Hello,
We are in the midst of changing our SOP for R on our local cluster. Network
access is only available from the head node but we don't want users downloading
and compiling packages on the head node. Previously, we asked that all package
installations & updates be done by the admins on request but this is
unsustainable. We've decided to internally mirror both CRAN and Bioconductor
and keep a core of the most used packages on R_HOME/library and update those
weekly. Users can also download/install/update other packages from
CRAN/Bioconductor using their own local library. What we are trying to work out
now are the system settings needed so that install.packages (and eventually
biocLite) automatically point to only our internal mirrors without having
specify them in the repos argument. After much googling, I think I've figured
out that we need to set up a ~R/etc/Rprofile.site file to change the default
CRAN repository to our local site using options(repos = c(CRAN = "file:////home!
/mirrors/no_backup/cran"). Is this all that is needed or do we also need to
modify ~R/etc/repositories, changing the URL locations to our local ones and
commenting out the other repositories? Anything else we need to do to get
install.packages("ape") to work seamlessly for the end user? We want to make it
so that only users who really know what they are doing can get around
installing from the local mirrors, even if they are accidentally running on the
head node.
Thanks for any help or suggestions,
Jenny
Jenny Drnevich, Ph.D.
Functional Genomics Bioinformatics Specialist
High Performance Biological Computing Program
and The Roy J. Carver Biotechnology Center
University of Illinois, Urbana-Champaign
NOTE NEW OFFICE LOCATION
2112 IGB
1206 W. Gregory Dr.
Urbana, IL 61801
USA
ph: 217-300-6543
fax: 217-265-5066
e-mail: drnev...@illinois.edu<mailto:drnev...@illinois.edu>
[[alternative HTML version deleted]]
______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.
