ProbNum implements probabilistic numerical methods in Python. Such methods solve numerical problems from linear algebra, optimization, quadrature and differential equations using probabilistic inference. This approach captures uncertainty arising from finite computational resources and stochastic input.

Probabilistic Numerics (PN) aims to quantify uncertainty arising from intractable or incomplete numerical computation and from stochastic input using the tools of probability theory. The vision of probabilistic numerics is to provide well-calibrated probability measures over the output of a numerical routine, which then can be propagated along the chain of computation.

To get started install ProbNum using pip.

pip install probnum

Alternatively, you can install the latest version from source.

pip install git+https://github.com/probabilistic-numerics/probnum.git

Note: This package is currently work in progress, therefore interfaces are subject to change.

For tips on getting started and how to use this package please refer to the documentation. It contains a quickstart guide and Jupyter notebooks illustrating the basic usage of implemented probabilistic numerics routines.

This repository is currently under development and benefits from contribution to the code, examples or documentation. Please refer to the contribution guidelines before making a pull request.

A list of core contributors to ProbNum can be found here.

This work is released under the MIT License.

Please submit an issue on GitHub to report bugs or request changes.