wiki:PublicAnalysisDatabase

Version 299 (modified by Benjamin Fuks, 3 years ago) ( diff )

--

MadAnalysis 5 Public Analysis Database (PAD) for recasting LHC results

WANTED: Analysis codes

Scientific reproducibility and data preservation solely depend on preserving analysis logic in a reinterpretable form. You can contribute to the HEP community by sharing the LHC recast you have implemented in the MadAnalysis 5 framework, through Public Analysis Database! Please send us your analysis code, detector card, info file and validation note to be included in PAD for public use.

More information and examples can be found in the proceedings of the second MadAnalysis 5 Workshop on LHC recasting in Korea. Analysis codes have been published, documented and got a DOI so that they can now be cited.

This page contains a collection of LHC analyses that have been implemented in the MadAnalysis 5 framework (in the expert mode) described in arXiv:1405.3982 and arXiv:1407.3278. Extensive step-by-step lecture notes can be found in arXiv:1808.00480, and documentation on the treatment of the theoretical uncertainties and high-luminosity extrapolations within the PAD can be found in arXiv:1910.11418.

The simulation of the response of the ATLAS and CMS detector can either be handled by Delphes 3, or through the new built-in SFS fast simulation relying on smearing and efficiency functions documented in arXiv:2006.09387 or here.

For each analysis, a commented C++ code is available for download, provided together with information on the analysis and references to relevant publications. Documentation on the careful validation associated with each implementation is also available.


The current version of MadAnalysis 5 is version 1.8. One of the analyses below however requires a newer version, v1.9.4 (or more recent) that can be obtained from LaunchPad.

In order to use the analyses implemented in the PAD, one needs to install a detector simulator and the corresponding library of recast analyses. Three options are available.

  • The simulation of the detector response can be dealt with Delphes 3. The PAD installation can be achieved by typing install delphes; install PAD in the MadAnalysis 5 shell, and by then restarting the program. This setup concerns most run 2 analyses currently implemented.
  • The simulation of the detector response can be dealt with the SFS infrastructure relying on smearing and efficiency functions. The PAD installation can here be achieved by typing install fastjet; install PADForSFS in the MadAnalysis 5 shell, and by then restarting the program. For now, this alternative to Delphes 3 is rather new and only few analyses use it.
  • The simulation of the detector response can be dealt with the MA5tune of Delphes 3 that is relevant for a good number of run 1 analyses. The PAD installation can be achieved by typing install delphesMA5tune; install PADForMA5tune in the MadAnalysis 5 shell, and by then restarting the program.

Although we have various options for taking care of the detector simulation (SFS, Delphes and DelphesMA5tune), we recommend, when developing a new recasting code, to use the latest MA5 version linked to either the SFS framework, or the standard Delphes 3. The implementation of a new analysis is detailed in arXiv:1808.00480, as well as here for what concerns the SFS. In a nutshell, new analyses (within the PAD or separately) can be added by running the script newAnalyzer.py (located in PAD/Build/SampleAnalyzer), passing as an argument the name of the analysis of interest. For instance,

 python newAnalyzer.py cms_sus_14_001 cms_sus_14_001

will create and modify the necessary files for adding the analysis named cms_sus_14_001. In particular, a pair of header and source C++ files cms_sus_14_001.cpp and cms_sus_14_001.h is now available in the sub-directory PAD/Build/SampleAnalyzer/User/Analyzer. We can now either implement the analysis on our own, or replace the newly created files with those shared by somebody else (or downloaded from the list below in case the automatic PAD installation introduced above is not used).

We strongly encourage users to publish their re-implemented LHC analyses on the MadAnalysis 5 dataverse. This way they become individually citable. Please send us an email to get your recast code added to the dataverse and to the PAD. Recommendations regarding the proper validation are given in arXiv:1407.3278 and arXiv:1808.00480. Note that for every analysis that is part of the PAD, a detailed validation note is required.

For any information on the implementation of an analysis, on the database or on the expert mode of the program, please visit the 'Answers' section of our LaunchPad page.


In order to test a new physics signal (corresponding to a Monte Carlo sample mysample.hep.gz or mysample.hepmc.gz), it is sufficient to start MadAnalysis 5 in the reco mode (./bin/ma5 -R) and type

    set main.recast = on
    set main.recast.store_root = False
    import <path-to-the-event-sample>
    submit

The interpreter then asks to edit the recasting card where one can tick/un-tick the analyses that need to be recast. One can also use the command

    set main.recast.card_path = <path-to-a-recasting-card>

to provide the recasting card directly. The output directory contains event samples after detector simulation if the store_root option has been set to True, as well as exclusions (1-CLs) and efficiency information for each signal region of each of the selected analyses. The cross section excluded at the 95% CL for each region is also provided as additional information.


Available Analyses

!! please properly cite all the re-implementation codes you are using; here are a PAD BibTeX file and a PADForMA5tune BibTex file for this purpose !!

ATLAS analyses, 13 TeV

Analysis Short Description Implemented by Code Validation note Version
ATLAS-SUSY-2015-06 Multijet + missing transverse momentum (3.2 fb-1) S. Banerjee, B. Fuks, B. Zaldivar MA5 dataverse PDF v1.3/Delphes3
ATLAS-SUSY-2016-07 Multijet + missing transverse momentum (36.1 fb-1) G. Chalons, H. Reyes-Gonzalez
J.Y. Araz
MA5 dataverse
MA5 dataverse
PDF Pythia files
Sec. 5.2 in 2006.09387
v1.7/Delphes3
v1.8/SFS
ATLAS-SUSY-2018-04 Staus in the ditau + met channel (139 fb-1) J. Lim, C.-T. Lu, J.-H. Park, J. Park MA5 dataverse PDF v1.8/Delphes3
ATLAS-SUSY-2018-06 Electroweakinos with Jigsaw variables (139 fb-1) J.H. Kim, T.G. Lee, J.W. Kim, H. Jang MA5 dataverse PDF v1.8/Delphes3
ATLAS-SUSY-2018-31 Sbottoms in the multibottom (including Higgs decays) + met channel (139 fb-1) J.Y. Araz, B. Fuks MA5 dataverse PDF v1.9/SFS
ATLAS-SUSY-2018-32 Electroweakinos/sleptons in the 2l + met channel (139 fb-1) J.Y. Araz, B. Fuks MA5 dataverse PDF v1.8/Delphes3
ATLAS-SUSY-2019-08 H (into b bbar) + 1 lepton + missing transverse momentum (139 fb-1) M. Goodsell MA5 dataverse PDF v1.8/Delphes3
ATLAS-EXOT-2015-03 Monojet (3.2 fb-1) D. Sengupta MA5 dataverse PDF v1.3/Delphes3
ATLAS-EXOT-2016-25 Mono-Higgs (36.1 fb-1) S. Jeon, Y. Kang, G. Lee, C. Yu MA5 dataverse PDF v1.6/Delphes3
ATLAS-EXOT-2016-27 Monojet (36.2 fb-1) D. Sengupta MA5 dataverse PDF v1.6/Delphes3
ATLAS-EXOT-2016-32 Monophoton (36.1 fb-1) S. Baek, T.H. Jung MA5 dataverse PDF v1.6/Delphes3
ATLAS-EXOT-2018-30 W' into lepton+neutrino (139 fb-1) K. Park, S. Lee, W. Jun, U. Min MA5 dataverse PDF v1.8/Delphes3
ATLAS-CONF-2016-086 b-pair + missing transverse momentum (13.3 fb-1) B. Fuks & M. Zumbihl MA5 dataverse PDF v1.6/Delphes3
ATLAS-CONF-2019-040 Jets + missing transverse momentum (139 fb-1) F. Ambrogi
J.Y. Araz
MA5 dataverse
MA5 dataverse
PDF
TBA
v1.8/Delphes3
v1.8/SFS

All detector parametrisations can be obtained from the MA5 dataverse links, together with the corresponding analysis codes.

CMS analyses, 13 TeV

Analysis Short Description Implemented by Code Validation note Version
CMS-SUS-16-033 Supersymmetry in the multijet plus missing energy channel (35.9 fb-1) F. Ambrogi and J. Sonneveld MA5 dataverse PDF v1.7/Delphes3
CMS-SUS-16-039 Electroweakinos in the SS2L, 3L and 4L channels (35.9 fb-1) B. Fuks and S. Mondal MA5 dataverse PDF v1.7/Delphes3
CMS-SUS-16-048 Compressed electroweakinos with soft leptons (35.9 fb-1) B. Fuks
J.Y. Araz
MA5 dataverse
MA5 dataverse
Sec. 19 in 2002.12220
Sec. 5.3 in 2006.09387
v1.8/Delphes3
v1.8/SFS
CMS-SUS-16-052 SUSY in the 1l + jets channel (36 fb-1) D. Sengupta MA5 dataverse PDF v1.6/Delphes3
CMS-SUS-17-001 Stops in the OS dilepton mode (35.9 fb-1) S.-M. Choi, S. Jeong, D.-W. Kang et al. MA5 dataverse PDF v1.6/Delphes3
CMS-SUS-19-006 SUSY in the HT/missing HT channel (137 fb-1) M. Mrowietz, S. Bein, J. Sonneveld MA5 dataverse PDF v1.8/Delphes3
CMS-EXO-16-010 Mono-Z-boson (2.3 fb-1) B. Fuks MA5 dataverse PDF v1.6/Delphes3
CMS-EXO-16-012 Mono-Higgs (2.3 fb-1) S. Ahn, J. Park, W. Zhang MA5 dataverse PDF v1.6/Delphes3
CMS-EXO-16-022 Long-lived leptons (2.6 fb-1) J. Chang MA5 dataverse PDF v1.7/Delphes3
CMS-EXO-17-015 Leptoquarks + dark matter in the 1mu+1jet+met channel (77.4 fb-1) A. Jueid and B. Fuks MA5 dataverse PDF v1.8/Delphes3
CMS-EXO-17-030 Pairs of trijet resonances (35.9 fb-1) Y. Kang, J. Kim, J. Choi, S. Yun MA5 dataverse PDF v1.8/Delphes3
CMS-EXO-20-004 Dark matter in the multi-jet+met channel (137 fb-1) A. Albert TBA TBA v1.9/Delphes3
CMS-HIG-18-011 Exotic Higgs decay in the 2 muons + 2 b-jet channel via 2 pseudoscalars (35.9 fb-1) J.B. Lee and J. Lee MA5 dataverse PDF v1.8/Delphes3
CMS-TOP-17-009 SM four-top analysis (35.9 fb-1) L. Darmé and B. Fuks MA5 dataverse PDF v1.7/Delphes3
CMS-TOP-18-003 SM four-top analysis (137 fb-1) L. Darmé and B. Fuks MA5 dataverse PDF v1.8/Delphes3

All detector parametrisations can be obtained from the MA5 dataverse links, together with the corresponding analysis codes.


ATLAS analyses, 8 TeV

Analysis Short Description Implemented by Code Validation note Version
ATLAS-SUSY-2013-02 (published) 0 leptons + 2-6 jets + MET G. Chalons, D. Sengupta MA5 dataverse PDF MA5tune
ATLAS-SUSY-2013-04 (published) 0 leptons + >6 jets + MET B. Fuks, M. Blanke, I. Galon MA5 dataverse PDF MA5tune
ATLAS-SUSY-2013-05 (published) stop/sbottom search: 0 leptons + 2 b-jets G. Chalons MA5 dataverse PDF (figures) MA5tune
ATLAS-SUSY-2013-11 (published) EWK-inos, 2 leptons + MET B. Dumont MA5 dataverse PDF (source) MA5tune
ATLAS-SUSY-2013-21 (published) 0 leptons + mono-jet/c-jets + MET G. Chalons, D. Sengupta MA5 dataverse PDF (source) MA5tune
ATLAS-SUSY-2014-10 (published) 2 leptons + jets + MET B. Dumont MA5 dataverse PDF (source) MA5tune
ATLAS-EXOT-2014-06 (published) mono-photons + MET D. Barducci MA5 dataverse PDF MadGraph cards MA5tune
v1.2/Delphes3
ATLAS-HIGG-2013-03 (published) ZH->ll+invisible B. Dumont MA5 dataverse PDF (source) MA5tune

All detector parametrisations can be obtained from the MA5 dataverse links, together with the corresponding analysis codes.

CMS analyses, 8 TeV

Analysis Short Description Implemented by Code Validation note Version
CMS-SUS-13-011 (published) stop search in the single lepton mode B. Dumont, B. Fuks, C. Wymant MA5 dataverse [1]
PDF (source) MA5tune
CMS-SUS-13-012 (published) gluino/squark search in jet multiplicity and missing energy S. Bein, D. Sengupta MA5 dataverse
PDF (source) MA5tune
CMS-SUS-13-016 (PAS) search for gluinos using OS dileptons and b-jets D. Sengupta, S. Kulkarni MA5 dataverse PDF(source) MA5tune
CMS-SUS-14-001 (published) Third-generation squarks in fully hadronic final states (monojet analysis) S. Sharma, S. Pandey MA5 dataverse PDF MA5tune
CMS-SUS-14-001 (published) Third-generation squarks in fully hadronic final states (top-tag analysis) S. Bein, P. Atmasiddha, S. Sharma MA5 dataverse PDF MA5tune
CMS-B2G-12-012 (published) T5/3 top partners in same-sign dilepton channel D. Barducci, C. Delaunay MA5 dataverse PDF (source), cards v1.2/Delphes3
CMS-B2G-12-022 (published) Monotops J. Guo, E. Conte, B. Fuks MA5 dataverse arXiv v1.2/Delphes3
CMS-B2G-14-004 (published) Dark matter with top quark pairs (single lepton) B. Fuks and A. Martini MA5 dataverse PDF MadGraph cards v1.2/Delphes3
CMS-EXO-12-047 (published) Monophoton J. Guo, E. Conte, B. Fuks MA5 dataverse PDF Pythia script v1.2/Delphes3
CMS-EXO-12-048 (published) Monojet J. Guo, E. Conte, B. Fuks MA5 dataverse PDF MadGraph cards v1.2/Delphes3

All detector parametrisations can be obtained from the MA5 dataverse links, together with the corresponding analysis codes.

[1] This analysis requires MINUIT libraries.

Attachments (99)

Note: See TracWiki for help on using the wiki.