will be the easiest way to fix things.
>>
>>
>> I'm not sure why you say that. He seems to have a fairly standard
>> /R.framework/ location for his /Resources/ and the fact that he still
>> posting in HTML suggests he is not likely to have compiled his R from
>> source.  A few years ago I would often solve problems by building from
>> source, but in recent years when there is a binary package available, it
>> generally installs without problems.
>>
>> (I do see that there is a ../bin/../2.13/  node in the CRAN mirror at ku,
>> which is not typical, so perhaps we are both wrong here. Now that I check,
>> both the Berkeley and the CMU CRAN mirrors seem to keep their Mac binaries a
>> lot longer than is required, so KU is just following US-local tradition. The
>> CMU mirror has binaries back to R 2.8.)
>>
>> Is your experience different?
>>
>> --
>> David.
>>>
>>>
>>> Michael
>>>
>>> On Fri, Mar 9, 2012 at 6:30 PM, Colstat<colstat at gmail.com>  wrote:
>>>>
>>>> I am trying to install package MCMCPack.   I am using Lion.
>>>>
>>>> R confirms:
>>>> trying URL '
>>>>
>>>> http://rweb.quant.ku.edu/cran/bin/macosx/leopard/contrib/2.13/MCMCpack_1.2-1.tgz
>>>> '
>>>> Content type 'application/x-gzip' length 8055525 bytes (7.7 Mb)
>>>> opened URL
>>>> ==================================================
>>>> downloaded 7.7 Mb
>>>>
>>>>
>>>> The downloaded packages are in
>>>>
>>>> /var/folders/qf/q83rlh3x2qjcp75c3phtj81c0000gn/T//Rtmp4qzUW6/downloaded_packages
>>>>
>>>>
>>>> seems like its downloaded, great!
>>>>
>>>>> require(MCMCpack)
>>>>
>>>> Loading required package: MCMCpack
>>>> Error in dyn.load(file, DLLpath = DLLpath, ...) :
>>>>   unable to load shared object
>>>>
>>>> '/Library/Frameworks/R.framework/Versions/2.13/Resources/library/MCMCpack/libs/x86_64/MCMCpack.so':
>>>>
>>>>
>>>> dlopen(/Library/Frameworks/R.framework/Versions/2.13/Resources/library/MCMCpack/libs/x86_64/MCMCpack.so,
>>>> 6): Library not loaded:
>>>> @rpath/R.framework/Versions/2.13/Resources/lib/libR.dylib
>>>>   Referenced from:
>>>>
>>>> /Library/Frameworks/R.framework/Versions/2.13/Resources/library/MCMCpack/libs/x86_64/MCMCpack.so
>>>>   Reason: image not found
>>>>
>>>> I have installed packages before, what is up with this one?
>>>>
>>>>        [[alternative HTML version deleted]]
>>>>
>>>> _______________________________________________
>>>> R-SIG-Mac mailing list
>>>> R-SIG-Mac at r-project.org
>>>> https://stat.ethz.ch/mailman/listinfo/r-sig-mac
>>>
>>>
>>> _______________________________________________
>>> R-SIG-Mac mailing list
>>> R-SIG-Mac at r-project.org
>>> https://stat.ethz.ch/mailman/listinfo/r-sig-mac
>>
>>
>> David Winsemius, MD
>> West Hartford, CT
>>
>
> _______________________________________________
> R-SIG-Mac mailing list
> R-SIG-Mac at r-project.org
> https://stat.ethz.ch/mailman/listinfo/r-sig-mac


-- 
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