Open main menu

Changes

Analysing Data on Armis or Great Lakes using Rmd Files

180 bytes added, 17:17, 10 July 2023
Clarified memory and time
=== Creating a Batch Script ===
To submit a job to the server you will need to create a batch script to submit your jobs. The Armis2 and Greatlakes servers use slurm to submit jobs.
Details of how to configure a slurm file can be found [https://arc.umich.edu/armis2/slurm-user-guide here]. Each script can include multiple Rmd files. A sample script is below. Make sure to replace <USER> with your slurm user id and entering a name for the job and your email.
<pre>
#SBATCH --nodes=1
#SBATCH --cpus-per-task=1
#SBATCH --mem-per-cpu=24G2G#SBATCH --time=00:5515:00
#SBATCH --account=<USER>
#SBATCH --partition=standard
echo "Running from $(pwd)"
 
# requires 30 mins and 64G
Rscript -e "rmarkdown::render('<FIRST_SCRIPT>.Rmd')"
Rscript -e "rmarkdown::render('<SECOND_SCRIPT>.Rmd')"
</pre>
Save this as a <JOB_NAME>.slurm file in your working folder on turbo. The <nowiki>mem-per-cpu</nowiki> and <nowiki>time</nowiki> commands should be adjusted to approximate how complex your job is (if it is too little memory or time you will get an update in the email response).
== Transferring Files ==