The following example can be used as a job script for running Stata jobs:
#!/bin/bash #SBATCH --job-name=my_stata_job #SBATCH --time=0-01:00:00 #SBATCH --partition=regular #SBATCH --ntasks=1 #SBATCH --mem=2gb module add StataSE/17 stata-se -b do my_stata_file.do
Note that the university currently only has a license for Stata/SE, which is limited to a single CPU. Therefore, always use --ntasks=1 in the job script to request a single core; requesting more cores will not improve the performance.
The input file for Stata, called my_stata_file.do in the previous job script, should be a Stata do file that contains all Stata commands. For example, it could look like:
sysuse auto, clear summarize graph twoway (scatter mpg weight) (lfit mpg weight) graph export plot.ps, replace
This script loads the Automobile dataset that is installed with Stata, prints a summary about the dataset, creates a graph and finally exports this graph to the output file plot.ps.