Open paulgessinger opened 1 year ago
In ATLAS. We'll soon have HI run where significant part of the data are UPC events. Typically a (very)few particles in event. Given AMVF is part of ATLAS reco already this may be affecting reconstruction (I do not think we care about it in HLT).
How does one reproduce this issue actually? Setup vertex ting chain with AMVF and particle gun to generate 4,3,2 muon events. Right? @andiwand @paulgessinger
to reproduce this issue you can run full_chain_odd.py
with AMVF instead of IVF. the underlying problem seems to be the high number of duplicated tracks. the vertex fitting blows up under these conditions which seems fair.
This issue/PR has been automatically marked as stale because it has not had recent activity. The stale label will be removed if any interaction occurs.
AMVF will produce an error in the current scenario. I guess it is because we loose one of the muons and then fail to fit the vertex. with 4 muons AMVF works but IVF even works with 2 but I am not sure why
Originally posted by @andiwand in https://github.com/acts-project/acts/issues/1299#issuecomment-1206199044