Skip to content


Switch branches/tags

Latest commit


Git stats


Failed to load latest commit information.
Latest commit message
Commit time


Experimental docker images for SfePy (Simple finite elements in Python) project. All images are available from the Docker hub Sfepy organization repositories.

Image sfepy-notebook

Docker image bootstrapped with official installation of Miniconda3 and SfePy package that is ready to use.

This image is intended as basic CLI to SfePy functionality or for testing Jupyter notebook/JupyterLab interface.


Docker Run Parameters

You can run this image interactively (/bin/bash) using the following command:

docker run --rm -it -p 8888:8888 sfepy/sfepy-notebook

You can also map any local folder (e.g. including <problem_description_file>) to docker image path (/home/sfepy/data) and run directly selected script:

docker run --rm -it -v $(pwd):/home/sfepy/data sfepy/sfepy-notebook sfepy_run simple <problem_description_file>

Alternatively, you can start a Jupyter Notebook/JupyterLab server and interact with SfePy via your browser:

docker run --rm -it -p 8888:8888 sfepy/sfepy-notebook

and from docker image prompt:

$ jupyter notebook


$ juppyter lab

You can then view the Jupyter notebook/JupyterLab by opening link in your browser (see more info on terminal output).

For further information see official SfePy and Docker instructions.

Useful Information
  • Docker container is based on debian:buster-slim distribution.
  • Use image tag to select specific SfePy release.
  • Inside a running image you can install additional packages or change settings with sudo command (no password required). These changes ARE NOT persistent across multiple docker run commands (see Docker volumes for more info).
  • conda installation prefix is /opt/conda.
  • SfePy sources can be found at /opt/conda/lib/python3.7/site-packages/sfepy.


  • The containers are still experimental and in the early development stage. Please use them with caution and forgiveness,

  • the containers are not size optimized (~2GB) and the pull may take some time,

  • only a limited set of optional SfePy components (e.g. no MPI, IGA) is included,

  • no graphical functions for displaying results visualizations are supported from shell prompt inside the docker container, but you still can do something:

    • run computations in "headless" mode:

      $ sfepy-run postproc data.vtk -o data.png -n
    • embed interactive mayavi 3D plotting into Jupyter notebook:

      from mayavi import mlab

Image sfepy-x11vnc-desktop

This is a Docker image for Ubuntu desktop with X11 and VNC with enhancements on security and features:

  • Accessible via standard web browser or VNC client.
  • Connection is protected by a unique random password for each session.
  • Supports dynamic resizing of the desktop and 24-bit true color.
  • Supports Ubuntu LTS release 20.04 with very fast launching.
  • Automatically shares the current work directory from host to the Docker image.
  • Automatically opens the default browser to show the running desktop.
  • Official Miniconda3 installation including Python 3.7+ and SfePy.


To run the Docker image, first download the script and save it to the working directory where you will store your codes and data.

After downloading the script, you can start the Docker image using the command:


This will download and run the Docker image and then launch your default web browser to show the desktop environment. The work directory by default will be mapped to the /home/sfepy/shared directory on your host.

For additional command-line options, use the command:

python -h
Docker Run Parameters

Please note the "circular definition" here: you need at least a basic Python installation to run the Docker image with the (advanced) Python installation. (Sorry for that ;). Of course, you can run Docker from the command line, but that is a really looong parameters line...

Useful Information
  • All volumes mapped into this image ARE persistent across multiple docker run commands (see Docker volumes for more info).
  • This image does not contain Jupyter notebook/JupyterLab components by default, everything else is the same as for the sfepy-notebook image.

Find Us


Experimental Docker images with sfepy.







No releases published


No packages published