Data science environments, for collaboration.
Flexible. Reproducible. Governable.
conda-store provides the familiarity and flexibility of conda environments, without compromising reliability for collaborative settings.
conda-store is built to work for all team members from individual data scientists to administrators, while making sure your team follows best practices throughout the environment life cycle: from initial environment creation, to using environments in a production machine.
- Flexiblity:
- Users can create and update environments with the Graphical UI or a YAML editor.
- The environments are automatically version-controlled and all versions are readily available.
- Reproduciblity:
- User can share environments quickly through the auto-generated artifacts including a lockfile, docker image, YAML file, and tarball.
- conda-store pins exact versions of all packages and their dependencies in all the auto-generated artifacts.
- Goverance:
- Users have access to admin-approved packages and channels for their work, and can request new ones when needed.
- Admins can insert or require certain packages and versions for organization-level compatibility.
- Admins can manage users' access-levels using "Namespaces", and allow users to share environments across (and only with) their team.
Learn more, including how to install, use, and contribute to conda-store in our documentation at conda.store.
- We are working on a new UI for conda-store at:
conda-incubator/conda-store-ui, and - a JupyterLab extension at:
conda-incubator/jupyterlab-conda-store.
To guarantee a welcoming and friendly community, we require all community members to follow our Code of Conduct.
conda-store is developed under the BSD-3 LICENSED.