Enabling Grid execution for R
Utilizing GridR and Bosco, you can submit remote processing from within your R environment, whether it is RStudio or the R command line to clusters on your campus or on national infrastructure such as the OSG or XSEDE.
GridR and Bosco can expand the realm of possibilities for your R processing. By moving R processing to remote clusters, your R workflows can run hundreds or thousands of simultaneous calculations, all managed from your laptop.
Using GridR and Bosco, you write the code on your laptop, and let the cluster do the heavy lifting.
# Initialize
library(GridR)
grid.init(service="bosco.direct", localTmpDir="tmp")
# boring function (replace with real work)
a <- function(s) { return (2*s) }
# Send the function to the remote cluster, wait for the result
grid.apply("x", a, c(1:10), batch=c(1))
With this code, you wrote the function and the inputs, and sent the function a to the cluster to be executed 10 times. After the calculations have completed, the output will be stored in the variable x. It was that easy to submit calculations to a remote cluster!
See more examples on the GridR Wiki
You can quickly install and try out Bosco + GridR using the quick install.
Support is provided by the @osg-bosco team. Modifications to GridR where done by Derek Weitzel (@djw8605).
Support for GridR is provided on the Bosco email list, bosco-discuss@opensciencegrid.org