Vertexing and DTF fitting with neutrals
Vertexing with neutrals seem to have issues at this point: https://indico.cern.ch/event/1098540/contributions/4746847/attachments/2394113/4093188/ParticleAdder.pdf Only simple mass cuts are being kept in the combiners.
For many radiative + neutral PID lines will need full DTF-based mass-constraints (KS0, ETA', D(s)), FD-significances, etc. Particularly true of the PID lines, where vertexing is critical in cleaning things up.
Neutral PID functors are available: !2736 (merged) However, these are trained on MC and from Run1/2 experience, it's better to train on calib data. So we need to make sure that the input calo variables are persisted. This includes the 3X3 (or better 5X5) cell energies for PID development, or at least the ones in the current IsPhoton
/CL
.
NB: I don't have much experience of IsNotE
, but since the PS/SPD are removed, this is no longer available.
I am creating both the exclusive RD radiative lines (hhG,hhKsG, hhhG) and porting the TurCal lines in the bdey_pid_neutral branch. Comments are welcome!