A global likelihood for the Standard Model Effective Field Theory
Switch branches/tags
Nothing to show
Clone or download
peterstangl and DavidMStraub Delta ll individual (#1)
* add SM likelihood to obstable

* add more options for obstable output

* remove theory_sm from obstable dict

* remove spaces around "=" and break too long line

* add some comments

* change "delta_ll" to "pull SM" and rename "pull" to "pull exp."

* raise error if invalid sort key is used
Latest commit 3d5aaad Nov 23, 2018
Type Name Latest commit message Commit time
Failed to load latest commit information.
smelli Delta ll individual (#1) Nov 23, 2018
.gitignore Initial commit Oct 16, 2018
.travis.yml travis: does not support Python 3.7 yet Oct 17, 2018
LICENSE Initial commit Oct 16, 2018
MANIFEST.in Initial commit Oct 16, 2018
README.md [README] Add one more arXiv number Oct 19, 2018
setup.py add multipledispatch to install requirements Oct 17, 2018


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.


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

python3 -m pip install smelli --user


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


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.


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.


In alphabetical order:

  • Jason Aebischer
  • Jacky Kumar
  • Peter Stangl
  • David M. Straub


smelli is released under the MIT license.