Python is an open-source, general purpose programming language.
Using Python on CARC systems
You can use Python in either interactive or batch modes. In either mode, first load the corresponding software module:
module load python
This loads the default version, currently 3.9.2, and is equivalent to
module load python/3.9.2. If you require a different version, specify the version of Python when loading. For example:
module load python/3.7.6
To see all available versions of Python, enter:
module spider python
The Python modules depend on the
gcc/8.3.0 module, which is loaded by default when logging in. This module needs to be loaded first because Python was built with the GCC 8.3.0 compiler.
If needed, the
gcc module should be loaded before loading a
module purge module load gcc/8.3.0 module load python/3.9.2
Or alternatively enter
module load usc and then load a
Installing a different version of Python
If you require a different version of Python that is not currently installed on CARC systems, please submit a help ticket and we will install it for you.
Many popular Python packages have already been installed and are available to use after loading one of the Python modules. Use the
pip list command to view them. You can install other Python packages that you need in your home or project directories (see the section on installing packages below).
Please note that we do not currently support the use of Jupyter notebooks on CARC systems.
Running Python in interactive mode
After loading the module, to run Python interactively on a login node, simply enter
python and this will start a new Python session. Using Python on a login node should be reserved for installing packages. Conversely, using Python interactively on a compute node is useful for more intensive work like exploring data, testing models, and debugging.
A common mistake for new users of HPC clusters is to run heavy workloads directly on a login node (e.g.,
endeavour.usc.edu). Unless you are only running a small test, please make sure to run your program as a job interactively on a compute node. Processes left running on login nodes may be terminated without warning. For more information on jobs, see our Running Jobs user guide.
To run Python interactively on a compute node, first use Slurm's
salloc command to reserve job resources on a node:
user@discovery1:~$ salloc --time=1:00:00 --ntasks=1 --cpus-per-task=8 --mem=16GB --account=<project_id> salloc: Pending job allocation 24737 salloc: job 24737 queued and waiting for resources salloc: job 24737 has been allocated resources salloc: Granted job allocation 24737 salloc: Waiting for resource configuration salloc: Nodes d05-04 are ready for job
Make sure to change the resource requests (the
--time=1:00:00 --ntasks=1 --cpus-per-task=8 --mem=16GB --account=<project_id> part after your
salloc command) as needed, such as the number of cores and memory required. Also make sure to substitute your project ID, which is of the form
<PI_username>_<id>. You can find your project ID in the CARC User Portal.
Once you are granted the resources and logged in to a compute node, load the modules and then enter
user@d05-04:~$ module load usc python/3.9.2 user@d05-04:~$ python Python 3.9.2 (default, Mar 19 2021, 09:12:17) [GCC 8.3.0] on linux Type "help", "copyright", "credits" or "license" for more information. >>>
Notice that the shell prompt changes from
user@<nodename> to indicate that you are now on a compute node (e.g.,
To run Python scripts from within Python, use the command
exec(open('script.py').read()). Alternatively, to run Python scripts from the shell, use the
python3 script.py command.
To exit the node and relinquish the job resources, enter
exit in Python and then enter
exit in the shell. This will return you to the login node:
>>> exit() user@d05-04:~$ exit exit salloc: Relinquishing job allocation 24737 user@discovery1:~$
Please note that compute nodes do not have access to the internet, so any data downloads or package installations should be completed on the login or transfer nodes, either before the interactive job or concurrently in a separate shell session.
Running Python in batch mode
To use Python in batch mode, there are a few steps to follow:
- Create a Python script
- Create a Slurm job script that runs the Python script
- Submit the job script to the job scheduler using
Your Python script should consist of the sequence of Python commands needed for your analysis. The
python command, available after a Python module has been loaded, runs Python scripts, and it can be used in the shell and in Slurm job scripts.
A Slurm job script is a special type of Bash shell script that the Slurm job scheduler recognizes as a job. For a job running Python, a Slurm job script should look something like the following:
#!/bin/bash #SBATCH --nodes=1 #SBATCH --ntasks=1 #SBATCH --cpus-per-task=8 #SBATCH --mem=16GB #SBATCH --time=1:00:00 #SBATCH --account=<project_id> module purge module load gcc/8.3.0 module load python/3.9.2 python script.py
Each line is described below:
|Command or Slurm argument||Meaning|
|Use Bash to execute this script|
|Syntax that allows Slurm to read your requests (ignored by Bash)|
|Use 1 compute node|
|Run 1 task (e.g., running a Python script)|
|Reserve 8 CPUs for your exclusive use|
|Reserve 16 GB of memory for your exclusive use|
|Reserve resources described for 1 hour|
|Charge compute time to <project_id>. You can find your project ID in the CARC User Portal|
|Clear environment modules|
|Load the |
|Load the |
Make sure to adjust the resources requested based on your needs, but remember that fewer resources requested leads to less queue time for your job. Note that to fully utilize the resources, especially the number of cores, you may need to explicitly change your Python code to do so (see the section on parallel programming below).
You can develop Python scripts and job scripts on your local machine and then transfer them to the cluster, or you can use one of the available text editor modules (e.g.,
micro) to develop them directly on the cluster.
Save the job script as
python.job, for example, and then submit it to the job scheduler with Slurm's
user@discovery1:~$ sbatch python.job Submitted batch job 13587
To check the status of your job, enter
squeue --me. For example:
user@discovery1:~$ squeue --me JOBID PARTITION NAME USER ST TIME NODES NODELIST(REASON) 170552 main python.j user R 1:01 1 d05-04
If there is no job status listed, then this means the job has completed.
The results of the job will be logged and, by default, saved to a file of the form
slurm-<jobid>.out in the same directory where the job script is located. To view the contents of this file, enter
less slurm-<jobid>.out, and then enter
q to exit the viewer.
For more information on job status and running jobs, see the Running Jobs user guide.
Installing Python packages
After loading a Python module (in this case, version 3), to install packages in your home directory, enter:
pip install <package_name> --user
By default, Python will install local (i.e., user) packages in your home directory (
To install Python packages in a library other than the default, you can use the
--target option with
pip. For example, to install a package in a project directory, enter:
pip install <package_name> --target /project/<project_id>/python/pkgs
<project_id> is your project's ID. You can find your project ID in the CARC User Portal.
To load packages from this location, ensure you have appended your
PYTHONPATH environment variable to include this directory:
To automatically set this variable when logging in to the cluster, add this line to your
~/.bashrc. Additionally, add this line to your Slurm job scripts that depend on the packages installed in this location.
You can also create project-specific package environments using virtual environments. To create a virtual environment, navigate to the directory where you want it to be installed, such as your home or project directory, and enter:
python -m venv <env_name>
<env_name> is the name of your environment. This will create an
<env_name> subdirectory in the current directory. To activate the environment, enter:
This will be reflected in your shell prompt:
Now when you install packages, they will automatically be installed in your
<env_name> environment and directory (e.g.,
./<env_name>/lib/python3.7/site-packages). Additionally, add a similar line to your Slurm job scripts that use this Python environment, but make sure to include the absolute path.
To deactivate the environment, enter
Parallel programming with Python
Python uses only one thread by default, but it also supports both implicit and explicit parallel programming to enable full use of multi-core processors and compute nodes. This also includes the use of shared memory on a single node or distributed memory on multiple nodes. On CARC systems, 1 thread = 1 core = 1 logical CPU (requested with Slurm's
Parallelizing your code to use multiple cores or nodes can reduce the runtime of your Python jobs, but the speedup does not necessarily increase in a proportional manner. The speedup depends on the scale and types of computations that are involved. Furthermore, sometimes using a single core is optimal. There is a cost to setting up parallel computation (e.g., modifying code, communications overhead, etc.), and that cost may be greater than the achieved speedup, if any, of the parallelized version of the code. Some experimentation will be needed to optimize your code and resource requests (optimal number of cores and amount of memory). Also keep in mind that your project account will be charged CPU-minutes based on the cores reserved for a job, even if all those cores are not actually used during the job.
Implicit parallelism is based on multi-threading, so that you do not need to explicitly call for parallel computation in your Python code. Multi-threaded Python packages and functions will automatically detect and use the number of threads. As a result, requesting multiple cores in your Slurm jobs with the
--cpus-per-task option will enable implicit parallelism via automatic multi-threading.
Explicit parallelism means explicitly calling for parallel computation in your Python code, either in relatively simple ways or potentially in more complex ways depending on the tasks to be performed. Many Python packages exist for explicit parallelism, designed for different types of tasks that can be parallelized.
The main Python packages for explicit parallelism are summarized in the following table:
|threading||For explicit multi-threading (I/O bound tasks)|
|multiprocessing||For explicit multi-processing (CPU bound tasks)|
|concurrent.futures||For asynchronous evaluations and workflows|
|mpi4py||For interfacing to MPI libraries|
|Cython||For interfacing to C or C++ code|
|h5py||For parallel I/O|
|pyslurm||For launching jobs via Slurm|
|dask||For asynchronous evaluations and workflows|
Please review the linked documentation above for examples and more information about how to use these packages and their functions.
If you have questions about or need help with Python, please submit a help ticket and we will assist you.
CARC Python workshop materials: