This is a Jupyter notebook packaged up inside a Docker container which comes preinstalled with a bunch of software needed to do a demo analysis of Planck data. Once the container is running, you can connect to the Jupyter notebook from your local browser.
There are no software dependencies needed to run this except for having Docker installed. If you don't already have it, on Linux you can follow the instructions here, or on Windows/Mac install Docker Toolbox from here. Docker requires a 64bit computer, and for Windows/Mac your processor needs to have hardware virtualization (called VT-x or AMD-v) available and enabled (this is the vast majority of any computer made in the last ~5 years).
With Docker installed, you can pull the image (a ~400Mb download):
docker pull marius311/jcdemo_argentina
To run it:
docker run -d -v $(pwd):/root/shared -p 8888:8888 marius311/jcdemo_argentina
At this point Juptyer is running and you simply need to point your browser to it. On Linux just go the URL 127.0.0.1:8888
. If you are running on Windows/Mac, replace 127.0.0.1
with the IP address reported when you start up the Docker Quickstart terminal.
The folder from where you do the run
command is shared between Docker and your local computer, so changes you make to notebooks will be saved there.
The build the image from the source code provided here and make changes, just check out this git repository, edit the Dockerfile or add/modify any notebook files, and from the top folder run,
docker build -t marius311/jcdemo_argentina .
You will also need to download and place the PICO datafile in the top folder for the build to work, I'll make this file available from a central location as soon as I have a more final version. -Marius