Jet Observable Library for Monte Carlo Validation

Description

Jets are collimated sprays of particles produced abundantly at the Large Hadron Collider. Comparing Monte Carlo generator predictions against data across a comprehensive set of jet observables is essential for validating and tuning simulations. However, the landscape of jet observables — ranging from simple kinematic variables (pT, mass, width) to sophisticated substructure quantities (N-subjettiness, energy correlation functions, Lund jet plane projections) — is large and fragmented across different codebases. This project aims to build a unified, well-documented Python/C++ library of jet observables that can be computed from standard event record formats. The library should provide a single interface to compute a wide catalog of observables, organized by category (IRC-safe, groomed, substructure, event shapes, etc.), and produce standardized output suitable for downstream comparison and plotting.

Duration

Total project length: 175/350 hours.

Task ideas

Survey the literature and existing tools (FastJet, EnergyFlow, etc.) to compile a comprehensive catalog of jet observables used in LHC analyses
Design a clean, modular Python API where each observable is a callable with standardized input/output
Implement core observables: jet kinematic variables, jet shapes, N-subjettiness ratios, energy correlation functions, groomed observables, Lund plane densities, etc
Ensure IRC safety annotations and metadata for each observable
Write unit tests using known analytic results or reference implementations
Provide example notebooks demonstrating usage on public datasets (e.g., from the LHC Olympics or JetClass)

Expected results:

A pip-installable Python library implementing at least 20–30 jet observables
Comprehensive documentation with observable definitions, references, and usage examples
Test suite with validation against reference implementations
Example notebooks for Monte Carlo validation workflows

Difficulty level

Intermediate

Requirements

Python/C++, familiarity with NumPy and scientific computing
Some knowledge of jet physics or willingness to learn from the literature
Experience with software engineering best practices (testing, documentation, packaging) is a plus

Test

Please use this link to access the test for this project.

Mentors

Steve Mrenna (Fermilab)
Konstantin Matchev (University of Alabama)
Tony Menzo (University of Alabama + Fermilab)
Ian Pang (Rutgers University)

Please DO NOT contact mentors directly by email. Instead, please email ml4-sci@cern.ch with Project Title and include your CV and test results. The mentors will then get in touch with you.

Corresponding Project

HEPSIM