Differences

This shows you the differences between two versions of the page.

--- software:r:farber [2018-04-26 13:20] – [R libraries and extensions] sraskar
+++ software:r:farber [2021-03-17 14:44] (current) – [matmul.qs file] anita
@@ Line 245: / Line 245: @@
 === Using IT's udbuild environment ===
-IT developed a formalization for installing modules called [[farber:udbuild]]
+IT developed a formalization for installing modules called [[/abstract/farber/install_software|udbuild]]
 which can simplify the installation of modules.  Here is an example ''udbuild''
 script which can be used to install a personal R library.
@@ Line 288: / Line 288: @@
 cuda 6.5 version into ''$WORKDIR/sw/r/add-ons/r3.1.1/testing/default-cuda-6.5''.
+====== R script in batch ======
+==== matmul.R script ====
+Consider the simple R script file to multiply a small 3x3 matrix
+<file R matmul.R>
+# Calculate and print small matrix AA'
+a <- matrix(1:12,3,4);
+a%*%t(a)
+</file>
+Let's test this R script using ''Rscript'' from the command line on a compute node.  Don't forget to set your [[general/userguide/04_compute_environ?&#using-workgroup-and-directories|workgroup]] to define your cluster group or //investing-entity// compute nodes before you use ''qlogin'' to get on a compute node. For example,
+<code bash>
+workgroup -g it_css
+qlogin
+vpkg_require r/3
+Rscript matmul.R
+</code>
+The output to the screen:
+<code>
+     [,1] [,2] [,3]
+[1,]  166  188  210
+[2,]  188  214  240
+[3,]  210  240  270
+</code>
+To return to the head node, type
+<code bash>
+exit
+</code>
+==== matmul.qs file ====
+To run a R script in batch instead of on the command line has nearly the same steps.
+Consider the queue submission script file:
+<file bash matmul.qs>
+#$ -N matmultiply
+# Add vpkg_require commands after this line:
+vpkg_require r/3
+# Syntax: Rscript [options] filename.R [arguments]
+Rscript matmul.R
+</file>
+Now to run the R script simply submit the job from the head node with the
+''qsub'' command.
+<code>
+qsub matmul.qs
+</code>
+You should see a notification that your job was submitted.  Something like this
+<code bash>
+Your job 2283886 ("matmultiply") has been submitted
+</code>
+After the code completes the output of the script will appear in the file
+''matmultiply.o2283886'' because ''-N matmultiply'' defines the name of the job in ''matmul.qs'' and appears in the notification above as ''("matmultiply")'' with ''2283886'' assigned as the job ID. Type
+<code>
+more matmultiply.o2283886
+</code>
+to display the contents of the output file on the screen.  For example,
+<code>
+Adding dependency `x11/RHEL6.1` to your environment
+Adding package `r/3.0.2` to your environment
+     [,1] [,2] [,3]
+[1,]  166  188  210
+[2,]  188  214  240
+[3,]  210  240  270
+</code>
+====== Using R script in batch array job ======
+===== sweep.R file =====
+Consider the simple script to print a fraction from the argument list
+<file R sweep.R>
+args <- commandArgs(trailingOnly = TRUE)
+# print fraction from argument list
+as.numeric(args[1])/as.numeric(args[2])
+</file>
+This is a R script with can be run from the command line on a compute node the commands
+<code bash>
+qlogin
+vpkg_require r/3
+Rscript sweep.R 5 200
+</code>
+The output to the screen:
+<code>
+[1] 0.025
+</code>
+===== sweep.qs file =====
+Consider the queue script file
+<file bash sweep.qs>
+#$ -N sweep
+#$ -t 1-200
+##
+## Parameter sweep array job to run the sweep.R  with
+##    lambda = 0,1,2. ... 199
+##
+# Add vpkg_require commands after this line:
+vpkg_require r/3
+date "+Start %s"
+echo "Host $HOSTNAME"
+let lambda="$SGE_TASK_ID-1"
+let taskCount=200
+# Syntax: Rscript [options] filename.R [arguments]
+Rscript --vanilla sweep.R $lambda $taskCount
+date "+Finish %s"
+</file>
+The ''date'' and ''echo Host'' lines are just a way of keeping track of when and where the jobs are run.
+There will be 200 array jobs all running the same script with different parameters (arguments).  The ''--vanilla'' option
+is used to prevent the multiple jobs from using the same disk space.
+To run this in batch you must submit the job from the head node with the
+''qsub'' command.
+<code>
+qsub sweep.qs
+</code>
+After the code completes the output of the script will appear in the files
+''sweep.o535064.1'' to ''sweep.o535064.200''. The number 535064 is the job ID assigned to your job when submitted, and 1 to 200 is the Task ID (e.g. corresponds to the ''-t 1-200'')
+<code>
+Adding dependency `x11/RHEL6.1` to your environment
+Adding package `r/3.0.2` to your environment
+[1] 0.025
+</code>
+<note tip>
+You will want to do more than just print out one fraction in your script.  The integer parameter can be used for
+a one dimensional parameter sweep, to construct unique input and output file names for each task,
+or as a seed for the R Random Number Generator (RNG).</note>
+==== Writing files from an array job ====
+You are running many jobs in the same directory.  Grid engine handles the standard output by writing to
+separate files with "dot taskid" appended to the jobid.  You need to take care of other file output in your R script.
+<note important>
+You need to make sure no two of your jobs will write to the same file.  Look at your R script to see if you
+are writing files.  Look for the ''**sink**'' command or any graphics writing commands such as ''**pdf**'' or ''**png**''.
+If you are using these R functions, then use a unique file name constructed from the task id.
+</note>
+==== vanilla option ====
+The command-line option ''--vanilla'' implies --no-site-file, --no-init-file and --no-environ.  This way you will not
+be reading or writing to the same files.  If you need initialization command, put them in your R script instead of in
+in the init-file ''.Rprofile''.  If you need some environment variables, export them in your bash script instead of assigning
+them in your environ file ''.Renviron''.