Contributing to Ibis

Set up a development environment

  1. Create a fork of the Ibis repository, and clone it.

    git clone<your-github-username>/ibis
  2. Download and install Miniconda

  3. Create a Conda environment suitable for ibis development:

    cd  ibis
    conda env create
  4. Activate the environment

    conda activate ibis-dev
  5. Install your local copy of Ibis into the Conda environment. In the root of the project run:

    pip install -e .

Find an issue to work on

If you are working with Ibis, and find a bug, or you are reading the documentation and see something wrong, or that could be clearer, you can work on that.

But sometimes, you may want to contribute to Ibis, but you don't have anything in mind. In that case, you can check the GitHub issue tracker for Ibis, and look for issues with the label good first issue. Feel free to also help with other issues that don't have the label, but they may be more challenging, and require knowledge of Ibis internals.

Once you found an issue you want to work on, write a comment with the text /take, and GitHub will assign the issue to yourself. This way, nobody else will work on it at the same time. If you find an issue that someone else is assigned to, please contact the assignee to know if they are still working on it.

Working with backends

Ibis comes with several backends. If you want to work with a specific backend, you will have to install the dependencies for the backend with conda install -n ibis-dev -c conda-forge --file="ci/deps/<backend>.yml".

If you don't have a database for the backend you want to work on, you can check the configuration of the continuos integration, where docker images are used for different backend. This is defined in .github/workflows/main.yml.

Run the test suite

To run Ibis tests use the next command:

PYTEST_BACKENDS="sqlite pandas" python -m pytest ibis/tests

You can change sqlite pandas by the backend or backends (space separated) that you want to test.

Style and formatting

We use flake8, black and isort to ensure our code is formatted and linted properly. If you have properly set up your development environment by running make develop, the pre-commit hooks should check that your proposed changes continue to conform to our style guide.

We use numpydoc as our standard format for docstrings.

Commit philosophy

We aim to make our individual commits small and tightly focused on the feature they are implementing. If you find yourself making functional changes to different areas of the codebase, we prefer you break up your changes into separate Pull Requests. In general, a philosophy of one Github Issue per Pull Request is a good rule of thumb, though that isn't always possible.

We avoid merge commits (and in fact they are disabled in the Github repository) so you may be asked to rebase your changes on top of the latest commits to master if there have been changes since you last updated a Pull Request. Rebasing your changes is usually as simple as running git pull upstream master --rebase and then force-pushing to your branch: git push origin <branch-name> -f.

Commit/PR messages

Well-structed commit messages allow us to generate comprehensive release notes and make it very easy to understand what a commit/PR contributes to our codebase. Commit messages and PR titles should be prefixed with a standard code the states what kind of change it is. They fall broadly into 3 categories: FEAT (feature), BUG (bug), and SUPP (support). The SUPP category has some more fine-grained aliases that you can use, such as BLD (build), CI (continuous integration), DOC (documentation), TST (testing), and RLS (releases).

Maintainer's guide

Maintainers generally perform two roles, merging PRs and making official releases.

Merging PRs

We have a CLI script that will merge Pull Requests automatically once they have been reviewed and approved. See the help message in dev/ for full details. If you have two-factor authentication turned on in Github, you will have to generate an application-specific password by following this guide. You will then use that generated password on the command line for the -P argument.

Access the Ibis "Merging PRs" wiki page for more information.


Ibis is released in two places:

Steps to release:

Create a new version

In the master branch, after the last commit to include in the release, create a tag:

Originally, Ibis used a version like v0.0.1, but the v was eventually dropped, and recently we have been using just the 0.0.1 format.

Push the tag to the remote branch:

The remote upstream is assumed to be the main Ibis repo (i.e.

Release to PyPI

Just after the tag (without pulling new commits from master, build the Python package:

This requires twine and wheel installed, which you should have if you created your environment with the repo environment.yml file.

The package will be built in the dist/ directory. To upload it to the PyPI server, use:

This will create the new package, and will be available immediately via pip install ibis-framework.

Release to conda-forge

The conda-forge package is released using the conda-forge feedstock repository:

We need to update its recipe in a pull request, and the new version will be automatically released. After cloning the feedstock repository, update its recipe with the one in the main Ibis repository:

Remove the comment at the header of the meta.yaml file. And update the next yaml values:

Once the recipe is final, run:

This will update the azure configuration files in the feedstock repository, and possibly other files. Open a pull request with all the changes.

The conda-forge package should be ready not long after the pull request is merged, and it can be installed with conda install -c conda-forge ibis-framework.

Finally, if extra changes have been required to the recipe, besides the version, build and sha256 mentioned before, copy the recipe to the Ibis repository. Keeping the header.