PPS nanoAOD #31531
PPS nanoAOD #31531
Conversation
… This was Mario's Peruzzi's idea to have this separate sequence.
|
The code-checks are being triggered in jenkins. |
|
-code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-31531/18510
Code check has found code style and quality issues which could be resolved by applying following patch(s)
|
| protonTable | ||
| +singleRPTable | ||
| +multiRPTable | ||
| ) |
There was a problem hiding this comment.
Do you need some of these precomputed in the miniAOD ?
if yes, we do not have in the previous miniAOD so you need to set the empty sequence for those
check here
(run2_miniAOD_80XLegacy | run2_nanoAOD_94XMiniAODv1 | run2_nanoAOD_94XMiniAODv2 | run2_nanoAOD_94X2016 | run2_nanoAOD_102Xv1 | run2_nanoAOD_106Xv1).toReplaceWith(protonTables, cms.Sequence())
There was a problem hiding this comment.
What do you mean by "precomputed"? I naively thought that nanoAOD is build from miniAOD and the nano tables are built in this step.
There was a problem hiding this comment.
The current master branch should be able to run Nano also with input from previously (since 80X release) made miniAOD campaigns EOY, UL and the future UL-reMiniAOD.
So if the input collections are available since 80X release and have stable meaning, we are ok.
Not sure when the PPS entered miniAOD.
There was a problem hiding this comment.
Thanks @mariadalfonso , I understand now. PPS has consistent miniAOD only as of the Run2 UL re-reco. What what would you recommend for the era-modifier condition?
There was a problem hiding this comment.
import the modifiers
from Configuration.Eras.Modifier_run2_miniAOD_80XLegacy_cff import run2_miniAOD_80XLegacy
from Configuration.Eras.Modifier_run2_nanoAOD_94X2016_cff import run2_nanoAOD_94X2016
from Configuration.Eras.Modifier_run2_nanoAOD_94XMiniAODv1_cff import run2_nanoAOD_94XMiniAODv1
from Configuration.Eras.Modifier_run2_nanoAOD_94XMiniAODv2_cff import run2_nanoAOD_94XMiniAODv2
from Configuration.Eras.Modifier_run2_nanoAOD_106Xv1_cff import run2_nanoAOD_106Xv1
just put an empty sequence when the PPS is not defined ;)
(run2_miniAOD_80XLegacy | run2_nanoAOD_94XMiniAODv1 | run2_nanoAOD_94XMiniAODv2 | run2_nanoAOD_94X2016 | run2_nanoAOD_102Xv1 | run2_nanoAOD_106Xv1).toReplaceWith(protonTables, cms.Sequence())
|
The code-checks are being triggered in jenkins. |
|
+code-checks Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-31531/18527
|
|
A new Pull Request was created by @jwill24 (Justin Williams) for master. It involves the following packages: PhysicsTools/NanoAOD @gouskos, @cmsbuild, @fgolf, @mariadalfonso, @santocch, @peruzzim can you please review it and eventually sign? Thanks. cms-bot commands are listed here |
|
can you confirm that it runs successfully the following: runTheMatrix.py -l limited -i all --ibeos; |
|
@mariadalfonso @perrotta are there some pending comments from your side? |
This code is only affecting nanoAOD, and reco is only a secondary actor here. |
|
I don't have further comments. I agree with the suggestion by @perrotta above. |
|
Three new tables are included for PPS: singleRP, multiRP, PPSLocalTrack added in the event tree, only data are filled. Data size increase is about +6.8% for dataUL18 and +0.2% for data UL17. Last presentation at XPOG https://indico.cern.ch/event/971101/#3-pps-mc-simulation-and-minina
We consider this a temporary update to allow gain experience, as in nano we need to store only high level objects that can be used by non experts. DQM plots attached below |
|
+xpog |
|
+1
|
Concerning this particular issue, please do let us know when such utilities are available. We would like to stress that, especially the crossing angle but also beta* are needed variables for analyses. We access them from the Event Setup. We should certainly check that we are saving them at the needed precision -- we have a well defined range of crossing angle and beta* values. So these will only take a small amount of bits per event. Any external utility should be able to correctly assign the beam parameters with a LS granularity. PPS, AlCaDB and others spent considerable effort in putting in place the (online+offline) DB infrastructure for this information to be readily available for reconstruction and analysis. What level of size reduction are we talking about by removing these variables.
|
|
merge |
|
+1 |
cmsbuild
added
analysis-approved
fully-signed
and removed
analysis-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 be automatically merged. |
|
@jwill24 and @jan-kaspar |
|
Many thanks for pointing this out! We schedule this task - what would be the optimal time line? |
|
@jan-kaspar |
|
OK, noted, many thanks! |
Done in #32616. |
PR description:
Provides the relevant PPS information in the nanoAOD in the form of FlatTables. The added content includes:
Updates regarding these additions can be found in the slides: https://indico.cern.ch/event/923848/contributions/3881642/attachments/2048614/3450447/jwilliams_nanoaod.pdf
PR validation:
These changes have been shown to function as expected when running over both data and MC -- adding the PPS tables when running with data and no additional changes when running with MC. The code has been used to process nanoAOD samples for the full Run II dataset (DoubleEG and EGamma streams). Slides of the analysis using these samples can be found here: https://indico.cern.ch/event/927936/contributions/3904334/attachments/2057549/3452032/pps_diphoton.pdf