This docker stack provides everything in Jupyter's datascience-notebook with some additional tools students will use in a Data Management for Analytics course (GWU ISTM 6212).
Please visit docker-stacks for detailed instructions.
These additional tools have been installed. See the Dockerfile for details.
- csvkit
- ipython-sql
- GNU less
- GNU nano
- nbgrader
- GNU Parallel
- PostgreSQL 9.6 client and library
- Psycopg2
Note that Docker won't allow a process like the postgresql daemon
to continue running, so the user account jovyan
is given sudo
rights. Within a notebook, or in a shell, jovyan
can restart
postgresql with sudo. Because of this, a build argument for the
document image allows for a password to be set at build time, on
the commandline, for the case where this environment might be
deployed on a public server. It's not real security, but it's
better than nothing.
To set the password in a local build:
% docker build --build-arg passwd=MYPASSWORD -t dchud/datamanagement-notebook .
If you don't specify passwd
as a --build-arg
, the password will
be empty.