smelli

A global likelihood for the Standard Model Effective Field Theory

Browse the documentation

Build Status Coverage Status

smelli – a global likelihood for precision constraints

smelli is a Python package providing a global likelihood function in the space of dimension-six Wilson coefficients in the Standard Model Effective Field Theory (SMEFT). The likelihood includes contributions from quark and lepton flavour physics, electroweak precision tests, and other precision observables.

The package is based on flavio for the calculation of observables and statistical treatment and wilson for the running, translation, and matching of Wilson coefficients.

Installation

The package requires Python version 3.5 or above. It can be installed with

python3 -m pip install smelli --user

Documentation

A brief user manual can be found in the paper cited below.

Citation

If you use smelli in a scientific publication, please cite

J. Aebischer, J. Kumar, P. Stangl, and D. M. Straub

“A Global Likelihood for Precision Constraints and Flavour Anomalies”

arXiv:1810.07698 [hep-ph]

Please also cite the publications on flavio and wilson, which are the pillars smelli is built on.

Bugs and feature requests

Please submit bugs and feature requests using Github’s issue system.

Contributing

The aim of the package is to provide a likelihood in the space of dimension-6 SMEFT Wilson coefficients using all relevant available experimental measurements. If you want to contribute additional observables, the easiest way is to implement the observable in flavio. Observables implemented there can be added to the likelihood simply by adding a corresponding entry in one of the observable YAML files.

Alternatively, also observables computed in any other standalone Python package can be incorporated in principle as long as it adheres to the WCxf standard. If you want to follow this route, please open an issue to start the discussion on how to integrate it.

Contributors

In alphabetical order:

License

smelli is released under the MIT license.