Speaker
Description
dilax is a software package for statistical inference using likelihood
functions of binned data. It fulfils three key concepts: performance,
differentiability, and object-oriented statistical model building.
dilax is build on JAX - a powerful autodifferentiation Python frame-
work. By making every component in dilax a “PyTree”, each compo-
nent can be jit-compiled (jax.jit), vectorized (jax.vmap) and differ-
entiated (jax.grad). This enables additionally novel computational
concepts, such as running thousands of fits simultaneously on a GPU
or differentiating through measurements of physical observables.
We present the key concepts of dilax, show its features, and discuss
performance benchmarks with toy datasets.
Significance
This project is a new statistics tool suited for typical measurements in LHC analyses. It focusses on performance, usability, and novel computing techniques such as autodifferentiation and vectorization of full fits. This project has not been presented so far. It introduces new concepts that have not been covered by other statistics libraries.
| Experiment context, if any | Use case for CMS, ATLAS, LHCb analyses |
|---|
