Thank you for your interest in contributing! There are a few ways you can contribute to the project:
A great way to contribute to the IPFX is by participating in the community discussion on the the forums. By asking new questions and answering questions from other users you will help growing the forum as a go-to collaborative scientific resource.
Is your question about the IPFX, or about Allen Institute data and tools more generally?
- If you have an IPFX question, first check the documentation (including the gallery of examples and the api reference) to make sure that your question is not already addressed.
- If you can't find an answer in the documentation, please create an issue on Github.
- If you have question about data and other tools, you should check our online help or the Allen Brain Map Community Forum.
- If you can't find what you are looking for with the aforementioned resources, you can submit your question using this form.
Before reporting a bug or requesting a feature, use Github's issue search to see if anyone else has already done so.
If there is no existing issue, create a new one. You should include:
- A brief, descriptive title
- A clear description of the problem
- If you are reporting a bug, your description should contain the following information:
- What you did (preferably the actual code or commands that you ran)
- What happened
- What you were expecting to happen
- How your system is configured (operating system, Python version)
If you are comfortable addressing this issue yourself, take a look at the guide to contributing code below.
Before suggesting a feature or enhancement, please check existing issues as you may find out you don't need to create one. When you create an enhancement suggestion, please include as many details as possible in the issue template.
When contributing a new feature to the IPFX, the maintenance burden is (by default) transferred to the IPFX team. This means that the benefit of the contribution must be weighed against the cost of maintaining the feature.
When suggesting a feature, consider:
- Is the change clearly explained and motivated?
- Would the enhancement be useful for most users?
- Is this a new feature that can stand alone as a third party project?
- How does this change impact existing users?
If you are able to improve the IPFX, send us your pull requests! Contributing code yourself can be a great way to include the features you want in the IPFX.
Navigate to the Github "issues" tab and start looking through issues. The IPFX team uses Github issues to track our internal development, so we recommend filtering to issues with the "good first issue" label or issues with the "help wanted" label. These are issues that we believe are particularly well suited for outside contributions, often because we won't get to them right away. If you decide to start on an issue, leave a comment so that other people know that you are working on it.
Code contributions should be submitted in the form of a pull request. Here are the steps:
- Create up an issue/branch/environment
- Sign Contributor License Agreement (CLA)
- Check if changes are consistent with Coding style
- Commit your code changes
- Write unit tests
- Run unit tests
- Update the "Unreleased" section in the CHANGELOG.md with the change made in this PR
- Make a pull request
- Go through the review process
-
Make sure that there is an issue tracking your work. See above for guidelines on creating effective issues.
-
Create a fork of the IPFX and clone it to your development environment.
-
Make a new branch for your code off of
master
. For consistency and use with visual git plugins, we prefer the following convention for branch naming:GH-<issue-number>/<bugfix/feature>/<short-description>
. For example:GH-712/bugfix/auto-reward-key GH-9999/feature/parallel-behavior-analysis
-
Create an environment and install necessary requirements:
requirements.txt
andrequirements-test.txt
-
Start writing code!
We follow PEP-8 guidelines for new python code. We also follow PEP-484 for type annotations. Before submitting a pull request, run flake8 and mypy linters to check the style of your code. All new code contributions should be compatible with Python 3.6+.
Docstrings for new code should follow the Numpy docstring standard. This allows us to ensure consistency in our auto-generated API documentation.
Commit messages should have a subject line, separated by a blank line and then paragraphs of approximately 72 char lines. For the subject line, shorter is better -- ideally 72 characters maximum. In the body of the commit message, more detail is better than less. See Chris Beams for more guidelines about writing good commit messages.
- Tag the issue number in your subject line. For Github issues, it's helpful to
use the abbreviation ("GH") to separate it from Jira tickets.
GH #1111 - Add commit message guidelines This contains more detailed information about the feature or bugfix. It's written in complete sentences. It has appropriate capitalization and punctuation. It's separated from the subject by a blank line.
- Limit commits to the most granular changes that make sense. Group together small units of work into a single commit when applicable. Think about readability; your commits should tell a story about your changes that everyone can follow.
All code you write should have unit tests, including bugfixes (since the presence of bugs likely indicates a gap in test coverage). We use pytest for running unit tests.
If you write a new file foo.py
, you should place its unit tests in test_foo.py
.
Follow the directory structure of the parent module(s) for your tests so that
they are easy to find. For example, tests for ipfx/dataset/foo.py
should be in ipfx/tests/dataset/test_foo.py
.
Testing guidelines
- Smaller, faster tests are better (and more likely to be run!)
- Tests should be deterministic
- Tests should be hermetic. They should be packed with everything they need and start any fake services they might need.
- Tests should work every time; use dependency injection to mock out flaky or long-running services.
- Make sure your tests pass locally first (
make test
orpython -m pytest <a test file or directory>
) - Update your forked repository and rebase your branch onto the latest
master
branch. - Target the latest release candidate branch for your PR. This branch has the format
rc/x.y.z
. - Use a brief but descriptive title.
- Include
Relates to: #issue_number
and a short description of your changes in the body of the pull request. - Support your changes with additional resources. Having an example notebook or visualizations can be very helpful during the review process.
Once your pull request has been made and your tests are passing, a member of the IPFX team will be assigned to review. Please be patient, as it can take some time to assign team members. Once your pull request has been approved, the IPFX team member will merge your changes into the latest release candidate branch. Your changes will be included in the next release cycle. Releases typically occur every 2-4 weeks.
If in doubt how to do anything, don't hesitate to ask a team member!