[BioC] can't get limma for R 2.12.0 x64?

Thu Nov 11 00:06:56 CET 2010

Hi Jenny,

Yes I did a complete check today and found a few more packages with
the same problem:

   ADaCGH2
   crlmm
   GeneR
   GenomicRanges
   IRanges
   qpgraph

The bi-arch versions of crlmm, IRanges and qpgraph are on their
way to the public repo (will be available thru biocLite() in 15 or
20 min.). For the 3 remaining packages, we'll need to wait that
the build system can actually produce the bi-arch versions (they
had failures during the latest builds).

Thanks for your patience and sorry for the inconvenience.

H.

On 11/10/2010 02:41 PM, Jenny Drnevich wrote:
> Hi Hervé,
>
> Maybe same problem as before, but now for IRanges? I don't know that I
> use this package directly, but I use affycoretools extensively, and it
> won't load properly without IRanges:
>
>  > biocLite("IRanges")
> Using R version 2.12.0, biocinstall version 2.7.4.
> Installing Bioconductor version 2.7 packages:
> [1] "IRanges"
> Please wait...
>
> Warning message:
> In getDependencies(pkgs, dependencies, available, lib) :
> package ‘IRanges’ is not available
>  > library(affycoretools)
> Error in library.dynam(lib, package, package.lib) :
> DLL 'IRanges' not found: maybe not installed for this architecture?
> Error: package/namespace load failed for 'affycoretools'
>  > sessionInfo()
> R version 2.12.0 (2010-10-15)
> Platform: x86_64-pc-mingw32/x64 (64-bit)
>
> locale:
> [1] LC_COLLATE=English_United States.1252 LC_CTYPE=English_United
> States.1252
> [3] LC_MONETARY=English_United States.1252 LC_NUMERIC=C
> [5] LC_TIME=English_United States.1252
>
> attached base packages:
> [1] grid stats graphics grDevices datasets utils methods base
>
> other attached packages:
> [1] limma_3.6.6 lattice_0.19-13 made4_1.24.0 scatterplot3d_0.3-31
> gplots_2.8.0
> [6] caTools_1.10 bitops_1.0-4.1 gdata_2.8.0 gtools_2.6.2 RColorBrewer_1.0-2
> [11] ade4_1.4-16 KEGG.db_2.4.5 GO.db_2.4.5 RSQLite_0.9-3 DBI_0.2-5
> [16] AnnotationDbi_1.12.0 affy_1.28.0 Biobase_2.10.0 RWinEdt_1.8-2
>
> loaded via a namespace (and not attached):
> [1] affyio_1.18.0 annaffy_1.22.0 annotate_1.28.0 biomaRt_2.6.0
> Category_2.16.0
> [6] genefilter_1.32.0 GOstats_2.16.0 graph_1.28.0 GSEABase_1.12.0
> preprocessCore_1.12.0
> [11] RBGL_1.26.0 RCurl_1.4-4.1 splines_2.12.0 survival_2.35-8 tools_2.12.0
> [16] XML_3.2-0.1 xtable_1.5-6
>  >
>
>
> Thanks,
> Jenny
>
>
> At 09:16 AM 11/10/2010, Jenny Drnevich wrote:
>> Got it installed. Thanks Hervé!
>>
>> Jenny
>>
>> At 07:00 PM 11/9/2010, Hervé Pagès wrote:
>>> On 11/09/2010 03:39 PM, Hervé Pagès wrote:
>>>> Hi Jenny, Martha,
>>>>
>>>> Thanks for reporting this problem.
>>>>
>>>> Because of a problem we had recently with our build system, the
>>>> Windows binary of limma that is currently available thru biocLite()
>>>> is single arch only (it has the 32-bit part but is missing the 64-bit
>>>> part). A new binary containing the 2 archs will be available in a few
>>>> minutes. Please reinstall with biocLite().
>>>
>>> The new limma binary (bi-arch) is now available:
>>>
>>> D:\biocbld\bbs-2.7-bioc>R\bin\R
>>>
>>> R version 2.12.0 RC (2010-10-11 r53293)
>>> Copyright (C) 2010 The R Foundation for Statistical Computing
>>> ISBN 3-900051-07-0
>>> Platform: x86_64-pc-mingw32/x64 (64-bit)
>>> [...]
>>> > source("http://bioconductor.org/biocLite.R")
>>> BioC_mirror = http://www.bioconductor.org
>>> Change using chooseBioCmirror().
>>> > biocLite("limma")
>>> Using R version 2.12.0, biocinstall version 2.7.4.
>>> Installing Bioconductor version 2.7 packages:
>>> [1] "limma"
>>> Please wait...
>>>
>>> trying URL
>>> 'http://www.bioconductor.org/packages/2.7/bioc/bin/windows/contrib/2.12/limma_3.6.6.zip'
>>>
>>> Content type 'application/zip' length 1280462 bytes (1.2 Mb)
>>> opened URL
>>> downloaded 1.2 Mb
>>>
>>> package 'limma' successfully unpacked and MD5 sums checked
>>>
>>> The downloaded packages are in
>>> C:\Users\biocbuild2\AppData\Local\Temp\2\RtmpYU9pzb\downloaded_packages
>>> > library(limma)
>>> > sessionInfo()
>>> R version 2.12.0 RC (2010-10-11 r53293)
>>> Platform: x86_64-pc-mingw32/x64 (64-bit)
>>>
>>> locale:
>>> [1] LC_COLLATE=English_United States.1252
>>> [2] LC_CTYPE=English_United States.1252
>>> [3] LC_MONETARY=English_United States.1252
>>> [4] LC_NUMERIC=C
>>> [5] LC_TIME=English_United States.1252
>>>
>>> attached base packages:
>>> [1] stats graphics grDevices utils datasets methods base
>>>
>>> other attached packages:
>>> [1] limma_3.6.6
>>>
>>> loaded via a namespace (and not attached):
>>> [1] tools_2.12.0
>>> >
>>>
>>> Also to answer your question about how the shared library for both
>>> 32- and 64-bit installations is working: bi-arch Windows packages
>>> that contain native code are compiled for the 2 archs and the
>>> resulting DLLs are put under <pkgtopdir>/libs/i386 and
>>> <pkgtopdir>/libs/x64 respectively. In addition those packages
>>> have an Archs field in their DESCRIPTION file that lists the
>>> supported archs, e.g.:
>>>
>>> Archs: i386, x64
>>>
>>> Packages that don't contain native code don't have the Archs field.
>>> Note that this field is not something the developer needs to be
>>> dealing with: it is automatically added when the binary package is
>>> built (hence you won't find any documentation about it in the
>>> "Writing R Extensions" manual).
>>>
>>> Please let me know if you have any question.
>>>
>>> H.
>>>
>>>
>>>>
>>>> Other packages might have the same problem but we'll need to wait
>>>> after the next build cycle is finished in order to do a more systematic
>>>> check.
>>>>
>>>> Cheers,
>>>> H.
>>>>
>>>>
>>>> On 11/09/2010 01:21 PM, mkbehnke at gmail.com wrote:
>>>>> I had the same problem but was too busy to pursue it at the time - I'm
>>>>> interested in the response too!
>>>>> Martha
>>>>>
>>>>> Sent from my Verizon Wireless BlackBerry
>>>>>
>>>>> -----Original Message-----
>>>>> From: Jenny Drnevich<drnevich at illinois.edu>
>>>>> Sender: bioconductor-bounces at stat.math.ethz.ch
>>>>> Date: Tue, 09 Nov 2010 15:15:39
>>>>> To:<bioconductor at stat.math.ethz.ch>
>>>>> Subject: [BioC] can't get limma for R 2.12.0 x64?
>>>>>
>>>>> Hi all,
>>>>>
>>>>> I'm just now getting around to upgrading to R 2.12.0. Last month I
>>>>> got a fancy new laptop, 64-bit Windows 7, and I was able to install R
>>>>> x64 2.11.1 from the separate binary and then use biocLite() just
>>>>> fine. With the new release of R 2.12.0, it seems they changed the way
>>>>> the R x64 is handled (from the Change Log):
>>>>>
>>>>>>
>>>>>> 64-bit NEWS
>>>>>
>>>>>> The 32- and 64-bit distributions have been merged:
>>>>>>
>>>>>> There is a combined installer (called 'R-<version>-win.exe'). When
>>>>>> used under 32-bit Windows this works as before and installs 32-bit
>>>>>> R. When used under 64-bit Windows there are options to install
>>>>>> either or both of 32- and 64-bit R: the default is to install both.
>>>>>>
>>>>>> The default package type is "win.binary" on both 32- and 64-bit
>>>>>> builds: a single repository contains binary packages for both
>>>>>> architectures. This is in place for CRAN, CRANextras and BioC: type
>>>>>> = "win64.binary" can still be used for any repositories which follow
>>>>>> the R 2.11.x convention. With the default filters (see
>>>>>> ?available.packages) packages will only be offered if they are
>>>>>> available for the current architecture.
>>>>>>
>>>>>> The default personal library, e.g. '~/R/win-library/2.12' is now the
>>>>>> same for both 32- and 64-bit R.
>>>>>
>>>>> I'm thinking it's great that I can easily have both 32-bit and 64-bit
>>>>> versions of R. I use the combined installer and install both 32- and
>>>>> 64-bit versions. I plan on primarily using the 64-bit version, so I
>>>>> first go into that and use biocLite(), but I got a warning message
>>>>> that package 'limma' is not available:
>>>>>
>>>>> > source("http://bioconductor.org/biocLite.R")
>>>>> BioC_mirror = http://www.bioconductor.org
>>>>> Change using chooseBioCmirror().
>>>>> > biocLite()
>>>>> Using R version 2.12.0, biocinstall version 2.7.4.
>>>>> Installing Bioconductor version 2.7 packages:
>>>>> [1] "affy" "affydata" "affyPLM" "affyQCReport"
>>>>> [5] "annaffy" "annotate" "Biobase" "biomaRt"
>>>>> [9] "Biostrings" "DynDoc" "gcrma" "genefilter"
>>>>> [13] "geneplotter" "GenomicRanges" "hgu95av2.db" "limma"
>>>>> [17] "marray" "multtest" "vsn" "xtable"
>>>>> Please wait...
>>>>>
>>>>> #skipped a bunch of output
>>>>>
>>>>> Warning message:
>>>>> In getDependencies(pkgs, dependencies, available, lib) :
>>>>> packages 'GenomicRanges', 'limma' are not available
>>>>>
>>>>> > biocLite("limma")
>>>>> Using R version 2.12.0, biocinstall version 2.7.4.
>>>>> Installing Bioconductor version 2.7 packages:
>>>>> [1] "limma"
>>>>> Please wait...
>>>>>
>>>>> Warning message:
>>>>> In getDependencies(pkgs, dependencies, available, lib) :
>>>>> package 'limma' is not available
>>>>>
>>>>> > sessionInfo()
>>>>> R version 2.12.0 (2010-10-15)
>>>>> Platform: x86_64-pc-mingw32/x64 (64-bit)
>>>>>
>>>>> locale:
>>>>> [1] LC_COLLATE=English_United States.1252
>>>>> [2] LC_CTYPE=English_United States.1252
>>>>> [3] LC_MONETARY=English_United States.1252
>>>>> [4] LC_NUMERIC=C
>>>>> [5] LC_TIME=English_United States.1252
>>>>>
>>>>> attached base packages:
>>>>> [1] stats graphics grDevices utils datasets methods base
>>>>>
>>>>> loaded via a namespace (and not attached):
>>>>> [1] tools_2.12.0
>>>>>
>>>>>
>>>>> The current page for limma says the windows binary is 32-& 64-bit:
>>>>> http://bioconductor.org/help/bioc-views/release/bioc/html/limma.html
>>>>>
>>>>> So then I went into the 32-bit R and used biocLite(), and it
>>>>> installed everything just fine, and I can load limma:
>>>>>
>>>>> > library(limma)
>>>>> > sessionInfo()
>>>>> R version 2.12.0 (2010-10-15)
>>>>> Platform: i386-pc-mingw32/i386 (32-bit)
>>>>>
>>>>> locale:
>>>>> [1] LC_COLLATE=English_United States.1252
>>>>> [2] LC_CTYPE=English_United States.1252
>>>>> [3] LC_MONETARY=English_United States.1252
>>>>> [4] LC_NUMERIC=C
>>>>> [5] LC_TIME=English_United States.1252
>>>>>
>>>>> attached base packages:
>>>>> [1] stats graphics grDevices utils datasets methods base
>>>>>
>>>>> other attached packages:
>>>>> [1] limma_3.6.6
>>>>>
>>>>> loaded via a namespace (and not attached):
>>>>> [1] tools_2.12.0
>>>>>
>>>>>
>>>>> If I go back to the x-64 R and try to load limma, I now get this
>>>>> message:
>>>>>
>>>>> > library(limma)
>>>>> Error: package 'limma' is not installed for 'arch=x64'
>>>>>
>>>>>
>>>>> I have no idea how the shared library for both 32- and 64-bit
>>>>> installations is going to work. I looked at both the R and BioC
>>>>> mailing lists, but no one seemed to post this problem yet, so I'm
>>>>> wondering what I'm doing wrong. Shouldn't it "just work", or do I
>>>>> need to get the 64-bit toolchain and compile limma from source for my
>>>>> 64-bit R?
>>>>>
>>>>> Thanks,
>>>>> Jenny
>>>>>
>>>>> [[alternative HTML version deleted]]
>>>>>
>>>>> _______________________________________________
>>>>> Bioconductor mailing list
>>>>> Bioconductor at stat.math.ethz.ch
>>>>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>>>>> Search the archives:
>>>>> http://news.gmane.org/gmane.science.biology.informatics.conductor
>>>>> _______________________________________________
>>>>> Bioconductor mailing list
>>>>> Bioconductor at stat.math.ethz.ch
>>>>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>>>>> Search the archives:
>>>>> http://news.gmane.org/gmane.science.biology.informatics.conductor
>>>
>>>
>>> --
>>> Hervé Pagès
>>>
>>> Program in Computational Biology
>>> Division of Public Health Sciences
>>> Fred Hutchinson Cancer Research Center
>>> 1100 Fairview Ave. N, M2-B876
>>> P.O. Box 19024
>>> Seattle, WA 98109-1024
>>>
>>> E-mail: hpages at fhcrc.org
>>> Phone: (206) 667-5791
>>> Fax: (206) 667-1319
>>
>> _______________________________________________
>> Bioconductor mailing list
>> Bioconductor at stat.math.ethz.ch
>> https://stat.ethz.ch/mailman/listinfo/bioconductor
>> Search the archives:
>> http://news.gmane.org/gmane.science.biology.informatics.conductor
>

-- 
Hervé Pagès

Program in Computational Biology
Division of Public Health Sciences
Fred Hutchinson Cancer Research Center
1100 Fairview Ave. N, M2-B876
P.O. Box 19024
Seattle, WA 98109-1024

E-mail: hpages at fhcrc.org
Phone:  (206) 667-5791
Fax:    (206) 667-1319