jelli is a Python package for building and evaluating likelihood functions in the Effective Field Theory (EFT) framework.
- EFT Framework: Construction of likelihoods in EFTs, such as the Standard Model Effective Field Theory (SMEFT) and Weak Effective Theory (WET).
- Flexibility: Supports arbitrary observable predictions provided in the POPxf data format, and a multitude of experimental likelihood assumptions.
- JAX Integration: Built on JAX for high-performance numerical computing.
- Differentiable: Fully differentiable likelihood functions due to JAX's autodiff, enabling efficient gradient and Hessian computations, gradient-based optimization and sampling, and more.
- Fast: Utilizes JAX's Just-In-Time (JIT) compilation for optimized performance.
- Multi-scale: Interfaced with rgevolve for fast renormalization group evolution using the evolution matrix formalism.
The package can be installed via pip:
pip install jelliThe documentation is available at https://jelli-pheno.github.io/.
A paper describing jelli is in preparation.
Please report bugs and request features via the GitHub issues page.
Authors:
- Aleks Smolkovič (@alekssmolkovic)
- Peter Stangl (@peterstangl)
jelli is licensed under the MIT License.
