A user-friendly library for Data Science (DS) in Python.

Our goal is to make DS more accessible to a wider audience by providing a simple, intuitive, and consistent API to solve common tasks on small to moderately sized datasets. As such, a major focus is to provide a learning tool for DS novices.

Instead of implementing DS methods from scratch, we use established DS libraries under the hood such as:

• polars for manipulation of tabular data,
• scikit-learn for classical machine learning,
• torch for deep learning, and
• matplotlib for visualization.

For more specialized tasks, we recommend using these or other DS libraries directly.

Note that this library is still in development and not yet ready for production. Expect breaking changes in the future without a major version bump (while in the 0.y.z version range). Feedback is very welcome, however! If you have a suggestion or find a bug, please open an issue. If you have a question, please use our discussion forum.

Get the latest version from PyPI:

pip install safe-ds

On a Windows PC with an NVIDIA graphics card, you may also want to install the CUDA versions of torch and torchvision:

pip install --upgrade torch torchvision --index-url https://download.pytorch.org/whl/cu121

You can find the full documentation here.

We welcome contributions from everyone. As a starting point, check the following resources:

• Setting up a development environment
• Project guidelines
• Contributing page

If you need further help, please use our discussion forum.