Run any standard Python code quality tool on a Jupyter Notebook
Table of contents
- Table of contents
In your virtual environment, run one of the following:
python -m pip install -U nbqa
conda install -c conda-forge nbqa
Here's an example of how to set up some pre-commit hooks: put this in your
.pre-commit-config.yaml file (see usage as pre-commit hook)
- repo: https://github.com/nbQA-dev/nbQA rev: 0.8.0 hooks: - id: nbqa-black args: [--nbqa-mutate] - id: nbqa-pyupgrade args: [--nbqa-mutate, --py36-plus] - id: nbqa-isort args: [--nbqa-mutate]
If you need to select specific versions of any of these linters/formatters,
add them to
Reformat your notebooks with black:
$ nbqa black my_notebook.ipynb --nbqa-mutate reformatted my_notebook.ipynb All done! ✨ 🍰 ✨ 1 files reformatted.
Sort your imports with isort:
$ nbqa isort my_notebook.ipynb --nbqa-mutate Fixing my_notebook.ipynb
Upgrade your syntax with pyupgrade:
$ nbqa pyupgrade my_notebook.ipynb --py36-plus --nbqa-mutate Rewriting my_notebook.ipynb
🥳 Used by
Take some inspiration from their config files
- alibi .pre-commit-config.yaml
- GoogleCloudPlatform/ai-platform-samples pyproject.toml
- intake-esm pyproject.toml .pre-commit-config.yaml
- LiuAlgoTrader: requirements/dev.txt
- mplhep: pyproject.toml .pre-commit-config.yaml
- NLP Profiler: pyproject.toml .pre-commit-config.yaml
- pandas-profiling .pre-commit-config.yaml
- PlasmaPy .pre-commit-config.yaml
- pyhf pyproject.toml .pre-commit-config.yaml
- PyMC3: pyproject.toml .pre-commit-config.yaml
- pytorch-forecasting: pyproject.toml .pre-commit-config.yaml
- Python compiled microcourse: .pre-commit-config.yaml
- ruptures: .pre-commit-config.yaml
- sktime: .pre-commit-config.yaml
Is your project missing? Let us know, or open a pull request!
Michael Kennedy & Brian Okken, hosts of the Python Bytes podcast:
This is really cool. I think it brings so much of the code formatting and code analysis, clean up to notebooks, which I think had been really lacking
Nikita Sobolev, CTO at wemake.services:
It is amazing!
well done on
nbqa@MarcoGorelli ! Will be super useful in CI
Matthew Feickert, Postdoc at University of Illinois working on LHC physics:
nbqa in your pre-commit hooks along with @codewithanthony 's pre-commit CI service is amazing! Everyone using Jupyter notebooks should be doing this.
Girish Pasupathy, Software engineer and now core-contributor:
thanks a lot for your effort to create such a useful tool
Simon Brugman, Data scientist & pandas-profiling dev:
nbQA helps us to keep notebooks to the same standards as the rest of the code. If you're serious about your code standards, you should keep them consistent across both notebooks and python scripts. Great addition to the ecosystem, thanks!
Bradley Dice, PhD Candidate in Physics & Scientific Computing:
nbqa is a clean, easy to use, and effective tool for notebook code style. Formatting and readability makes a huge difference when rendering notebooks in a project's documentation!
James Lamb, engineer @saturn_cloud, LightGBM maintainer
today I learned about
nbqa, a command-line tool to run linters like
flake8over #Python code in @ProjectJupyter notebooks. Thanks to @jayyqi for pointing me to it. So far, I really really like it.
I will give write-access to anyone who makes a useful pull request - see the contributing guide for details on how to do so.
Thanks goes to these wonderful people (emoji key):
This project follows the all-contributors specification. Contributions of any kind welcome!