Choose timezone
Your profile timezone:
Informal CMS/ATLAS SMEFT fitting exercise meeting
→
Europe/Zurich
Interplay between propagator corrections and top spin: madspin does not implement SMEFT corrections, but not using it would not include spin correlations, etc. One could generate the full matrix element (pp -> WbWb + X ?) in MG, but this is computationally intensive. Other options - can we factorize prod/decay a la Higgs? Has been tested in CMS - to follow-up.
Common gitlab - could be useful for dumping yaml/json files with parametrisations.
Z-pole: derived by M. Trott (https://arxiv.org/abs/2102.02819) as part of work with ATLAS. Another theory parameterization is available (https://arxiv.org/abs/2012.02779 , with code here: https://gitlab.com/kenmimasu/fitrepo). A comparison showed good agreement between the two.
YAML: should store also quadratic terms, and possibly also the stat uncertainties where relevant. For Higgs, should also have the values for production and decay separately, before merging (and linearization) of the two.
Uncertainties: could add the Higgs theory uncertainties as NPs (in addition to the lumi). For ATLAS this is already folded in the covariance matrix of the Higgs measurement, but some of the uncertainties could be subtracted out of the covariance matrix and then added back as NPs.
Compare linear-only to start with, then move to linear+quadratic (with linear + diagonal-quadratic as an intermediate step ?)
For now avoid loop processes in Higgs -- both collaborations should have the technology to do this, but this is more time consuming and may benefit from an additional validation round.
Goals for the next round:
- combined Gaussian workspaces
- EFT parameterizations for all bins
- EFT-reparameterized workspaces
- single-parameter measurements (linear, and linear+quadratic if available)
- possibly also eigenvector-based measurements, but this can be for the next round.
- Start testing inputs from other expt. (if time allows)
Mailing lists: will create 2: a public one for general discussions, and one restricted to CMS+ATLAS only for technical discussions of experimental issues.
There are minutes attached to this event.
Show them.
