Add DiMuonNoIP same sign lines and fix minor issues
Goes with lhcb-datapkg/PRConfig!329 (merged).
This MR adds 9 lines to the DiMuonNoIP lines which have rates of:
| line | rate (kHz) | prescale | stream |
|---|---|---|---|
| SS mass range 1 (0-740 MeV) | 3.2 | 0.1 | turbo |
| SS mass range 2 (740 MeV - 1.1 GeV) | 2.5 | 0.1 | turbo |
| SS mass range 3 (1.1 - 3 GeV) | 5.1 | 0.1 | turbo |
| SS mass range 4 (3 - 3.2 GeV) | 0.5 | 0.1 | turbo |
| SS mass range 5 (3.2 - 9 GeV) | 2.2 | 0.1 | turbo |
| SS mass range 6 (9 - 110 GeV) | 0.02 | 0.1 | turbo |
| SS displaced | 2.5 | 0.1 | turbo |
| full OS | 0.098 | 0.001 | full |
| full SS | 0.124 | 0.001 | full |
The rates were calculated using the UpgradeHLT1FilteredWithGEC sample. These lines can be prescaled further if necessary.
These lines are for the dark photon analysis and use the cuts used in run two for the same lines.
Some other minor fixes were made to the DiMuonNoIP lines as a whole including moving the IPchi2 cut to the muons and fixing the opposite sign decay descriptor. It is also moved to the qee folder. This removes the DiMuonNoIP module, necessitating changes in options files that try to import this(/all) module(s); see also lhcb-datapkg/PRConfig!329 (merged).
The OS lines that were already merged have rates around:
| line | rate (kHz) | prescale | stream |
|---|---|---|---|
| mass range 1 (0-740 MeV) | 8.8 | 1 | turbo |
| mass range 2 (740 MeV - 1.1 GeV) | 7.1 | 1 | turbo |
| mass range 3 (1.1 - 3 GeV) | 81.7 | 1 | turbo |
| mass range 4 (3 - 3.2 GeV) | 9.7 | 1 | turbo |
| mass range 5 (3.2 - 9 GeV) | 24.6 | 1 | turbo |
| mass range 6 (9 - 110 GeV) | 0 | 1 | turbo |
| displaced | 27.4 | 1 | turbo |
| prescaled all mass | 10.3 | 0.1 | turbo |
Please note that the last one is completely included in the mass ranges currently, but is there in case of future prescaling.
The size of a dimuon event was measured using the bandwidth test to be 4 kB, making the bandwidth of both the OS and SS lines much less than the rates suggest.
Using the HltEfficiencyChecker tool, the efficiencies for J/psi -> mumu and Upsilon -> mumu were measured as 0.647 +/- 0.116 and 0.867 +/- 0.051 respectively.
Merge request reports
Activity
added RTA label
added qee label
Resolved by Luke GrazetteFYI @lugrazet incoming QEE lines.
- Last reply by Luke Grazette
- Resolved by Kate Abigail Richardson
Hi @kaaricha.
From looking at the current Rate tests, that these lines were previously down as 'other', so from QEE's perspective you are adding in a significant amount of rate for OS lines as well as the creation of these SS lines, and thus we will need the rate estimates for both sets of lines in the description.
Even just looking at the prescaled SS rates they're much larger than qee's current rate, ~15 kHz compared to the current 0.5kHz.I agree that only saving the dimuon event to tape will reduce the bandwidth, but turbo events are still, by my understanding, around ~50kB, meaning with this much rate that would correspond to ~750 MB/s (not a trivial amount). Would it be possible for you to check the size of these
.mdffiles?
On the topic of rates, the 9-110 GeV mass window has been reported with 0 kHz, I'm curious how many events you ran over and thus what sort of bound that gives us on this rate?.I would normally want some efficiencies from signal MC per line, however I'm not entirely sure that there's good signal samples for some of these mass ranges. At least for the Jpsi and Upsilon mass regions I think seeing some respective efficiencies will be necessary.
The code for the lines themselves seem okay to me. I think a somewhat more compact version could be done via a loop over the mass ranges as there seems to only small differences per line, however we do have the deadline for new Hlt2 lines looming soon, so the priority is discussing and controlling the rates/bandwidths.
- Last reply by Kate Abigail Richardson
assigned to @lugrazet
Resolved by Sebastien Ponce/ci-test
added ci-test-triggered label
- Resolved by Luke Grazette
Seems like we now have a fail in the pipeline : build-docs
- Last reply by Luke Grazette
- [2023-05-26 16:09] Validation started with lhcb-master-mr#8038
- [2023-05-30 20:26] Validation started with lhcb-master-mr#8074
Edited by Software for LHCbadded hlt2-throughput-decreased label
added 2 commits
- Resolved by Luke Grazette
/ci-test
- Last reply by Kate Abigail Richardson
removed hlt2-throughput-decreased label