Skip to content

[Bioc-devel] error loading Biostrings

12 messages · Robert Scharpf, Vincent Carey, Martin Morgan +3 more

Original
Robert Scharpf
# 
Hello,

I'm having trouble loading the Biostrings package (the Biostrings.so shared library) after installation of the most recent R-devel (Oct 22).  There is no error on the BioC build report, but BioC is building packages from the Oct 5 R-devel. 

Thanks,

Rob

biocLite("Biostrings", type="source")
[...]

*** installing help indices
** building package indices
** installing vignettes
   'Biostrings2Classes.Rnw' 
   'MultipleAlignments.Rnw' 
   'PairwiseAlignments.Rnw' 
   'matchprobes.Rnw' 
** testing if installed package can be loaded
Error in dyn.load(file, DLLpath = DLLpath, ...) : 
  unable to load shared object '/home/bst/student/rscharpf/Library/R/2.16-bioc-devel/library/Biostrings/libs/Biostrings.so':
  /home/bst/student/rscharpf/Library/R/2.16-bioc-devel/library/Biostrings/libs/Biostrings.so: undefined symbol: rcont2
Error: loading failed

Session info:

R Under development (unstable) (2012-10-22 r60997)
Platform: x86_64-unknown-linux-gnu (64-bit)

locale:
 [1] LC_CTYPE=en_US.iso885915       LC_NUMERIC=C                  
 [3] LC_TIME=en_US.iso885915        LC_COLLATE=en_US.iso885915    
 [5] LC_MONETARY=en_US.iso885915    LC_MESSAGES=en_US.iso885915   
 [7] LC_PAPER=C                     LC_NAME=C                     
 [9] LC_ADDRESS=C                   LC_TELEPHONE=C                
[11] LC_MEASUREMENT=en_US.iso885915 LC_IDENTIFICATION=C           

attached base packages:
[1] stats     graphics  grDevices datasets  utils     methods   base     

other attached packages:
[1] BiocInstaller_1.9.4

loaded via a namespace (and not attached):
[1] tools_2.16.0
Martin Morgan
#
On 10/23/2012 07:54 AM, Vincent Carey wrote:
would seem to be this

r60988 | ripley | 2012-10-21 13:57:56 -0700 (Sun, 21 Oct 2012) | 1 line
Changed paths:
    M /trunk/doc/NEWS.Rd
    M /trunk/src/appl/Makefile.in
    M /trunk/src/appl/Makefile.win
    D /trunk/src/appl/rcont.c
    M /trunk/src/include/R_ext/Applic.h

remove deprecated entry point rcont2

but I don't know why it would still be in the revision Vince mentions. Also a 
little more below...
it would have; probably you have a library that was shared between R-2.15 and 
R-devel, or you installed from svn -- biocLite (and install.packages, I believe) 
would have found the repository .../2.16/pkg.tgz

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Kasper Daniel Hansen
    

    

      
      #
    

  


  

      
I was using

R Under development (unstable) (2012-10-22 r60997) -- "Unsuffered Consequences"
Copyright (C) 2012 The R Foundation for Statistical Computing
ISBN 3-900051-07-0
Platform: x86_64-unknown-linux-gnu (64-bit)

Kasper

        
On Tue, Oct 23, 2012 at 11:16 AM, Martin Morgan <mtmorgan at fhcrc.org> wrote:

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

  
  


  


      

    

    

      
      

        


  

        

      Steve Piccolo
    

    

      
      #
    

  


  

      
I'm getting the same error on a fresh install of R-devel.

            

      
      
      
    

        
R Under development (unstable) (2012-10-23 r61007)
Platform: x86_64-unknown-linux-gnu (64-bit)

locale:
 [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C
 [3] LC_TIME=en_US.UTF-8        LC_COLLATE=en_US.UTF-8
 [5] LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8
 [7] LC_PAPER=C                 LC_NAME=C
 [9] LC_ADDRESS=C               LC_TELEPHONE=C
[11] LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base

other attached packages:
[1] BiocInstaller_1.9.4

loaded via a namespace (and not attached):
[1] tools_2.16.0




On 10/23/2011 Tue, Oct 23, 2011 9:18 AM, "Kasper Daniel Hansen"

        
<kasperdanielhansen at gmail.com> wrote:

        

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

  
  


  


      

    

    

      
      

        
      

    

    

      
      

        


  

        

      Hervé Pagès
    

    

      
      #
    

  


  

      
Thanks all. I'll take a look at this (need to update my R-devel first).

H.

        
On 10/24/2012 08:26 AM, Vincent Carey wrote:

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

        

  
    

      
      
    

  


  


      

    

    

      
      

        


  

        

      Hervé Pagès
    

    

      
      #
    

  


  

      
OK I can reproduce this now. The rcont2 C function used to be in the R
C API but was recently moved to the stats package, so it's not
accessible anymore. For it to be accessible stats would need to
register it with REGISTER_CCALLABLE so the C code in other packages
(like Biostrings) would be able to put stats in their LinkingTo field
and then call rcont2.

But the C code in Biostrings trying to call rcont2 is the code behind
an R function (g.test(), non exported) that does a G-test.
This g.test() function is originally from Peter Hurd. See:

   https://stat.ethz.ch/pipermail/r-sig-ecology/2008-July/000275.html

It looks very much like stats::chisq.test() and its implementation
is clearly based on the implementation of stats::chisq.test() which
is also based on rcont2.

g.test() should not be implemented in Biostrings to start with so my
question is: does anybody know if there is already such a function
in the stats package or in a CRAN package? Sounds like it belongs to
stats but I cannot find it.

I don't want to copy the rcont2 C code from stats to Biostrings.
If there is no g.test() in stats or on CRAN, then I'd like to drop
support for G-test in dinucleotideFrequencyTest() (will go thru
deprecation first).

Thanks,
H.

        
On 10/24/2012 08:33 AM, Herv? Pag?s wrote:

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

      
      
    

            

      
      
      
    

        

  
    

      
      
    

  


  


      

    

    
1 day later

    

      
      

        


  

        

      Hervé Pagès
    

    

      
      #
    

  


  

      
This is fixed in Biostrings 2.27.5.  H.

        
On 10/24/2012 09:36 PM, Herv? Pag?s wrote: