Rivet is a C++ and Python framework for preservation and reinterpretation of particle collider analyses, suitable for comparison with Monte Carlo event generator predictions, and an enabler of physics model development, validation, and optimisation in both Standard Model and BSM.

Rivet is built on the dual concepts of analyses – routines which implement published experiment analyses – and projections – routines which compute observables from MC event records. It is designed to be efficient via automatic caching of projection results for each event, and very "physicist friendly" with analysis codes containing minimal boilerplate code and maximum physics.

Key classes

Analysis: base for implementation of analysis logic
Particle: representation of (meta)particles
Jet: representation of recursively clustered particle jets
FourMomentum: a 4-momentum object
Event: container of all particles
ParticleBase: core functionality for particles and jets

Key projections

ParticleFinder: retrieve all final-state particles matching cuts
FinalState: retrieve all final-state particles matching cuts
PromptFinalState: retrieve FS particles
JetFinder: base class for jet reconstruction
FastJets: interface to the FastJet clustering package
LeptonFinder: finder of "physical" bare leptons plus ~colinear photons
MissingMomentum: fiducial definition of momentum imbalances

Key tools

Modules
Filtering utilities
Container utilities
String utilities
PID utilities

Detector projections

SmearedParticles: detector efficiency and smearing functions for particles
SmearedJets: detector efficiency and smearing functions for jets
SmearedMET: detector efficiency and smearing functions for missing ET

Advanced classes (for API embedding)

AnalysisHandler: coordinator of Analysis objects in a Run
AnalysisInfo: direct access to analysis metadata (most also via Analysis)
Run: aggregator of multiple incoming event files into a single stream

For more information, please see the documentation via https://rivet.hepforge.org/