On Fri, Mar 8, 2013 at 12:46 PM, Martin Morgan <mtmorgan at="" fhcrc.org=""> wrote: > On 03/08/2013 12:23 PM, Dan Tenenbaum wrote: >> >> On Fri, Mar 8, 2013 at 12:13 PM, Nigel Horne <nigel_horne at="" hotmail.com=""> >> wrote: >>> >>> Dan, >>> Thanks for getting back to me. >>>>> >>>>> Can you send your sessionInfo()? >>>>> >>>>> My guess is your version of R is too old--Biobase 2.18.0 requires R >>>>> 2.15. >>> >>> Ah, OK, this is R2.14. I was wondering if I needed something more >>> recent. Updating here would be a large undertaking so before I start could >>> I get a 'yes you definitely need R2.15'? >>> >> >> There is a version of Biobase that works with R 2.14. >> You should probably get that version. This page links to the package >> pages for that release (which contain version numbers at the bottom): >> >> http://www.bioconductor.org/packages/2.10/BiocViews.html#___Software >> >> But as Martin says, it's really not recommended to install packages in >> this way. >> You want to make sure that all your packages are from the same >> Bioconductor release so they will work with each other. Mixing and >> matching packages from different releases is a sure recipe for >> problems. > > > Piecing things together a bit, an R session > > source("http://bioconductor.org/biocLite.R") > > gets enough information, specific to the version of R you're in, to proceed > to find the packages that are available to your installation of R, and then > to identify the dependencies of the package(s) you're interested in. > > urls = contrib.url(biocinstallRepos()) > avail = available.packages(urls) > need = utils:::getDependencies("Biobase", TRUE, avail) > > (the first argument to getDependencies can be a vector of packages). The > packages can be downloaded to a local directory > > dest = tempdir() > download.packages(need, dest, repos=biocinstallRepos()) > > and then managed as needed. > Other options: http://www.bioconductor.org/about/mirrors/mirror-how-to/ describes how to rsync the entire BioC 2.10 repository to your local machine. Also, maybe your institution already has an internal BioC mirror. NCI/NIH has one but I don't know if it is available internally. You could ask the maintainer of that mirror whose contact information is listed here: http://www.bioconductor.org/about/mirrors/ Dan > Martin > > >> >> Dan >> >> >> >> >>>> Also the recommended route to installing packages, without worrying >>>> about >>>> dependencies, is just >>>> >>>> source("http://bioconductor.org/biocLite.R") >>>> biocLite("Biobase") >>> >>> Unfortunately that won't work in our set-up because we run a CM system, >>> so software is pushed out from a central server. >>> Regards, >>> -Nigel >>> >>> -- >>> Nigel Horne (NIH, Bethesda, MD, USA) >>> http://www.ncbi.nlm.nih.gov/ >>> >>> [[alternative HTML version deleted]] >>> >>> _______________________________________________ >>> Bioconductor mailing list >>> Bioconductor at r-project.org >>> https://stat.ethz.ch/mailman/listinfo/bioconductor >>> Search the archives: >>> http://news.gmane.org/gmane.science.biology.informatics.conductor >> >> >> _______________________________________________ >> Bioconductor mailing list >> Bioconductor at r-project.org >> https://stat.ethz.ch/mailman/listinfo/bioconductor >> Search the archives: >> http://news.gmane.org/gmane.science.biology.informatics.conductor >> > > > -- > Computational Biology / Fred Hutchinson Cancer Research Center > 1100 Fairview Ave. N. > PO Box 19024 Seattle, WA 98109 > > Location: Arnold Building M1 B861 > Phone: (206) 667-2793

Martin/Dan, > > But as Martin says, it's really not recommended to install packages in > > this way. > > You want to make sure that all your packages are from the same > > Bioconductor release so they will work with each other. Mixing and > > matching packages from different releases is a sure recipe for > > problems. > > Piecing things together a bit, an R session > > source("http://bioconductor.org/biocLite.R") > > gets enough information, specific to the version of R you're in, to proceed to > find the packages that are available to your installation of R, and then to > identify the dependencies of the package(s) you're interested in. > > urls = contrib.url(biocinstallRepos()) > avail = available.packages(urls) > need = utils:::getDependencies("Biobase", TRUE, avail) > > (the first argument to getDependencies can be a vector of packages). The > packages can be downloaded to a local directory > > dest = tempdir() > download.packages(need, dest, repos=biocinstallRepos()) > > and then managed as needed. I appreciate what you're saying, however that doesn't map into the model that we have where we use a CM system to push things out from a central server: no software is installed directly on the development or production machines, so like it or not I can't go down that route. Regards, -Nigel [[alternative HTML version deleted]]