SciPost Submission Page

Fast simulation of detector effects in Rivet

by Andy Buckley, Deepak Kar, Karl Nordstrom

Submission summary

As Contributors: Andy Buckley · Deepak Kar
Arxiv Link: (pdf)
Date accepted: 2020-01-28
Date submitted: 2019-12-24 01:00
Submitted by: Buckley, Andy
Submitted to: SciPost Physics
Academic field: Physics
  • High-Energy Physics - Phenomenology
Approaches: Experimental, Computational


We describe the design and implementation of detector-bias emulation in the Rivet MC event analysis system. Implemented using C++ efficiency and kinematic smearing functors, it allows detector effects to be specified within an analysis routine, customised to the exact phase-space and reconstruction working points of the analysis. A set of standard detector functions for the physics objects of Runs 1 and 2 of the ATLAS and CMS experiments is also provided. Finally, as jet substructure is an important class of physics observable usually considered to require an explicit detector simulation, we demonstrate that a smearing approach, tuned to available substructure data and implemented in Rivet, can accurately reproduce jet-structure biases observed by ATLAS.

Ontology / Topics

See full Ontology or Topics database.

Jets Large Hadron Collider (LHC)

Published as SciPost Phys. 8, 025 (2020)

Author comments upon resubmission

We present a resubmitted version, improved thanks to the referees' comments. Thanks again for your consideration.

List of changes

1. Extension of validation plots and discussion to include tau and photon observables, using ttgamma events.
2. Inclusion of Rivet & Delphes analysis routines and Delphes steering cards, as ancillary material.
3. Clarification of motivation issues like unfolding as an ill-posed problem, the discussion of smearing accuracy (and its geometric representation in Fig 1), definition of ghost association, and various suggested improvements to phrasing.
4. Adding citations to CheckMATE and MadAnalysis5 in addition to Gambit, as active and public examples of recasting code using fast-simulated reco-level quantities (other suggested codes are non-public and/or use unfolded observables).
5. Removal of poorly-defined error estimates on fitted substructure smearing variables: the numerical values are less important than the observation that a simple smearing ansatz like this can obtain good results. Undoubtedly, further focused development could improve further, but that is beyond the scope of this in-principle demonstrator.

Reports on this Submission

Anonymous Report 3 on 2020-1-22 Invited Report


The revised version sufficiently addresses the raised points.

  • validity: high
  • significance: high
  • originality: good
  • clarity: high
  • formatting: good
  • grammar: good

Report 2 by Tilman Plehn on 2020-1-20 Invited Report


Thank you for taking into account my comments, all cool now!

  • validity: -
  • significance: -
  • originality: -
  • clarity: -
  • formatting: -
  • grammar: -

Anonymous Report 1 on 2019-12-30 Invited Report

  • Cite as: Anonymous, Report on arXiv:1910.01637v3, delivered 2019-12-30, doi: 10.21468/SciPost.Report.1422


An improved version of an already good paper. My only (and entirely optional) comment/question is that now you have added citations to checkmate etc do you also want to add one to Contur, as an example of a BSM application which could directly make use of rivet routines that exploit your work?

  • validity: -
  • significance: -
  • originality: -
  • clarity: -
  • formatting: -
  • grammar: -

Author Andy Buckley on 2020-01-24
(in reply to Report 1 on 2019-12-30)

Thanks for the suggestion. We could do this, but it would be a bit pre-emptive as Contur has not yet made public use of reco-level analyses (although we are currently involved in this extension, e.g. via Les Houches projects). Since our reference of other BSM recasting tools was limited to those already have fast detector-emulation tools, and we did not reference other currently unfolded/particle-level fits which could use reco-level Rivet results, it would seem like special treatment and self-reference so we think it better to leave it as-is this time.

Anonymous on 2020-01-24
(in reply to Andy Buckley on 2020-01-24)

Fair enough. All fine.

Login to report or comment