Contributing¶

Code of Conduct¶

By participating in Nilearn, you agree to abide by the NIPY Code of Conduct

How to get help?¶

If you have issues when using Nilearn, or if you have questions on how to use it, please don’t hesitate to reach out!

There are currently three ways to interact with the Nilearn team: through the neurostars forum, our github issues, and through our weekly drop-in hours, usually every Wednesday from 4pm to 5pm UTC. We post on our X account in advance to let you know if the drop-in hours are happening that week.

If you have a usage question, that is if you need help troubleshooting scripts using Nilearn, we would appreciate it if you either ask it during the drop-in hours or create a topic on neurostars with the “nilearn” tag. Asking questions or reporting issues is always valuable because it will help other users having the same problem. So, please don’t hold onto a burning question!

We ask that you don’t open an issue on GitHub for usage questions. We use our GitHub issue board for bug reports, feature requests, and documentation changes only.

How to help the project?¶

If you are interested in contributing to the Nilearn project, we thank you very much. Note that there are multiple ways to help us, and not all of them require writing code.

Report bugs or discuss enhancement ideas¶

We welcome open discussion around improvements—both to the documentation as well as to the code base—through our GitHub issue board!

If you think you have discovered a bug, please start by searching through the existing issues to make sure it has not already been reported. If the bug has not been reported yet, create an new issue including a minimal runnable example to showcase it (using nilearn.datasets) as well as your OS and Nilearn version.
If you have an idea for a new feature, check if it is in the Scope of the project and feel free to open a new issue to discuss it.
If you think the documentation can be improved, please open a new issue to discuss what you would like to change! This helps to confirm that your proposed improvements don’t overlap with any ongoing work.

Answer questions¶

Another way to help the project is to answer questions on neurostars, or comment on github issues. Some issues are used to gather user opinions on various questions, and any input from the community is valuable to us.

Review Pull Requests¶

Any addition to the Nilearn’s code base has to be reviewed and approved by several people including at least two Core developers. This can put a heavy burden on Core developers when a lot of pull requests are opened at the same time. We welcome help in reviewing pull requests from any community member. We do not expect community members to be experts in all changes included in pull requests, and we encourage you to concentrate on those code changes that you feel comfortable with. As always, more eyes on a code change means that the code is more likely to work in a wide variety of contexts!

Contribute code¶

If you want to contribute code:

For new features, please be sure to create a new issue first, to discuss whether it can be included and its specifications.
To help with known issues, please check good first issues to get started, known bugs, or proposed enhancements.

Please see the How to contribute to nilearn section for more detailed information, including instructions for Setting up your environment and a description of the Contribution Guidelines.

How do we decide what code goes in?¶

The following sections explain the Scope of the project and Who makes decisions, which jointly determine whether potential contributions will be accepted into the project.

Scope of the project¶

Nilearn is an Open-source Python package for visualizing and analyzing human brain MRI data. It provides statistical and machine-learning tools for brain mapping, connectivity estimation and predictive modelling. It brings visualization tools with instructive documentation & open community.

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. To assess new features, our main concern is their usefulness to a number of our users. To make Nilearn high-quality and sustainable we also weigh their benefits (i.e., new features, ease of use) with their cost (i.e., complexity of the code, runtime of the examples).

Exhaustive criteria used in the review process are detailed in the contribution guide below. Be sure to read and follow them so that your code can be accepted quickly.

As a rule of thumb:

To be accepted, new features must be in the scope of the project and correspond to an established practice (typically as used in scientific publications).
It must have a concrete use case, illustrated with a simple example in the Nilearn documentation to teach it easily to end-users.
It must be thoroughly tested, and respect coding conventions of the existing codebase.
Features introducing new dependencies will generally not be accepted.

Adding atlases and datasets¶

Downloaders for new atlases datasets are usually discouraged.

There is no intention to provide a comprehensive collection of downloaders for the most widely used atlases and datasets. This would be outside the scope of this project. We will consider adding fetchers only for atlases and light datasets which are needed to demo and teach features.

Warning

Issues requesting to add a new atlas or dataset that are not linked to the development of a new feature or example will be closed as being out of scope.

The following projects are dedicated to host atlases and accept contributions:

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. Discussions are public, held on issues and pull requests in Github. All modifications of the codebase are ultimately checked during a reviewing process, where maintainers or contributors make sure they respect the Contribution Guidelines. To be merged, a pull request usually needs to be accepted by two maintainers. In case a consensus does not emerge easily, the decisions are made by the Core developers, i.e., people with write access to the repository, as listed here.

How to contribute to nilearn¶

This project, hosted on https://github.com/nilearn/nilearn/, is a community effort, and everyone is welcome to contribute. We value very much your feedback and opinion on features that should be improved or added. All discussions are public and held on relevant issues or pull requests. To discuss your matter, please comment on a relevant issue or open a new one.

The best way to contribute and to help the project is to start working on known issues such as good first issues, known bugs or proposed enhancements. If an issue does not already exist for a potential contribution, we ask that you first open a new issue before sending a Pull Requests to discuss scope and potential design choices in advance.

Contribution Guidelines¶

When modifying the codebase, we ask every contributor to respect common guidelines. Those are inspired from scikit-learn and ensure Nilearn remains simple to understand, efficient and maintainable. For example, code needs to be tested and those tests need to run quickly in order not to burden the development process. To keep continuous integration efficient with our limited infrastructure, running all the examples must lead to downloading a limited amount of data (gigabytes) and execute in a reasonable amount of time (less than an hour). Those guidelines will hence be enforced during the reviewing process. The section Setting up your environment will help you to quickly get familiar with the tools we use for development and deployment.

	Which PR ?	Guidelines
PR Structure	Any	Clear name Link issue through mention :”Closes #XXXX” Clearly outline goals and changes proposed Doesn’t include “unrelated” code change Add entry in “doc/changes/latest.rst”
Coding Style	Any	Variables, functions, arguments have clear names Easy to read, PEP8 compliant Code formatted with ruff Public functions have docstring (numpydoc format) Low redundancy No new dependency Backward compatibility All internal imports are absolute, not relative Impacted docstrings have versionadded and/or versionchanged directives as needed. These should use the current dev version.
Tests	Bugfixes New features	Test type is adapted to function behavior Tests pass continuous integration Coverage doesn’t decrease Fast, using small mocked data Atomic (one per function) and seeded For Bugfixes: non-regression test
Examples	New features	Clearly showcase benefits Run in a few seconds Use light data (generated or from Nilearn) Renders well after build
Documentation	Any	Simple and didactic Links to relevant examples Renders well after build Doesn’t include code

PR Structure¶

A new pull request must have a clear scope, conveyed through its name, a reference to the issue it targets (through the exact mention “Closes #XXXX”), and a synthetic summary of its goals and main steps. When working on big contributions, we advise contributors to split them into several PRs when possible. This has the benefit of making code changes clearer, making PRs easier to review, and overall smoothening the whole process. No changes unrelated to the PR should be included.

When relevant, PR names should also include tags if they fall in various categories. When opening a PR, the authors should include the [WIP] tag in its name, or use github draft mode. When ready for review, they should switch the tag to [MRG] or can switch it back to normal mode. Other tags can describe the PR content : [FIX] for a bugfix, [DOC] for a change in documentation or examples, [ENH] for a new feature and [MAINT] for maintenance changes.

Changelog¶

Changelog entries in nilearn/doc/changes/latest.rst should adhere to the following conventions:

Entry in the appropriate category
Single line per entry
Add a “badge” corresponding to the change type (see below)
Finish with a link to the PR and the author’s profile
New contributors to add their details to the authors section of the CITATION.cff (see below)

List of badges:

:bdg-primary:`Doc`
:bdg-secondary:`Maint`
:bdg-success:`API`
:bdg-info:`Plotting`
:bdg-warning:`Test`
:bdg-danger:`Deprecation`
:bdg-dark:`Code`

Example entry in nilearn/doc/changes/latest.rst:

- :bdg-dark:`Code` Fix off-by-one error when setting ticks in :func:`~plotting.plot_surf` (:gh:`3105` by `Dimitri Papadopoulos Orfanos`_).

Associated entry in CITATION.cff:

authors:

  - given-names: Dimitri Papadopoulos
    family-names: Orfanos
    website: https://github.com/DimitriPapadopoulos
    affiliation: NeuroSpin, C.E.A., Université Paris-Saclay, Gif-sur-Yvette, France
    orcid: https://orcid.org/0000-0002-1242-8990

Coding Style¶

The nilearn codebase follows PEP8 styling. The main conventions we enforce are :

line length < 80
spaces around operators
meaningful variable names
function names are underscore separated (e.g., a_nice_function) and as short as possible
public functions exposed in their parent module’s init file
private function names preceded with a “_” and very explicit, see also Guidelines for Private Functions
classes in CamelCase
2 empty lines between functions or classes

You can check that any code you may have edited follows these conventions by running ruff.

Documentation style¶

Each function and class must come with a “docstring” at the top of the function code, using numpydoc formatting. The docstring must summarize what the function does and document every parameter.

If an argument takes in a default value, it should be described with the type definition of that argument.

See the examples below:

def good(x, y=1, z=None):
    """Show how parameters are documented.

    Parameters
    ----------
    x : :obj:`int`
          X

    y : :obj:`int`, default=1
          Note that "default=1" is preferred to "Defaults to 1".

    z : :obj:`str`, default=None

    """


def bad(x, y=1, z=None):
    """Show how parameters should not be documented.

    Parameters
    ----------
    x :
          The type of X is not described

    y : :obj:`int`
          The default value of y is not described.

    z : :obj:`str`
          Defaults=None.
          The default value should be described after the type.
    """

Additionally, we consider it best practice to write modular functions; i.e., functions should preferably be relatively short and do one thing. This is also useful for writing unit tests.

Writing small functions is not always possible, and we do not recommend trying to reorganize larger, but well-tested, older functions in the codebase, unless there is a strong reason to do so (e.g., when adding a new feature).

APIs of nilearn objects¶

Estimated Attributes¶

Attributes that have been estimated from the data should always have a name ending with trailing underscore. For example the coefficients of some regression estimator would be stored in a coef_ attribute after fit has been called.

The estimated attributes are expected to be overridden when you call fit a second time.

This follows the scikit-learn convention.

Guidelines for Private Functions¶

We start a name with a leading underscore to indicate that it is an internal implementation detail, not to be accessed directly from outside, of the enclosing context:

the parent module (for a submodule name),
or the module (for the name of a top-level function, class or global variable),
or the class (for a method or attribute name).

Moreover, modules explicitly declare their interface through the __all__ attribute, and any name not listed in __all__ should not be accessed from outside the module

In some cases when private functions are in a private module (filename beginning with an underscore), but are used outside of that file, we do not name them with a leading underscore.

Example:

nilearn
├── image.py             # part of public API
├── __init__.py
├── maskers              # part of public API
│   ├── __init__.py
│   ├── nifti_masker.py  # part of public API
│   └── _validation.py   # private to the maskers module
└── _utils.py            # private to the nilearn module

Code inside maskers._validation.py:

import numpy as np  # not part of the public API

__all__ = ["check_mask_img", "ValidationError"]  # all symbols in the public API


def check_mask_img(mask_img):
    """Public API of _validation module

    can be used in nifti_masker module
    but not the image module (which cannot import maskers._validation),
    unless maskers/__init__.py imports it and lists it in __all__
    to make it part of the maskers module's public API
    """

    return _check_mask_shape(mask_img) and _check_mask_values(mask_img)


def _check_mask_shape(mask_img):
    """Private internal of _validation, cannot be used in nifti_masker"""


def _check_mask_values(mask_img):
    """Private internal of _validation, cannot be used in nifti_masker"""


class ValidationError(Exception):
    """Public API of _validation module"""


class _Validator:
    """Private internal of the _validation module"""

    def validate(self, img):
        """Public API of _Validator"""

    def _validate_shape(self, img):
        """Private internal of the _Validator class.

        As we don't use the double leading underscore in nilearn we
        cannot infer from the name alone if it is considered to be
        exposed to subclasses or not.

        """

Guidelines for HTML and CSS¶

We use prettier to format HTML and CSS.

This is implemented via a pre-commit hook (see below) that can be run with

pre-commit run --all-files prettier

Pre-commit¶

We use pre-commit to run a set of linters and autoformatters on the codebase.

To install pre-commit, run:

pip install pre-commit

Then run the following to install the pre-commit hooks:

pre-commit install

Pre-commit will then run all those hooks on the files you have staged for commit. Note that if some of those hooks fail you may have to edit some files and stage them again.

Tests¶

When fixing a bug, the first step is to write a minimal test that fails because of it, and then write the bugfix to make this test pass. For new code you should have roughly one test function per function covering every line and testing the logic of the function. They should run on small mocked data, cover a representative range of parameters.

Hint

It is easier to write good unit tests for short, self-contained functions. Try to keep this in mind when you write new functions. For more information about this coding approach, see test-driven development.

We use pytest to run our tests.

If you are not familiar with pytest, have a look at this introductory video by one of the pytest core developer.

In general tests for a specific module (say nilearn/image/image.py) are kept in a tests folder in a separate module with a name that matches the module being tested (so in this case nilearn/image/tests/test_image.py).

When you have added a test you can check that your changes worked and didn’t break anything by running pytest nilearn. To do quicker checks it’s possible to run only a subset of tests:

pytest -v nilearn/module/tests/test_module.py

Fixtures¶

If you need to do some special “set up” for your tests (for example you need to generate some data, or a NiftiImage object or a file…) you can use pytest fixtures to help you mock this data (more information on pytest fixtures in this video).

Fixture are recognizable because they have a @pytest.fixture decorator. Fixtures that are shared by many tests modules can be found in nilearn/conftest.py but some fixures specific to certain modules can also be kept in that testing module.

Before adding new fixtures, first check those that exist in the test modules you are working in or in nilearn/conftest.py.

Seeding¶

Many tests must be seeded to avoid random failures. When your test use random numbers, you can seed a random number generator with numpy.random.default_rng like in the following examples:

def test_something():
    # set up
    rng = np.random.default_rng(0)
    my_number = rng.normal()

    # the rest of the test

You can also use the rng fixture.

def test_something(rng):
    # set up
    my_number = rng.normal()

    # the rest of the test

Documentation¶

Documentation must be understandable by people from different backgrounds. The “narrative” documentation should be an introduction to the concepts of the library. It includes very little code and should first help the user figure out which parts of the library he needs and then how to use it. It must be full of links, of easily-understandable titles, colorful boxes and figures.

Examples take a hands-on approach focused on a generic usecase from which users will be able to adapt code to solve their own problems. They include plain text for explanations, python code and its output and most importantly figures to depict its results. Each example should take only a few seconds to run.

To build our documentation, we are using sphinx for the main documentation and sphinx-gallery for the example tutorials. If you want to work on those, check out next section to learn how to use those tools to build documentation.

Continuous integration¶

Please note that if one of the following markers appear in the latest commit message, the following actions are taken.

Commit Message Marker	Action Taken by CI
[skip ci]	Gtihub CI is skipped completely. Several other options are also possible, see github documentation).
[skip test]	Skip running the tests.
[skip doc]	Skip building the doc.
[test nightly]	Run tests on the nightly build of Nilearn’s dependencies.
[full doc]	Run a full build of the documentation (meaning that all examples will be run).
[example] name_of_example.py	Run partial documentation build but will run the requested example.
[force download]	Force a download of all the dataset required for the build of the documentation.

Setting up your environment¶

Installing¶

Here are the key steps you need to go through to copy the repo before contributing:

fork the repo from github (fork button in the top right corner of our main github page) and clone your fork locally:

git clone git@github.com:<your_username>/nilearn.git

(optional but highly recommended) set up a virtual environment to wor in using whichever environment management tool you’re used to and activate it. For example:

python3 -m venv nilearn
source nilearn/bin/activate

or:

conda create -n nilearn pip
conda activate nilearn

install the forked version of nilearn

Recommendation

When you install nilearn, it will use the version stored in the version control system (in this case git) to get the version number you would see if you typed in your terminal:

pip show nilearn

or:

python -c "import nilearn; print(nilearn.__version__)"

To make sure that you get the correct version number, you must fetch all the git tags from the nilearn github repository, by running the following commands:

# add the nilearn repo as an "upstream" remote
git remote add upstream https://github.com/nilearn/nilearn.git
# fetch all the tags
git fetch --all
# check that you got all the tags
git tag --list

You can then install nilearn in editable mode:

pip install -e '.[dev]'

This installs your local version of Nilearn, along with all dependencies necessary for developers (hence the [dev] tag). For more information about the dependency installation options, see pyproject.toml. The installed version will also reflect any changes you make to your code.

check that all tests pass with (this can take a while):

pytest nilearn

(optional) install pre-commit hooks to run the linter and other checks before each commit:

pre-commit install

Contributing¶

Here are the key steps you need to go through to contribute code to nilearn:

open or join an already existing issue explaining what you want to work on
on your fork, create a new branch from main:

git checkout -b your_branch

implement changes, lint and format

Recommendation

To lint your code and verify PEP8 compliance, you can run ruff locally on the changes you have made.

ruff check --fix <path_to_edited_file>

To format your code, you can also use ruff and run:

ruff format <path_to_edited_file>

Note that if you installed pre-commit and the pre-commit hooks, those commands will be run automatically before each commit.

commit your changes on this branch (don’t forget to write tests!)
run the tests locally (to go faster, only run tests which are relevant to what you work on with, for example):

pytest -v nilearn/plotting/tests/test_surf_plotting.py

push your changes to your online fork:

git push

in github, open a pull request from your online fork to the main repo (most likely from your_fork:your_branch to nilearn:main).
check that all continuous integration tests pass

For more details about the Fork Clone Push workflows, read here.

Building documentation¶

If you wish to build documentation:

First, ensure that you have installed sphinx and sphinx-gallery. When in your fork top folder, you can install the required packages using:

pip install '.[doc]'

Then go to nilearn/examples or nilearn/doc and make needed changes using reStructuredText files
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

The full build can take a very long time. So if you don’t need the plots, a quicker option is:

make html-noplot

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.
You can now open a Pull Request from Nilearn’s Pull Request page.
Request the CI builds the full documentation from your branch:

git commit --allow-empty -m "[full doc] request full build"

Tip

When generating documentation locally, you can build only specific files to reduce building time. To do so, use the filename_pattern:

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

Additional cases¶

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 is more than one map, 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.

How to contribute a dataset fetcher¶

The nilearn.datasets module provides functions to download some neuroimaging datasets, such as nilearn.datasets.fetch_haxby or nilearn.datasets.fetch_atlas_harvard_oxford.

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 a new 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.tests._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.

Maintenance¶

More information about the project organization, conventions, and maintenance process can be found there : Maintenance.