Scientific Computing with Python
Austin, Texas • July 11-17, 2016

SciPy 2016 Sprints Schedule

Join us for a 2-day open source hackathon!  Sit with the authors of open source packages you have been using and contribute a feature you have a desire for.  Or help push a new and exciting project that will change your life.  Don't know how to contribute to a project?  No problem, we'll teach you at the Sprint tutorial.

We encourage you to fill out the sprints form in order to have your sprint (or sprint idea/request) published on this page!

See the bottom of the page for details on each of the currently planned sprints:

  • Gensim
  • scikit-image
  • scikit-learn
  • scikit-build
  • yt
  • matplotlib
  • conda-forge
  • Nipype

Sprints are free for all, but please register to receive a badge and access to the sessions.

8:00 AM - 9:00 AMBreakfast
Served in the Tejas Room on Level 2
9:00 AM - 9:30 AMSprint Kickoff
Room 204
Sprint leaders will speak briefly about their projects and rooms will be assigned.
9:30 AM - 11:00 AM"How to Sprint" Tutorial
Room 204
A brief tutorial for new sprinters on how to work with open source projects and use git and github to coordinate with other contributors.
11:00 AM - 6:00 PMSprints (Rooms assigned at kickoff)
No location
6:30 PM - 8:30 PMTexas BBQ Dinner
Scholz's Beer Garten, 1607 San Jacinto Blvd (walking distance from AT&T Center)

8:00 AM - 9:00 AMBreakfast
Served in the Tejas Room on Level 2
9:00 AM - 6:00 PMSprints (Rooms assigned at kickoff)
No location

Sprint Session Details

Name of package or theme

Description of the goal(s) of the sprint, tasks planned and why it is important.

Minimum level of Python expertise needed

Is familiarity with package required?


Gensim is a popular Natural Language Processing library using Machine Learning techniques like LDA Topic Modelling and word2vec. We need people wishing to learn about these techniques to come and write tutorials, documentation and tests together with more experienced package developers over these 2 days. There is also a plan to refactor our word2vec and Latent Dirichlet Allocation implementations in order to incorporate slight variations without code duplication. The variations are respectively fastsent and Supervised LDA.




scikit-image is an image processing toolbox for Python. During the sprints, we work on everything: bug fixes, documentation & website, new features, and whatever you fancy. We welcome beginners and experts alike, and will give mentorship as needed. At the end of the day, the goal is to have fun and to learn from one another.




The point of this sprint is to make new contributors familiar with the scikit-learn development process and show how easy it can be to contribute. More experienced contributors welcome, of course.




scikit-build is a package that makes is easier to build scientific packages that are CPython extensions that require compilation of C/C++ or Fortran code. In this sprint, we will apply scikit-build to new projects.




This sprint serves as an opportunity for yt developers to fix bugs, work on new features, and improve yt as a group in person. We also encourage new users or people who are interested in yt to stop by and ask questions about yt. If you would like to get involved with yt development, we will also be able to point you to easy issues to get started on so you can get familiar with the codebase.




Primarily, we'll be work on integrating and refining all of the new style changes in matplotlib to get version 2.0 out the door. New developers are also welcome --we will begin with a "start contributing to matplotlib" tutorial.




We will be sprinting on conda-forge (a community powered packaging solution for the conda package manager). The types of things we can achieve:

* packaging software that you care about

* enhancing or fixing bugs for existing packages

* extending the conda-forge infrastructure to improve administration of packages

All levels of experience welcome - from first time conda users right through to compiler gurus.




Nipype ( is a general purpose dataflow framework, with specific support for medical imaging. It allows running large scale computational workflows on local and clustered environments.

The goal of the sprint will be to split the nipype project into easily installable and reusable components, and to integrate with other python projects (neurovault and neurosynth). By the time the sprint happens, we will label all issues by python/scientific expertise level.

Sprint topics will include:

- establishing framework for real world testing of data (data package management)

- support for docker containers and the common workflow language

- provenance tracking

- restful API services