Nilearn development process

How to help?

  • You are new to python and you don’t know how to do xy

  • If you discovered a bug, but don’t know how to fix it

  • If you discovered a bug and know how to fix it, but don’t know how to get your code onto github (ie, you have some Python experience but have never have used git/github, or have never written tests)

  • You want to contribute code

    • See below

How do we decide what codes goes in?

Scope of the project

Nilearn strives to develop open and powerful statistical analysis of brain volumes (as produced by multiple modalities: MRI, PET, and others). Its focus is to reach end users of the methods (as opposed to methods developers).

Nilearn targets ease of use, but as Python code. In other words, we will not add graphical user interfaces, but we want our code to be as easy to understand as possible, with easy prototyping and debugging, even for beginners in Python.

We are parsimonious in the way we add features to the project, as it puts on weight. Criteria include:

  • It must be in the scope of the project

  • It must correspond to an established practice (typically as used in scientific publications)

  • It must have a concrete use case that can be demo-ed simply with nilearn: an example, on real data, understandable by end-users.

Part of the decision will also be about weighing the benefits (i.e., new features or ease of use for the users) with the cost (i.e., complexity of the code, runtime of the examples).

In practice:

  • The new feature must be demoed in an example in a way that shows its benefit to new users.

  • Because our infrastructure is limited, running all the examples must lead to downloading a limited amount of data (gigabytes) and execute in a reasonable amount of time (a few hours)

  • The new feature must be thoroughly tested (it should not decrease code coverage)

  • The new feature may not introduce a new dependency

Special cases:

  • A downloader for a new atlas: we are currently being very lenient for this: if the atlas is published and can be used in an example, we will accept the pull request (but see below for specifics).

  • A downloader for a new dataset: the larger the dataset is, the less likely we are to consider including it. Datasets are meant to demo and teach features, rather than be the basis of research.

How to contribute a feature

To contribute a feature, first create an issue, in order to discuss whether the feature can be included or not, and the specifications of this feature. Once agreed on the feature, send us a pull request.

There are specific guidelines about how to write code for the project. They can be found in the contributors guide, below.

Special case: How to contribute a dataset fetcher

The nilearn.datasets package provides functions to download some neuroimaging datasets, such as fetch_haxby or fetch_atlas_harvard_oxford. The goal is not to provide a comprehensive collection of downloaders for the most widely used datasets, and this would be outside the scope of this project. Rather, this package downloads data that is required to showcase nilearn features in the example gallery.

Downloading data takes time and large datasets slow down the build of the example gallery. Moreover, downloads can fail for reasons we do not control, such as a web service that is temporarily unavailable. This is frustrating for users and a major issue for continuous integration (new code cannot be merged unless the examples run successfully on the CI infrastructure). Finally, datasets or the APIs that provide them sometimes change, in which case the downloader needs to be adapted.

As for any contributed feature, before starting working on a new downloader, we recommend opening an issue to discuss whether it is necessary or if existing downloaders could be used instead.

To add a new fetcher, nilearn.datasets.utils provides some helper functions, such as get_dataset_dir to find a directory where the dataset is or will be stored according to the user’s configuration, or _fetch_files to load files from the disk or download them if they are missing.

The new fetcher, as any other function, also needs to be tested (in the relevant submodule of nilearn.datasets.tests). When the tests run, the fetcher does not have access to the network and will not actually download files. This is to avoid spurious failures due to unavailable network or servers, and to avoid slowing down the tests with long downloads. The functions from the standard library and the requests library that nilearn uses to download files are mocked: they are replaced with dummy functions that return fake data.

Exactly what fake data is returned can be configured through the object returned by the request_mocker pytest fixture, defined in nilearn.datasets._testing. The docstrings of this module and the Sender class it contains provide information on how to write a test using this fixture. Existing tests can also serve as examples.

Special case: How to contribute an atlas

We want atlases in nilearn to be internally consistent. Specifically, your atlas object should have three attributes (as with the existing atlases):

  • description (bytes): A text description of the atlas. This should be brief but thorough, describing the source (paper), relevant information related to its construction (modality, dataset, method), and if there are more than one maps, a description of each map.

  • labels (list): a list of string labels corresponding to each atlas label, in the same (numerical) order as the atlas labels

  • maps (list or string): the path to the nifti image, or a list of paths

In addition, the atlas will need to be called by a fetcher. For example, see here.

Finally, as with other features, please provide a test for your atlas. Examples can be found here

Who makes decisions

We strongly aim to be a community oriented project where decisions are made based on consensus according to the criteria described above. Decisions are made public, through discussion on issues and pull requests in Github.

The decisions are made by the core-contributors, ie people with write access to the repository, as listed here

If there are open questions, final decisions are made by the Temporary Benevolent Dictator, currently Gaël Varoquaux.

How to contribute to nilearn

This project is a community effort, and everyone is welcome to contribute.

The project is hosted on https://github.com/nilearn/nilearn

The best way to contribute and to help the project is to start working on known issues. See Good first issues to get started. If the issue is not already assigned to someone, you can request it by commenting “take”. This will automatically assign you the issue.

If an issue does not already exist for a potential contribution, we ask that you first open an issue before sending a Pull Requests.

Opening an issue

Nilearn uses issues for tracking bugs, requesting potential features, and holding project discussions.

Core developers can assign labels on issues, such as:

  • Discussion These issues discuss ongoing discussions on the project where community feedback is requested.

  • Enhancement These issues discuss potential enhancements or additions to the project.

  • Bug These issues detail known bugs in the Nilearn code base.

Pull Requests

We welcome pull requests from all community members. We follow the same conventions as scikit-learn. You can find the recommended process to submit code in the scikit-learn guide to contributing code.

Retrieving the latest code

We use Git for version control and GitHub for hosting our main repository. If you are new on GitHub and don’t know how to work with it, please first have a look at this to get the basics.

You can check out the latest sources with the command:

git clone git://github.com/nilearn/nilearn.git

or if you have write privileges:

git clone git@github.com:nilearn/nilearn.git

Installing the latest code

In order to ensure that any code changes are reflected in your installation, navigate to your cloned Nilearn base directory and install using the following command:

pip install -e .

Coding guidelines

Nilearn follows the coding conventions used by scikit-learn. Please read them before you start implementing your changes.

Contributing to the documentation

To build our documentation, we are using sphinx for the main documentation and sphinx-gallery for the example tutorials. If you want to make changes to the example tutorials, please do the following :

  1. First, ensure that you have installed sphinx and sphinx-gallery. You can install the requirements using nilearn/requirements-build-docs.txt.

  2. Fork the Nilearn repository and clone your fork.

  3. Then go to nilearn/examples

  4. Make your changes using reStructuredText files

  5. You can now go to nilearn/doc and build the examples locally:

    make html-strict

    or, if you do not have make install (for instance under Windows):

    python3 -m sphinx -b html -d _build/doctrees . _build/html

  6. Visually review the output in nilearn/doc/_build/html/auto_examples/. If all looks well and there were no errors, commit and push the changes.

  7. You can now open a Pull Request from Nilearn’s Pull Request page.

For more details about the Fork Clone Push worksflow, read here <https://guides.github.com/activities/forking/>_

TIPS : To reduce building time, we suggest you to use the filename_pattern to build just one specific file:

python3 -m sphinx -D sphinx_gallery_conf.filename_pattern=plot_decoding_tutorial.py -b html -d _build/doctrees . _build/html

How to make a release?

This section describes how to make a new release of Nilearn. It is targeted to the specific case of Nilearn although it contains generic steps for packaging and distributing projects. More detailed information can be found on packaging.python.org.

We assume that we are in a clean state where all the Pull Requests (PR) that we wish to include in the new release have been merged. For example, make sure all deprecations that are supposed to be removed with this new version have been addressed. Furthermore, if this new release comes with dependency version bumps (Python, Numpy…), make sure to implement and test these changes beforehand. Ideally, these would have been done before such as to update the code base if necessary. Finally, make sure the documentation can be built correctly.

Prepare the release

Switch to a new branch locally:

git checkout -b REL-x.y.z

First we need to prepare the release by updating the file nilearn/doc/whats_new.rst to make sure all the new features, enhancements, and bug fixes are included in their respective sections. We also need to write a “Highlights” section promoting the most important additions that come with this new release, and add the version tag just above the corresponding title:

.. _vx.y.z:

x.y.z
=====
**Released MONTH YEAR**

HIGHLIGHTS
----------

- Nilearn now includes functionality A
- ...

Next, we need to bump the version number of Nilearn by updating the file nilearn/version.py with the new version number, that is edit the line:

__version__ = x.y.z.dev

to be:

__version__ = x.y.z

We also need to update the website news section by editing the file nilearn/doc/themes/nilearn/layout.html. The news section typically contains links to the last 3 releases that should look like:

<h4> News </h4>
    <ul>
        <li><p><strong>November 2020</strong>:
            <a href="whats_new.html#v0-7-0">Nilearn 0.7.0 released</a>
        </p></li>
        <li><p><strong>February 2020</strong>:
            <a href="whats_new.html#v0-6-2">Nilearn 0.6.2 released</a>
        </p></li>
        <li><p><strong>January 2020</strong>:
            <a href="whats_new.html#v0-6-1">Nilearn 0.6.1 released</a>
        </p></li>
    </ul>

Here, we should remove the last entry and add the new release on top of the list.

In addition, we can have a look at MANIFEST.in to check that all additional files that we want to be included or excluded from the release are indicated. Normally we shouldn’t have to touch this file.

Add these changes and submit a PR:

git add doc/whats_new.rst nilearn/version.py
git commit -m "REL x.y.z"
git push origin REL-x.y.z

Once the PR has been reviewed and merged, pull from main and tag the merge commit:

git checkout main
git pull upstream main
git tag x.y.z
git push upstream --tags

Build the distributions and upload them to Pypi

First of all we should make sure we don’t include files that shouldn’t be present:

git checkout x.y.z

If the workspace contains a dist folder, make sure to clean it:

rm -r dist

In order to build the binary wheel files, we need to install wheel:

pip install wheel

And, in order to upload to Pypi, we will use twine that you can also install with pip:

pip install twine

Build the source and binary distributions:

python setup.py sdist bdist_wheel

This should add two files to the dist subfolder:

  • one for the source distribution that should look like PACKAGENAME-VERSION.tar.gz

  • one for the built distribution that should look like PACKAGENAME-PACKAGEVERSION-PYTHONVERSION-PYTHONCVERSION-PLATFORM.whl

Optionally, we can run some basic checks with twine:

twine check dist/*

We are now ready to upload to Pypi. Note that you will need to have an account on Pypi (you can create one here), and be added to the maintainers of Nilearn. If you satisfy these conditions, you should be able to run:

twine upload dist/*

Once the upload is completed, make sure everything looks good on Pypi. Otherwise you will probably have to fix the issue and start over a new release with the patch number incremented.

At this point, we need to upload the binaries to GitHub and link them to the tag. To do so, go to the Nilearn GitHub page under the “Releases” tab, and edit the x.y.z tag by providing a description, and upload the distributions we just created (you can just drag and drop the files).

Build and deploy the documentation

We now need to update the documentation:

cd doc
make install

This will build the documentation (beware, this is time consuming…) and push it to the GitHub pages repo.

Post-release

At this point, the release has been made. We can now update the file nilearn/version.py and update the version number by increasing the patch number and appending .dev:

__version__ = x.y.(z+1).dev

We can also update the file doc/whats_new.rst by adding a title and the usual New, Enhancements, and Bug Fixes sections for the version currently under developement:

x.y.z+1.dev
=========

NEW
---

Fixes
-----

Enhancements
------------

.. _vx.y.z:

x.y.z
=====
...