ATLAS DC2 SUSY Challenge

Updated March 29 2004


Introduction

The goals of the SUSY Working Group for DC2 are to validate the new simulation and reconstruction software and carry out sample physics analyses using data generated at two SUSY points.

The first model point is that used at DC1 and is described here. The DC1 analysis was written up in ATL-COM-PHYS-2003-055.
The second is a 'coannihilation' model with soft taus as described here.
Further details of the generation phase can be found here


Plans and Timescales

(1) Before beginning of May: Validation of 8.0.x

Now - As many people as possible should download Frank's analysis code and try to understand what it does. A test file is available here.

When 8.0.x is ready and working, use to reconstruct existing DC1 sample at LBL/BNL (should take no more than 1 night). Make root ntuples available to people either to run FEP code at e.g. BNL (if they have accounts) or to download and analyse locally using FEP code. b-tagging may not be present in 8.0.x (we hope it is) but we should not get hung up on it if not. Fortran b-tagging code exists but probably not worth the effort of translation unless there is a specific volunteer. Need volunteers to look at specific physics objects (as in DC1) and produce plots (see below). Same goes for physics plots (should be same people).

(2) Between beginning of May and middle of July (pre-production): Validation of new G4 simulation and first new physics plots.

As soon as DC2 pre-production samples are ready run 8.0.x at LBL/BNL and make output root ntuples available. Same people who validated 8.0.x with DC1 sample should then repeat with new DC2 data from DC1 model point (20k events) - produce same plots. First use no pile-up data (to validate simulation), then use pile-up (to re-validate 8.0.x in presence of pile-up). Feed back results to validation team asap. Also study co-annihilation events for validation and new physics plots (jobOptions here , auxiliary file here ).

(3) After middle of July:

Use 9.0.x locally and/or at LBL/BNL to reconstruct DC2 samples - validate 9.0.x with(out) pile-up. Output will be in AOD form and will have to be analysed using PAT framework - requires porting of analysis code (from FEP package etc.) into PAT framework. Will use main reconstruction effort as well but will not wait for it as may well have problems (we may need to hack the code ourselves locally). Make same validation plots for 9.0.x and physics plots.

Continuous production phase: At least 100k co-annihilation events (jobOptions here , auxiliary file here ) + other model points as we wish (please give thought to this). One good suggestion is a Focus Point model. Repeat simulation/reconstruction effort to produce lots more nice physics plots!


Tasks

The following tasks need volunteers:

(i) Validate tauola with atlfast output at co-annihilation point. (Now)
Atlfast ntuple is here

(ii) Download and experiment with analysis code from FEP (ALL)
Root ntuple analysis code is now available here, together with a test file.

(iii) Analyse reconstruction output sarting with 8.0.x, 9.0.x etc. to produce validation plots as in ATL-COM-PHYS-2003-055:
. electrons/photons (Fabiola)
. muons
. taus
. jets
. etmiss (Dan)

(iv) Produce physics plots (these people should also work on (iii)) as in ATL-COM-PHYS-2003-055:
. jet signatures
. electron/muon signatures
. tau signatures
. combined signatures

(v) experiment with local installation of reconstruction using kits (speak to Fred Akesson) (now)

(vi) port FEP analysis code to PAT framework (before July).

Aim should be to produce ATLAS note(s) summarising validation and physics studies as was done for DC1, as well as continuous feedback to validation.

Logistics and Progress

Root ntuple analysis code is now available here, together with a test file.

Agendas and Minutes

15 March 2004: DC2 SUSY Challenge Phone Conference





Mailing list of ATLAS SUSY Working Group


Dan Tovey   Modified 29 March 2004