Calendar

Semester 1

Sprint 0

Reading assignments for Sprint 0

1. Read Code of Conduct.
2. Read How to choose a project.
3. Read about smart goals.
4. Read Figure checklist and make a perfect figure in your slide.
5. Read Constituents of Decisions. and 3 additional consituents for decisions under uncertainty.
6. Read 10 Simple Rules for Designing Learning Machines.
7. Read 10 Simple Rules to Developing a New Method.
8. Read contribution guidelines for the repository you are considering contributing to, for example, here is a link to sklearn’s contribution guidelines. All the approved repositories have detailed contribution guidelines.

Final Project for Sprint 0

Each team will submit a two-page research proposal (in a Google doc) describing their plan for the year, including

1. a neurobiological question to investigate for the year,
2. a computational approach to solving it,
3. in a separate page, for each student, a list of links to each of the github issues in one of the approved repositories that they will resolve as their final project for the fall semester.

During the first month, if you do not identify an existing issue that you’d like to address, you can get approval to generate a new issue and address it. If you are proceeding, make sure to follow the detailed issue guidelines for the appropriate repository.

Once the TA’s review all the proposed issues to address, you claim them in the appropriate repository.

Sprint 1

Two weeks before the assigned final data, make your pull request (PR) that addressed the issue that you claimed. You will PR into a NeuroDataDesign forked version of the appropriate repository. The TAs will approve your PR, assuming it follows the PR Guidelines established. The PR can be as simple as updating documentation, fixing a bug, or as complex as contributing a previously published algorithm (but it cannot be contributing a new algorithm).

No later than the date and time of the final (as listed by the registrar), each team will append to their google document the following lists of links (one per teammate):

1. The merged PR.
2. The tutorial included in your PR.

In addition, each team will make a final 20 minute presentation, where a single computer will be used to run each of the PR’ed Jupyter notebooks.

Semester 2

In semester 2, each individual can elect to do either a pair of code sprints or data sprints

Code Sprint 1

The deliverable for each team is to append to their google document the following lists of links (one per teammate):

1. The merged PR into the main repo (not the neurodatadesign fork).
2. The issue that you will be addressing in Sprint 3.
3. The two Jupyter notebooks described in 10 Simple Rules to Developing a New Method, PR’ed into the NeuroDataDesign fork of the appropriate repository.

Code Sprint 2

The deliverable for each team is to append to their google document the following lists of links (one per teammate):

1. The merged pull request (PR) that addressed the issue that you claimed. You will PR into a NeuroDataDesign forked version of the appropriate repository. The TAs will approve your PR, assuming it follows the PR Guidelines established. This PR cannot be as simple as updating documentation or fixing a bug; rather it must be developing a new algorithm.
2. The tutorial included in your PR.
3. If your algorithm is new (and therefore cannot be contributed directly to an existing repository), provide a link to PyPi showing you deposited it in the appropriate package management system.

Data Sprint 1

• A 4-page conference report with publication quality figures and 10+ references.
• A notebook to reproduce each of the quantitative results from the report.

Data Sprint 2

• An 8-page conference report with publication quality figures and 10+ references.
• A notebook to reproduce each of the quantitative results from the report.
• A 20-minute presentation presenting the motivation, challenge, action, and result of your work.
• A live demo during your presentation.
• A link to your data derivatives deposited in an open access repository.