cookiecutter-data-science and cookiecutter-docker-science
These are complementary tools: cookiecutter-data-science provides the project structure and organization, while cookiecutter-docker-science adds containerization capabilities that can wrap around that structure for reproducible deployment.
About cookiecutter-data-science
drivendataorg/cookiecutter-data-science
A logical, reasonably standardized, but flexible project structure for doing and sharing data science work.
Built on the [cookiecutter](https://cookiecutter.readthedocs.io/) templating framework, it provides a Python package (`ccds` CLI) that scaffolds projects with standardized directories for data pipelines, notebooks, models, and documentation, along with `pyproject.toml` and Makefile conventions for reproducibility. The template separates raw/processed data, intermediate artifacts, and source code into a modular Python package structure with dedicated modules for dataset handling, feature engineering, and model training/inference.
About cookiecutter-docker-science
docker-science/cookiecutter-docker-science
Cookiecutter template for data scientists working with Docker containers
Provides scaffolding with separate development and production Dockerfiles, abstracted data source ingestion (S3, NFS, URL), and Make targets for container lifecycle management and Jupyter integration. Supports live code editing in host editors via volume mounts while maintaining reproducible environments, with development images using bind mounts and release images packaging complete projects. Streamlines common ML workflows including dependency management, notebook execution, linting, and testing within containerized workflows.
Related comparisons
Scores updated daily from GitHub, PyPI, and npm data. How scores work