Stata is a software tool for performing statistical analysis of data. The Stata software is licensed and has been provided by the Department of Political Science.
Using Stata on RCC Resources
To use Stata, you must first load the module:
module load stata
The Stata binaries are:
stata- the batch version of STATA
stata-mp- the batch multi-processor version of STATA
xstata- the gui version of STATA (Spear only)
xstata-mp- the gui multi-processor version of STATA (Spear only)
You can run Stata in three different ways:
- Interactive on the HPC Compute Nodes
- Batch jobs on the HPC Compute Nodes
- Interactive on the Spear Cluster (GUI or CLI)
Interactive on the HPC Compute Nodes
To use interactive mode on the HPC Compute Nodes, you will login into one of the compute nodes directly and type the stata commands just as you would on your computer. This is a good way for troubleshooting. However, you must stay logged in while your jobs are running. For long running jobs, we recommend using batch mode (see below).
For interactive use of STATA, use the following command on an HPC login node:
module load stata srun --pty -t30:00 -n1 stata
This command will connect you to a node where STATA is installed.
Batch mode on the HPC
In batch mode, you submit your job to the system and it will run independently, like other HPC jobs. This is the preferred way of submitting large Stata jobs.
The following is an example of a batch job submission script (
job.sh) that will run a Stat job via SLURM with a do-file called sample.do.
#!/bin/bash #SBATCH -p stata_q #SBATCH -J MYSTATAJOB module load stata stata -b do sample.do
If you save this file as
job.sh, you can submit it using this command on your login node:
Interactive on the Spear Cluster
The Spear nodes provide a good way to use GUI version of STATA. However, only short jobs are allowed to run on the Spear cluster.
To use the GUI version, login to Spear and run:
For more information about Stata, refer to the official documentation: http://www.stata.com/support/documentation.