Choose timezone
Your profile timezone:
di-boson
Minutes LHC di-boson group meeting, 8.11.2013:
Organisation:
Agreed on having regular meetings every 4 weeks to accelerate the progress.
Next meeting 6.12.2013, 16h00 CET
Limits representation:
ATLAS proposes to have lambda dependent limit representation for observed and expected limits.
CMS proposes that as a baseline limits without form factors or other unitarization schemes always be shown and where necessary and well defined lambda dependent limits be shown.
CMS agrees that showing limits as a function of lambda is a good proposal, but prefers this representation to specific cases, e.g. where unitarization is necessary to avoid a divergent cross section and where there is a clearly defined correct choice of the exponents for a lambda dependent form factor.
CMS suggested that where a clear choice of unitarization schemes are unavailable the non-unitary region should be understood using separate programs and reported. (See CMS WVg analysis)
ATLAS prefers to give only numerical results for unitarised scenarios, as unitarity violating limits are non-physical.
No conclusive common position has been reached at this point and the discussion between ATLAS and CMS on this topic are ongoing.
Optimisation:
- ATLAS prefers to optimize the selection for unitarised aGC implementations.
- ATLAS and CMS agree that optimization procedures should include unitarization or be designed not to be highly optimized in the unitary region.
- CMS and ATLAS agree that where available k-matrix should be used.
- CMS prefers not to use Lambda dependent form factors - see notes below.
Combination issues:
Use f_ij to exchange in a first step. More detailed information may be needed (e.g. fine grid).
ATLAS and CMS agreed that the relative difference should not exceed 10^-3 to pass for delta LL when comparing limits obtained from limit setting codes of ATLAS and CMS using the same input data.
For 7 TeV ZZ combine ATLAS (4l) with CMS (4l) as a test case not for public release.
Proceed to more complete combination of ZZ results including 7 TeV ATLAS (4l, 2l2v) with 7 and 8 TeV CMS (4l) and any new ZZ results for public release.
Other remarks:
CMS felt it is difficult to agree on form factor which would not change when analysis is updated, which makes difficult comparison with previous result.
CMS remarked K-matrix has been made recently available in Madgraph. Validation is ongoing.
Action Items:
AI: Optimisation:
- not clear how, as exercise optimize for lambda = inf and lambda=X, where X is chosen such that unitarity is ensured.
- Plot the expected limits as function of lambda and quantify the difference.
AI: Encourage theorists to implement k-matrix for 2->2 scattering and triboson production for all major generators.
AI: How is luminosity correlation handled between experiments ? talk for next meeting.
AI: theoretical cross section calculation the same in CMS and ATLAS? if difference exists probably need to be propagated as systematics.
AI: Need details on how to translate to EFT in charge and neutral case.
AI: First combination of ZZ to be presented in the next meeting in four weeks.
Organisation:
Agreed on having regular meetings every 4 weeks to accelerate the progress.
Next meeting 6.12.2013, 16h00 CET
Limits representation:
ATLAS proposes to have lambda dependent limit representation for observed and expected limits.
CMS proposes that as a baseline limits without form factors or other unitarization schemes always be shown and where necessary and well defined lambda dependent limits be shown.
CMS agrees that showing limits as a function of lambda is a good proposal, but prefers this representation to specific cases, e.g. where unitarization is necessary to avoid a divergent cross section and where there is a clearly defined correct choice of the exponents for a lambda dependent form factor.
CMS suggested that where a clear choice of unitarization schemes are unavailable the non-unitary region should be understood using separate programs and reported. (See CMS WVg analysis)
ATLAS prefers to give only numerical results for unitarised scenarios, as unitarity violating limits are non-physical.
No conclusive common position has been reached at this point and the discussion between ATLAS and CMS on this topic are ongoing.
Optimisation:
- ATLAS prefers to optimize the selection for unitarised aGC implementations.
- ATLAS and CMS agree that optimization procedures should include unitarization or be designed not to be highly optimized in the unitary region.
- CMS and ATLAS agree that where available k-matrix should be used.
- CMS prefers not to use Lambda dependent form factors - see notes below.
Combination issues:
Use f_ij to exchange in a first step. More detailed information may be needed (e.g. fine grid).
ATLAS and CMS agreed that the relative difference should not exceed 10^-3 to pass for delta LL when comparing limits obtained from limit setting codes of ATLAS and CMS using the same input data.
For 7 TeV ZZ combine ATLAS (4l) with CMS (4l) as a test case not for public release.
Proceed to more complete combination of ZZ results including 7 TeV ATLAS (4l, 2l2v) with 7 and 8 TeV CMS (4l) and any new ZZ results for public release.
Other remarks:
CMS felt it is difficult to agree on form factor which would not change when analysis is updated, which makes difficult comparison with previous result.
CMS remarked K-matrix has been made recently available in Madgraph. Validation is ongoing.
Action Items:
AI: Optimisation:
- not clear how, as exercise optimize for lambda = inf and lambda=X, where X is chosen such that unitarity is ensured.
- Plot the expected limits as function of lambda and quantify the difference.
AI: Encourage theorists to implement k-matrix for 2->2 scattering and triboson production for all major generators.
AI: How is luminosity correlation handled between experiments ? talk for next meeting.
AI: theoretical cross section calculation the same in CMS and ATLAS? if difference exists probably need to be propagated as systematics.
AI: Need details on how to translate to EFT in charge and neutral case.
AI: First combination of ZZ to be presented in the next meeting in four weeks.
There are minutes attached to this event.
Show them.
