Play with the SPT3G software suit on any operating system within an isolated Docker container. This repository was created to enable users to prototype and run spt3g code to a limited extent on local machines especially when access to the internet may be limited. This container is built on ubuntu:20.04 with python:3.8.10 and comes with:

• Full SPT3G software environment
• Jupyter and IPython
• Local copy of spt3g_software documentation

Install the appropriate version of Docker for your operating system : https://docs.docker.com/engine/install

We can then build the docker image using the provieded Dockerfile:

git clone [email protected]:amanchokshi/spt3g-docker.git
cd spt3g-docker

# Copy or clone the spt3g_software repo here
# Need to be a part of the SPT3G Colab for this
git clone [email protected]:SouthPoleTelescope/spt3g_software.git

# Build the image with any tag [-t]
docker build . -t achokshi/spt3g

Create a directory to mount to the docker container. Once mounted it will be available to your host computer and within the docker container, serving as a location to put data and save outputs to:

mkdir ~/spt3g-docker

Run the container with:

docker run --rm -it \
    -v ~/spt3g-docker:/root/spt3g-docker \
    -p 8888:8888 -p 3141:3141 \
    --hostname spt3g \
    achokshi/spt3g

The --rm flag shuts down and cleans up the running container once you're done, while the -it flag provides an interactive terminal. The -v --volume flag mounts the local ~/spt3g-docker volume to the corresponding /root/spt3g-docker directory within the docker container. The -p or flag publishes the container's port(s) to the host. Port 8888 is dedicated for Jupyter while port 3141 is reserved to host a local copy of spt3g_software documentation. achokshi/spt3g is the tag of the docker image which we pulled in previous steps. If the image runs succesfully you will see:

##############################################

Welcome to the SPT3G Docker Container
SPT3G Environment Variables Set
SPT3G Docs Availabel at http://localhost:3141

##############################################

root@spt3g:~#

Files which you put within ~/spt3g-docker will now be available to you. A local version of the documentation is available at http://localhost:3141 once the above commands have been succesfully executed.

Jupyter notebooks are a common tool for interactive data analysis and can be run from this container with:

jupyter-lab --port=8888 --no-browser --ip=0.0.0.0 --allow-root

Any notebooks or data you create will be saved in your ~/spt-docker directory and be available after the container is shutdown.

You should run tests to see that nothing broke in the build:

root@spt3g:~# cd spt3g_software/
root@spt3g:~/spt3g_software# cd build/
root@spt3g:~/spt3g_software/build# make test
Running tests...
..
..
100% tests passed, 0 tests failed out of 54

Hooray!!