agentelinux/jupyter-graphlab repository overview

An easy way to get started with data science, using Docker. Containers that come preinstalled with ipython notebook and key packages in the scientific stack.

• Include Graphlab Create lib for python.

⁠Getting started

• Run sh build_container.sh python2 (switch python2 for another container to build that one`).
  • You can also pull the container by running docker pull dataquestio/python2-starter. Replace python2 with any other container you want.
• Run docker run -d -p 8888:8888 -v ORIGIN_FOLDER:/home/ds/notebooks dataquestio/python2-starter
  • Replace python2-starter with the container you want.
  • Replace ORIGIN_FOLDER with a folder on your local machine that you want to persist notebooks in.
• Open your browser and start working with IPython notebook.
  • On Linux, the url will be localhost:8888.
  • On Windows/OSX, run docker-machine ip default (replace default with the name of your machine). Then, you'll be able to access IPython notebook at CONTAINER_IP:8888.

⁠Contributions

Contributions are welcome -- please submit a PR if you want to modify the Dockerfiles, or add to the requirements.