Building BioC 2.2 with R 2.7.0
2
0
Entering edit mode
@atro-tossavainen-1104
Last seen 7.1 years ago
GO annaffy simpleaffy GO annaffy simpleaffy • 1.0k views
0
Entering edit mode
@kasper-daniel-hansen-2979
Last seen 3 months ago
United States
0
Entering edit mode
0
Entering edit mode
0
Entering edit mode
@groot-philip-de-1307
Last seen 7.1 years ago
0
Entering edit mode
> I encountered the same problem with genefilter. Genefilter depends on > tkWidgets which in turn depends on access to the x-server (stupid, but > this is true for several Bioconductor packages now; it's rather annoying). > So: use the xhost command to properly arrange your X display properties > and the genefilter problem might be solved. I neglected to mention (as I didn't see it being related at all) that all of my sessions always have X forwarding via ssh enabled. If this was the issue, I would have had the same problem on all platforms, and furthermore, I am able to run other X programs from the Solaris host on my desktop without issues. Whatever it is, it's not this. Now that you mention it, however, I need to point out that raw X11 on the network using xauth (bad) or xhost (really, really bad) is some- thing that, in my ever so humble opinion, should have died out in 1995 or so when Tatu Yl?nen published ssh. I don't know why people still insist on shooting themselves in the foot by running raw X11 on the network without SSH protection, nor do I particularly need to know. -- Atro Tossavainen (Mr.) / The Institute of Biotechnology at Systems Analyst, Techno-Amish & / the University of Helsinki, Finland, +358-9-19158939 UNIX Dinosaur / employs me, but my opinions are my own. < URL : http : / / www . helsinki . fi / %7E atossava / > NO FILE ATTACHMENTS
0
Entering edit mode
Here's the line that generated genefilter.so: (I'm using Sun's Studio 11 compilers and enabled the use of the sunperf library while building R.) CC -G -o genefilter.so half_range_mode.o init.o nd.o pAUC.o rowPAUCs.o rowttests.o ttest.o -R/opt/SUNWspro/lib/v8plus:/opt/SUNWspro/lib:/opt/ SUNWspro/lib/v8plus:/opt/SUNWspro/lib -L/opt/SUNWspro/lib/v8plus -L/opt/SUNWspro/prod/lib/v8plus -L/opt/SUNWspro/lib -L/opt/SUNWspro/prod/lib -L/usr/ccs/lib -lfui -lfai -lfai2 -lfsumai -lfprodai -lfminlai -lfmaxlai -lfminvai -lfmaxvai -lfsu -lsunmath -lmtsk -lm I think all of the various -R, -L and -l parameters should have been reduced into "-library=sunperf" above. It might not be an issue. I removed the existing genefilter installation and tried building again simply with biocLite("genefilter"), which resulted in exactly the same problem. I also tried building the .so by hand: CC -G -o genefilter.so half_range_mode.o init.o nd.o pAUC.o rowPAUCs.o rowttests.o ttest.o -library=sunperf which resulted in exactly the same problem when loading via > library(genefilter) Loading required package: Biobase Loading required package: tools Welcome to Bioconductor Vignettes contain introductory material. To view, type 'openVignette()'. To cite Bioconductor, see 'citation("Biobase")' and for packages 'citation(pkgname)'. Loading required package: survival Loading required package: splines Error in dyn.load(file, DLLpath = DLLpath, ...) : unable to load shared library '/.../R/2.7.0/lib/R/library/genefilter/libs/genefilter.so': ld.so.1: /.../R/2.7.0/lib/R/bin/exec/R: fatal: relocation error: file /.../R/2.7.0/lib/R/library/genefilter/libs/genefilter.so: symbol __1cDstdJexceptionG__vtbl_: referenced symbol not found Error: package/namespace load failed for 'genefilter' -- Atro Tossavainen (Mr.) / The Institute of Biotechnology at Systems Analyst, Techno-Amish & / the University of Helsinki, Finland, +358-9-19158939 UNIX Dinosaur / employs me, but my opinions are my own. < URL : http : / / www . helsinki . fi / %7E atossava / > NO FILE ATTACHMENTS
0
Entering edit mode
Hi Atro, Thanks for all your efforts in trouble shooting. If I understand, the GO issue is now resolved but genefilter is still problematic. Could you try using gcc to compile rather than sunperf? It looks like there are some header file issues that may be complicating the issue and it will take a few days to fix them and test properly (given a bunch of other constraints). best wishes Robert Atro Tossavainen wrote: > Here's the line that generated genefilter.so: > > (I'm using Sun's Studio 11 compilers and enabled the use of the sunperf > library while building R.) > > CC -G -o genefilter.so half_range_mode.o init.o nd.o pAUC.o rowPAUCs.o rowttests.o ttest.o -R/opt/SUNWspro/lib/v8plus:/opt/SUNWspr o/lib:/opt/SUNWspro/lib/v8plus:/opt/SUNWspro/lib -L/opt/SUNWspro/lib/v8plus -L/opt/SUNWspro/prod/lib/v8plus -L/opt/SUNWspro/lib -L/opt/SUNWspro/prod/lib -L/usr/ccs/lib -lfui -lfai -lfai2 -lfsumai -lfprodai -lfminlai -lfmaxlai -lfminvai -lfmaxvai -lfsu -lsunmath -lmtsk -lm > > I think all of the various -R, -L and -l parameters should have been > reduced into "-library=sunperf" above. It might not be an issue. > > I removed the existing genefilter installation and tried building again > simply with biocLite("genefilter"), which resulted in exactly the same > problem. I also tried building the .so by hand: > > CC -G -o genefilter.so half_range_mode.o init.o nd.o pAUC.o rowPAUCs.o rowttests.o ttest.o -library=sunperf > > which resulted in exactly the same problem when loading via > >> library(genefilter) > Loading required package: Biobase > Loading required package: tools > > Welcome to Bioconductor > > Vignettes contain introductory material. To view, type > 'openVignette()'. To cite Bioconductor, see > 'citation("Biobase")' and for packages 'citation(pkgname)'. > > Loading required package: survival > Loading required package: splines > Error in dyn.load(file, DLLpath = DLLpath, ...) : > unable to load shared library '/.../R/2.7.0/lib/R/library/genefilter/libs/genefilter.so': > ld.so.1: /.../R/2.7.0/lib/R/bin/exec/R: fatal: relocation error: file /.../R/2.7.0/lib/R/library/genefilter/libs/genefilter.so: symbol __1cDstdJexceptionG__vtbl_: referenced symbol not found > Error: package/namespace load failed for 'genefilter' > -- Robert Gentleman, PhD Program in Computational Biology Division of Public Health Sciences Fred Hutchinson Cancer Research Center 1100 Fairview Ave. N, M2-B876 PO Box 19024 Seattle, Washington 98109-1024 206-667-7700 rgentlem at fhcrc.org
0
Entering edit mode
Robert, > Thanks for all your efforts in trouble shooting. If I understand, > the GO issue is now resolved but genefilter is still problematic. Yes indeed. > Could you try using gcc to compile rather than sunperf? As the Bioconductor packages are built with whatever was used to compile R itself, this meant having to compile R from scratch with gcc, which I wasn't really looking forward to. However, at the moment, that looks like the only way I could get a working genefilter going on Solaris/SPARC, and I also found I could substitute the gcc-compiled genefilter.so for the Sun Studio generated one in the otherwise Sun Studio generated install. While this is far from practical, it at least allowed me to have a working set of R and Bioconductor for Solaris/SPARC for now. > > ld.so.1: /.../R/2.7.0/lib/R/bin/exec/R: fatal: relocation error: file /.../R/2.7.0/lib/R/library/genefilter/libs/genefilter.so: symbol __1cDstdJexceptionG__vtbl_: referenced symbol not found As far as I can guess, this might mean that g++ was used to compile something somewhere even though I've specified Sun Studio all along. I'd need to rebuild all of R once more and log everything to verify. -- Atro Tossavainen (Mr.) / The Institute of Biotechnology at Systems Analyst, Techno-Amish & / the University of Helsinki, Finland, +358-9-19158939 UNIX Dinosaur / employs me, but my opinions are my own. < URL : http : / / www . helsinki . fi / %7E atossava / > NO FILE ATTACHMENTS
0
Entering edit mode
Hi, I'm seeking help here regarding updating the Rcurl for macOSX to a newer version so it'll accomodate to bioconductor 2.7. The current version of Rcurl 0.8-1 (in bioconductor 2.7 for macOSX) is built for bioconductor 2.6; Is there anyone who could help to put the bioconductor 2.7 compatible version of Rcurl into the database? So we could use packages depending on Rcurl? Any kind help is greatly appreciated! Thanks in advance. yan Atro Tossavainen wrote: >Robert, > > > >> Thanks for all your efforts in trouble shooting. If I understand, >>the GO issue is now resolved but genefilter is still problematic. >> >> > >Yes indeed. > > > >> Could you try using gcc to compile rather than sunperf? >> >> > >As the Bioconductor packages are built with whatever was used to compile >R itself, this meant having to compile R from scratch with gcc, which >I wasn't really looking forward to. However, at the moment, that looks >like the only way I could get a working genefilter going on Solaris/SPARC, >and I also found I could substitute the gcc-compiled genefilter.so for >the Sun Studio generated one in the otherwise Sun Studio generated >install. While this is far from practical, it at least allowed me to >have a working set of R and Bioconductor for Solaris/SPARC for now. > > > >>> ld.so.1: /.../R/2.7.0/lib/R/bin/exec/R: fatal: relocation error: file /.../R/2.7.0/lib/R/library/genefilter/libs/genefilter.so: symbol __1cDstdJexceptionG__vtbl_: referenced symbol not found >>> >>> > >As far as I can guess, this might mean that g++ was used to compile >something somewhere even though I've specified Sun Studio all along. >I'd need to rebuild all of R once more and log everything to verify. > > >
0
Entering edit mode
Hi, > I'm seeking help here regarding updating the Rcurl for macOSX to a > newer version so it'll accomodate to bioconductor 2.7. The current > version of Rcurl 0.8-1 (in bioconductor 2.7 for macOSX) is built for > bioconductor 2.6; Is there anyone who could help to put the > bioconductor 2.7 compatible version of Rcurl into the database? So > we could use packages depending on Rcurl? Any kind help is greatly > appreciated! When this was brought up earlier, I think the consensus was that since this isn't a bioconductor hosted package, you'd better inquire over at R-help. That said, someone also suggested earlier to install it straight from the source via this incantation: > install.packages("RCurl", repos = "http://www.omegahat.org/R") I don't think that worked for me, and I ended up d/ling the source package and installing it manually, by first d/ling it and uncompressing it. You'll get an RCurl folder. At the command line, you can then: $R CMD BUILD RCurl$ R CMD CHECK RCurl_0.9-3.tar.gz $R CMD INSTALL RCurl_0.9-3.tar.gz I'm not sure if this is the best way, but I seem to have a fully functioning RCurl again since the biomaRt package relies on that, and that works now. Btw - you can get RCurl here: http://www.omegahat.org/RCurl/ HTH, -steve ADD REPLY 0 Entering edit mode Steve and Yan, We just uploaded source, Windows binary, and MacOS X Tiger binary packages for RCurl 0.9-3 to the Bioconductor CRAN-like repository http://bioconductor.org/packages/2.2/extra. This repository is make available when you use R 2.7.0 and type source("http://bioconductor.org/biocLite.R") biocLite("RCurl") Let me know if this meets your needs. Cheers, Patrick Steve Lianoglou wrote: > Hi, > >> I'm seeking help here regarding updating the Rcurl for macOSX to a >> newer version so it'll accomodate to bioconductor 2.7. The current >> version of Rcurl 0.8-1 (in bioconductor 2.7 for macOSX) is built for >> bioconductor 2.6; Is there anyone who could help to put the >> bioconductor 2.7 compatible version of Rcurl into the database? So we >> could use packages depending on Rcurl? Any kind help is greatly >> appreciated! > > When this was brought up earlier, I think the consensus was that since > this isn't a bioconductor hosted package, you'd better inquire over at > R-help. > > That said, someone also suggested earlier to install it straight from > the source via this incantation: > > > install.packages("RCurl", repos = "http://www.omegahat.org/R") > > I don't think that worked for me, and I ended up d/ling the source > package and installing it manually, by first d/ling it and > uncompressing it. You'll get an RCurl folder. At the command line, you > can then: > >$ R CMD BUILD RCurl > $R CMD CHECK RCurl_0.9-3.tar.gz >$ R CMD INSTALL RCurl_0.9-3.tar.gz > > I'm not sure if this is the best way, but I seem to have a fully > functioning RCurl again since the biomaRt package relies on that, and > that works now. > > Btw - you can get RCurl here: http://www.omegahat.org/RCurl/ > > HTH, > -steve > > _______________________________________________ > 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
0
Entering edit mode
Patrick and Steve, Thank you for all your kind help. I installed using source. It works fine. yan Patrick Aboyoun wrote: > Steve and Yan, > We just uploaded source, Windows binary, and MacOS X Tiger binary > packages for RCurl 0.9-3 to the Bioconductor CRAN-like repository > http://bioconductor.org/packages/2.2/extra. This repository is make > available when you use R 2.7.0 and type > > source("http://bioconductor.org/biocLite.R") > biocLite("RCurl") > > Let me know if this meets your needs. > > > Cheers, > Patrick > > > > Steve Lianoglou wrote: > >> Hi, >> >>> I'm seeking help here regarding updating the Rcurl for macOSX to a >>> newer version so it'll accomodate to bioconductor 2.7. The current >>> version of Rcurl 0.8-1 (in bioconductor 2.7 for macOSX) is built for >>> bioconductor 2.6; Is there anyone who could help to put the >>> bioconductor 2.7 compatible version of Rcurl into the database? So >>> we could use packages depending on Rcurl? Any kind help is greatly >>> appreciated! >> >> >> When this was brought up earlier, I think the consensus was that >> since this isn't a bioconductor hosted package, you'd better inquire >> over at R-help. >> >> That said, someone also suggested earlier to install it straight from >> the source via this incantation: >> >> > install.packages("RCurl", repos = "http://www.omegahat.org/R") >> >> I don't think that worked for me, and I ended up d/ling the source >> package and installing it manually, by first d/ling it and >> uncompressing it. You'll get an RCurl folder. At the command line, >> you can then: >> >> $R CMD BUILD RCurl >>$ R CMD CHECK RCurl_0.9-3.tar.gz >> \$ R CMD INSTALL RCurl_0.9-3.tar.gz >> >> I'm not sure if this is the best way, but I seem to have a fully >> functioning RCurl again since the biomaRt package relies on that, and >> that works now. >> >> Btw - you can get RCurl here: http://www.omegahat.org/RCurl/ >> >> HTH, >> -steve >> >> _______________________________________________ >> 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 > >