Optimizations to the Scouting Data Formats #37646
Conversation
|
-code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-37646/29444
Code check has found code style and quality issues which could be resolved by applying following patch(s)
|
|
Please fix the issue with |
dsperka
changed the title
|
+code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-37646/29446
|
|
A new Pull Request was created by @dsperka for master. It involves the following packages:
@Martin-Grunewald, @Dr15Jones, @smuzaffar, @makortel, @cmsbuild, @missirol can you please review it and eventually sign? Thanks. cms-bot commands are listed here |
There was a problem hiding this comment.
To make the changes apparent in a backward compatible way we added new members to the data formats to flag these changes.
Technically already adding members breaks the reading of old streamer files (and for ROOT more complex evolution would work as well).
Has any data been written in streamer files already with the earlier version?
| <version ClassVersion="3" checksum="1086011373"/> | ||
| <version ClassVersion="4" checksum="1250202632"/> | ||
| </class> | ||
| <class name="Run3ScoutingMuon" ClassVersion="4"> | ||
| <version ClassVersion="3" checksum="3882497397"/> | ||
| <version ClassVersion="4" checksum="4206297195"/> | ||
| </class> | ||
| <class name="Run3ScoutingPhoton" ClassVersion="4"> | ||
| <class name="Run3ScoutingPhoton" ClassVersion="5"> | ||
| <version ClassVersion="5" checksum="1238010495"/> |
There was a problem hiding this comment.
Please move below 4 and use consistent indentation.
Co-authored-by: Marino Missiroli <m.missiroli@cern.ch>
…mssw into from-CMSSW_12_3_0_pre6
|
I don't know if streamer files have been written with previous versions. If yes and thats a problem, we need some mechanism to evolve these data formats. It is not reasonable to assume there will be no changes through the next 4 years. |
|
-code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-37646/29519
Code check has found code style and quality issues which could be resolved by applying following patch(s)
|
|
Pull request #37646 was updated. @Martin-Grunewald, @Dr15Jones, @smuzaffar, @makortel, @cmsbuild, @missirol can you please check and sign again. |
|
please test |
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-88f11e/24343/summary.html The following merge commits were also included on top of IB + this PR after doing git cms-merge-topic:
You can see more details here: Comparison SummarySummary:
|
|
@cmsbuild, please test For cleaner comparisons |
|
+1 Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-88f11e/24350/summary.html Comparison SummarySummary:
|
|
+core
|
|
+hlt |
cmsbuild
added
fully-signed
hlt-approved
and removed
hlt-pending
pending-signatures
labels
|
This pull request is fully signed and it will be integrated in one of the next master IBs (tests are also fine). This pull request will now be reviewed by the release team before it's merged. @perrotta, @dpiparo, @qliphy (and backports should be raised in the release meeting by the corresponding L2) |
|
+1 |
PR description:
This PR makes some optimizations to the scouting data format. We store the track quantities for pfCandidates as relative to the pfCandidate itself, which results in a better compression. We also dropped pfCandidate mass, because it is redundant w.r.t. pdgId. We also only store non-zero rechit energies, resulting in smaller data size when also storing the associated detIds.
To make the changes apparent in a backward compatible way we added new members to the data formats to flag these changes.
These changes, together with other updates to the HLT configuration not in this PR, result in a scouting event size of ~7.7 kb.