Single-jet triggers at low PU

Unlike the single-jet triggers at high PU in Run 2, the single-jet triggers at low PU seem to exhibit differences. For instance, HLT_HIPFJet25 and HLT_PHIAK4PFJet15 in Run17H have different PreL1min and PreL1max because they are seeded by different L1.

Proposal in private exchanges with Silvio Donato on 31-March 2023:

• use the prescale from brilcal and use the kinematic matching at high pT as a cross check of the goodness of the method.
• For HLT paths using multiple L1 seeds, force a single L1 seed and consider only correspoding L1 seed prescale

added to epic &6

added On hold label

changed the description

added changeminor typephysics labels

For a single trigger like HLT_HIPFJet25_v*, brilcal output:

Singularity> brilcalc trg --prescale --hltpath HLT_HIPFJet25_v*  -r 306929
+--------+-------+----------+-------------+--------------------+-------+---------------------------------------------------+
| run    | cmsls | prescidx | totprescval | hltpath/prescval   | logic | l1bit/prescval                                    |
+--------+-------+----------+-------------+--------------------+-------+---------------------------------------------------+
| 306929 | 1     | 0        | 0.00        | HLT_HIPFJet25_v1/0 | OR    | L1_SingleJet8_BptxAND/0.00 L1_ZeroBias/4001.00    |
| 306929 | 2     | 2        | 0.00        | HLT_HIPFJet25_v1/0 | OR    | L1_SingleJet8_BptxAND/0.00 L1_ZeroBias/13610.00   |
| 306929 | 15    | 13       | 2722.00     | HLT_HIPFJet25_v1/2 | OR    | L1_SingleJet8_BptxAND/1920.00 L1_ZeroBias/1361.00 |
| 306929 | 22    | 16       | 2722.00     | HLT_HIPFJet25_v1/2 | OR    | L1_SingleJet8_BptxAND/1440.00 L1_ZeroBias/1361.00 |
| 306929 | 36    | 19       | 1920.00     | HLT_HIPFJet25_v1/2 | OR    | L1_SingleJet8_BptxAND/960.00 L1_ZeroBias/1361.00  |
| 306929 | 47    | 21       | 1280.00     | HLT_HIPFJet25_v1/2 | OR    | L1_SingleJet8_BptxAND/640.00 L1_ZeroBias/1361.00  |
| 306929 | 217   | 2        | 0.00        | HLT_HIPFJet25_v1/0 | OR    | L1_SingleJet8_BptxAND/0.00 L1_ZeroBias/13610.00   |
| 306929 | 221   | 3        | 0.00        | HLT_HIPFJet25_v1/0 | OR    | L1_SingleJet8_BptxAND/0.00 L1_ZeroBias/13610.00   |
+--------+-------+----------+-------------+--------------------+-------+---------------------------------------------------+

The logic is or. For different luminosity sections (LS), we use different L1 triggers.

Can we get the lumi section information in the Ttree, and determine which prescale should be used?

The LS is stored in RecEvent::lumi (https://dasanalysissystem.docs.cern.ch/structDAS_1_1RecEvent.html#a9d7de4191265f2f721dbcf5783f4a75f), but perhaps we need to store some L1 information too in the n-tuples. If needed, you can introduce another vector<bool>.

I have myself limited experience with the L1 trigger---but perhaps we can force the least prescaled L1 trigger and ignore the other?

Use the least prescaled L1 trigger is right. Always check if the trigger can be used in HLTinfo and it's prescale value.

Maybe an example how to treat these kinds of trigger: indico

Not sure why you have provided this link... it's about JERC business... I recall a presentation in a WG meeting where the mystery of the combination of the triggers was explained with a stack plot showing a nicely smooth spectrum. That would be more relevant, I think.

https://indico.cern.ch/event/1496966/#2-normalisation-of-runh

added statusongoing label and removed statuson hold label

assigned to @zhye

added statusfinalising label and removed statusongoing label