Public | Automated Build

Last pushed: a year ago
Short Description
Jupyter Notebook server for (data) science. Single user and insecure, Python3, Alpine-based.
Full Description

Docker files

Standard Jupyter port 8888 is exposed, map it to your local port.
You can plug your own persistent notebooks as volume to /notebooks mount.
Example start of the minimal container::

docker run -d \
    --name jupyter \
    -p 8888:8888 \
    -v ${PWD}:/notebooks \

Extending image at run-time
The entrypoint to image is jupyter-notebook (not really configurable
for now) and the server is run under root user.
Together with Terminal available in Jupyter server, this gives you
possibility to run any shell command from inside the container, including
adding system packages with apk add ...
or Python packages with python3 -m pip install ....


.. image::
:target: alpine-jupyter-minimal-py3_dockerhub_

.. image::
:target: alpine-jupyter-minimal-py3_dockerhub_

.. image::
:target: alpine-jupyter-minimal-py3_dockerhub_

.. _alpine-jupyter-minimal-py3_dockerhub:

Basic insecure single user Jupyter Notebook server deployment, Python3-based.
Apart from Python stdlib following Python packages are installed:

  • notebook
  • ipywidgets
  • requests

This makes this image already useful for teaching/showcasing
and some development around Web APIs.


.. image::
:target: alpine-jupyter-sci-py3_dockerhub_

.. image::
:target: alpine-jupyter-sci-py3_dockerhub_

.. image::
:target: alpine-jupyter-sci-py3_dockerhub_

.. _alpine-jupyter-sci-py3_dockerhub:

In addition to the minimal install, next scientific Python packages are added:

  • numpy
  • pandas
  • sympy
  • networkx
  • plotly ( integration, >= 1.9 has offline mode)
  • cufflinks

matplotlib, scipy and scikits will be added later
as they require non-trivial build dependencies


  • better startup/shutdown behavior, handle signals and arguments

    • dedicated startup script?
    • move default settings to jupyter config file?
Docker Pull Command
Source Repository