Public | Automated Build

Last pushed: 2 years ago
Short Description
Short description is empty for this repo.
Full Description

ontouchstart/qstk: a lightweight Docker image for QuantSoftware ToolKit

ontouchstart/qstk is a lightweight Docker image for QuantSoftware ToolKit built from
ontouchstart/QuantSoftwareToolkit, a fork from

For your convenience, both vim and emacs are also installed in this image.


Simple CLI

docker run --name qstk --rm -it ontouchstart/qstk bash


Since there is no GUI in docker container, we set matplotlibrc to

backend      : PDF

There are many ways to get the PDF files out of the container. For example: docker cp command or Data volumes.

But there is also an easy way to access the content inside the container via a web browser. Here I will give an example in boot2docker for Mac OS X:

First you need to do a port forwarding from container port 8000 to the localhost port 8000:

boot2docker ssh -L 8000:localhost:8000

Then you will be in a boot2docker shell like:

boot2docker: 1.2.0
             3.16.1-config-file : e75396e - Fri Aug 22 06:45:30 UTC 2014

You can run docker in a separate shell or inside boot2docker shell itself:

docker run -p 8000:8000 --name qstk --rm -it ontouchstart/qstk bash

Once in the docker container, you can run SimpleHTTPServer in the background

cd QuantSoftwareToolkit/
python -m SimpleHTTPServer >& /dev/null &

Open http://localhost:8000 in your browser to see the directory list under QuantSoftwareToolkit.

CLI input + GUI output

There is another way to work efficiently with QSTK in a CLI input + GUI output workflow.

It can be demonstrated in the
GitHub repo of this docker image. The basic ideas are:

  • Put Python scripts (such as,
    which was modified from the original QSTK
    for output destinations) in the place like input directory.

  • Use UNIX pipe to send the script to the docker container running python.

  • Copy the output from docker container via docker cp command.

  • Remove the docker container.

You can see how this works in a sample shell script
The result will be displayed nicely in GitHub as


Docker Pull Command
Source Repository