Public | Automated Build

Last pushed: a year ago
Short Description
Deprecated. IPython notebook server with SciPy stack. See
Full Description

Dockerized Notebook + SciPy Stack

Docker container for the SciPy stack and configured IPython notebook server.


Assuming you have docker installed, run this to start up a notebook server on https://localhost.

docker run -d -p 443:8888 -e "PASSWORD=MakeAPassword" ipython/scipyserver

You'll now be able to access your notebook at https://localhost with password MakeAPassword (please change the environment variable above).

Hacking on the Dockerfile

Clone this repository, make changes then build the container:

docker build -t scipyserver .
docker run -d -p 443:8888 -e "PASSWORD=MakeAPassword" scipyserver

Use your own certificate

This image looks for /key.pem. If it doesn't exist a self signed certificate will be made. If you would like to use your own certificate, concatenate your private and public key along with possible intermediate certificates in a pem file. The order should be (top to bottom): key, certificate, intermediate certificate.


cat hostname.key intermidiate.cert > hostname.pem

Then you would mount this file to the docker container:

docker run -v /path/to/hostname.pem:/key.pem -d -p 443:8888 -e "PASSWORD=pass" ipython/scipyserver

Using HTTP

This docker image by default runs IPython notebook in HTTPS. If you'd like to run this in HTTP,
you can use the USE_HTTP environment variable. Setting it to a non-zero value enables HTTP.


docker run -d -p 80:8888 -e "PASSWORD=MakeAPassword" -e "USE_HTTP=1" ipython/notebook
Docker Pull Command
Source Repository

Comments (8)
2 years ago

ee69352e1c03: Error pulling dependent layers
2015/07/11 17:47:57 Error pulling image (latest) from ipython/scipyserver, HTTP code 400

2 years ago

In case anyone wants to read a nice tutorial on how to deploy scipyserver using kitematic, I found this very useful

2 years ago

Also if you want to install extra packages on Python 2.7 you'll need to use pip2 as the default pip points at a Python 3 pip (see issue.

2 years ago

When running on Windows you have to setup the port forwarding.
I followed this SoF answer:

boot2docker ssh -vnNTL 443:localhost:443

which allowed me to access the notebook on https://localhost.

3 years ago

Should not run as root. Running as root even within a Docker isn't really the safest option. So when starting as service it should run for example as www-data, a given user, a made up user, or pretty much anything that isn't root.

3 years ago


I have built a jupyterhub image which contains R (as well as scipystack and julia). It can be run using:

docker run -p 8000:8000 simonbiggs/jupyter
3 years ago

Does this contain the R kernel?

3 years ago

How can I run IPython using docker and access it through a computer in the same LAN?