Conversation
steering-files/src/main/resources/org/hps/steering/recon/PhysicsRun2019_pass2_recon.lcsim
Show resolved
Hide resolved
steering-files/src/main/resources/org/hps/steering/recon/PhysicsRun2019_pass2_recon.lcsim
Show resolved
Hide resolved
|
I was digging into this line-by-line to make the corresponding MC file, and I noticed something I wanted to run by both of you. The beam variables under KalmanPatRecDriver are: The beam variables under ReconParticleDriver_Kalman are: and the actual beam parameters are: Are these differences meaningful / should we updated this? |
|
Ah good catch! Yes, we should update this. For the 2021 dataset processing, are using a dynamic beam spot loaded through the conditions DB. In order for that to work, we will need this first pass so we can characterize the beam spot and then add it to the conditions DB. So for this first pass, we'll need to use a fixed one. @EBerzin , from your multi-track vertex fits, can you put in sensible (x,y) beam spot positions and uncertainties? The uncertainty on the 100 track fit will be quite a bit smaller than what we'd expect for 2-track vertex fits, so for the uncertainty I'd make it artificially large for the time being, OR we could use the values from 2021 which are probably pretty good (sigmaX 0.055 and sigmaY 0.045). It's not too important at this point. Forsure we need to make sure the vertex Z position is the same across both drivers. @mghrear Note: for MC, we've typically generated beamspots at (x,y) = (0,0) until we have a better idea of how the data looks. |
|
I updated the beam positions and used the 2021 uncertainties. |
|
Do we need a separate PR to get those changes in? |
Added ability to skip sensor in RawTrackerHitFitterDriver.
Added steering files with new seeding strategies, masking L5b.