Preliminary steps for UVA folks to get up and running:
Make sure you understand how the generate and use the Fun4Sim MonteCarlo setup. You don't need the reconstruction aspect but you need the generator-Geant chain to simulate different situations.
Look at the Generator options. There are three but Pythia and Legacy are the most relavant here.
https://github.com/E1039-Collaboration/e1039-wiki/wiki/Get-Started-on-SpinQuest-Simulation
After you learn about the different generators, use both to make some J/psi events generated from different vertex locations. Use the true vertex information to check that you generated what you though you generated by plotting the true vertex in each case. Also plot the true momentum distributions and plot the trueJ/psi peak from the true momentum components and make sure everything makes sense.
We are going to be moving both beam and target and looking at the following asymmetries so you should do a similar test for moving the target position. In this case you must use reconstruction to see anything.
Put the information to calculate phi in the CS frame.
Put the information on how to calculate asymmetries.
After you have confirmed that you can move the beam and target where you want them and you can reconstruct the J/psi mass peak, perform the following:
a.) Make an artificial asymmetry for the various cases: up-down for J/psi, and left-right for J/psi.
b.) Measure the amount of beam shift or target shift you need in order to see an asymmetry that you can detect.
C.) Develop a technique to measure these asymmetries for a few statistics as possible.
These tasks will need to be done for the online reconstruction and monitoring effort for SpinQuest.
- Level 1
- Vertex strip charts for single and dimuon
- Momentum Display
- Kinematic variables for the dimuon (xF, x1, x2,qT,cosphi, theta)
- Hit displays for every detector
- Event display for every reconstructed track
- DY and J/Psi L-R asymmetry
- DY and J/Psi Up-down asymmetry
You want both accumulated asymmetries and per spill, this tells you both the trend and how that compares to whats coming in.
So what should the final set of GUIs look like:
There should be one central display that has the most critical information all in one display.
There are two types of information that need to be displayed fast. Information already in the file that you can use
directly (direct information), then raw information that you need to process or reconstruct (processed information) to do something with.
First the information will be pass through the Qtracker filter so that we focus in on the most interesting events, then we can plot those events or
information from integrating of the events in a spill or integration over the spills in a run.
Direct information:
a.) One view should monitor Hodoscope in-time window and record drifts in real time with respects to trigger TDC.
b.) Detector occupancy plots organized geometrically for intuitive feel of whats happening (DC, HODO, PROP)
c.) Closest Beam profile monitor with inferred positions in X, Y of the beam also showing profile.
d.) Intensity information from both the beam-Cherenkov and the luminosity monitor for each event, these can just be histos and strip charts.
Processed Information:
a.) Plot of the Dimuon mass spectrum per spill and integrated over all spills in that run.
b.) Plot of 4-asymmetries per spill and integrated over all spills in that run.
c.) Set of kinematic variable with momentum distributions
d.) Vertex position of selected dimuons and single muons
e.) Estimated number of interaction with target (strip chart)
f.) Running calculation of J/psi and DY Sivers Asymmetry
g.) Qtracker track selection path and event display showing selected tracks and quality of tracks for good Dimuons
h.) ktracker display
High Level Analysis Information:
A.) Basic DNN anomaly detection
B.) Event display showing regularity and irregularity over time with statistical analysis
C.) Discord chat bot monitoring all changes that alerts users to system changes, anomalies and irregularities and can answer questions
D.) Similar monitoring for cryogenics with train chat bot to watch for system changes, anomalies and irregularities and can answer questions
These are so far the basics of what we need. There is certainly a lot more that we would like some of which is listed here: