ePSproc + JupyterLab builds with Docker

Dockerfiles for ePSproc, based on Jupyter Docker Stacks images, are in `/docker`` <https://github.com/phockett/ePSproc/tree/dev/docker>`_. For related tools and more Docker builds, see also the Open Photoionization Docker Stacks repo.

Quick build container only

Basic container + ePSproc install.

docker build -t jupyterlab_epsproc .
docker run -p 8888:8888 jupyterlab_epsproc

where the port mapping is host:container.

Basic container + ePSproc + PEMtk install.

Includes PEMtk install, with libmsym build for symmetrized harmonics support.

docker build -t jupyterlab_epsproc_pemtk -f Dockerfile-PEMtk .
docker run -p 8888:8888 jupyterlab_epsproc_pemtk

where the port mapping is host:container.

Other options

Terminal in container

Use exec -it <container> bash to attach to a running container, or run -it <container> bash to spin one up, and connect to the terminal.

E.g. for named container as above: docker exec -it jupyterlab_epsproc bash

Build with Compose (includes some extra options)

Compose + local install (for development)

See docker-compose.local.yml for options, including port and storage mapping.

The default case pulls source image, installs dependencies, and mounts local folders for source code and notebooks - see the compose file to set the mount paths.

docker-compose -f docker-compose.local.yml build
docker-compose -f docker-compose.local.yml up

Once up, run installlocal.sh in the container to install the local code in editable mode.

A one-liner for this (assuming a github directory mounted with local code): docker exec jupyterlab-ePSproc-dev /home/jovyan/github/ePSproc/docker/localinstall.sh, or run via a terminal in the container, e.g. docker exec -it jupyterlab-ePSproc-dev bash to connect.

Note the script defaults to ~/github for the install, pass a different path if required, e.g. docker exec jupyterlab-ePSproc-dev /home/jovyan/github/ePSproc/docker/localinstall.sh /path/to/repos.

Compose + ePSproc install

See docker-compose.ePSproc.yml for options, including port and storage mapping.

docker-compose -f docker-compose.ePSproc.yml build

NOTE: this compose version is currently not working.

Notes

For more details & options, see:

For use with full JupyterHub deployment, see https://github.com/phockett/jupyterhub-docker

TODO

Further testing.
Different install types. (Here using pip install git+git://github.com/phockett/ePSproc@dev).
Deploy to Docker Cloud/Jupyter Stacks repos.