Cookiecutter template for datascience projects.
To start a new project from this cookiecutter, install cookiecutter (duh) and then do;
cookiecutter gh:eliavw/cookiecutter-datascience
This is still a work in progress and will likely never be finished. However, everything can always be done better.
Some ideas for new, extra features are;
- CI on github
- Edit
setup.py
so that it parsesdependencies-deploy.yaml
and adds them tosetup.cfg
automatically. Also, as a policy; adopt pip first, conda second. For python packages, everything has to be in pip anyways. - Auto-generate API docs from docstrings
- Unittest docs and code style
- Generate tests from notebooks
- git hooks
This repo is also basically an example of how you can also make your own cookiecutter!