[HGCAL] Deprecate MultiCluster

[lecriste]

PR description:

This PR replaces the usage of HGCal MultiCluster with Trackster.
No changes are expected, apart from

• naming shift ("MultiCluster" -> "Trackster") in DQM output.
• O(10%) difference in the MEs using the MultiCluster/Trackster energy (which may change of 0.01% for numerical precision) due to overflow migration; I verified that the unbinned histogram mean is identical.

PR validation:

runTheMatrix -l limited

[cmsbuild]

-code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-33488/22211

• This PR adds an extra 84KB to repository

Code check has found code style and quality issues which could be resolved by applying following patch(s)

• code-format:
  https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-33488/22211/code-format.patch
  e.g. curl https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-33488/22211/code-format.patch | patch -p1
  You can also run scram build code-format to apply code format directly

[cmsbuild]

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-33488/22213

• This PR adds an extra 84KB to repository

[cmsbuild]

A new Pull Request was created by @lecriste (Leonardo Cristella) for master.

It involves the following packages:

RecoHGCal/TICL
RecoParticleFlow/PFClusterProducer
Validation/HGCalValidation

@perrotta, @andrius-k, @kmaeshima, @ErnestaP, @kpedro88, @cmsbuild, @srimanob, @jfernan2, @ahmad3213, @slava77, @jpata, @rvenditti can you please review it and eventually sign? Thanks.
@mmarionncern, @sethzenz, @bsunanda, @clelange, @felicepantaleo, @cbernet, @vandreev11, @rovere, @lgray, @cseez, @apsallid, @sobhatta, @pfs, @hatakeyamak, @seemasharmafnal this is something you requested to watch as well.
@silviodonato, @dpiparo, @qliphy you are the release manager for this.

cms-bot commands are listed here

[felicepantaleo]

@cmsbuild please test

[rovere]

Ciao @lecriste thanks for this.
Would you mind taking care also of this?

[jfernan2]

OK, but mean and stddev of that (e.g.) ME is changing at the 10% level, while the changes you quote are at 0.01% level, is this really expected?

[lecriste]

Yes, the differences I quoted enter as denominator in the fraction used to fill those MEs.

[jfernan2]

+1

[perrotta]

Yes, the differences I quoted enter as denominator in the fraction used to fill those MEs.

Just to understand: if the ratio has a difference itself of order 10%, while the denominator has a difference of the order of 0,01%, shouldn't you explain why the numerator (what else?) has this 10% difference?

[jfernan2]

Yes, the differences I quoted enter as denominator in the fraction used to fill those MEs.

In any case, @lecriste you should describe this change in the PR description for the records, since in principle the PR implied a folder renaming only. Thanks

[lecriste]

@perrotta the numerator is unchanged. The difference in the fraction (of the same order of the one in the denominator) produces an overflow migration resulting in a different Mean and StdDev of the affected MEs (only the last bin differs between the two MEs). As I wrote in the PR description, the unbinned histogram mean is identical at the numerical precision level.

[perrotta]

+1

• No reco related updates since previous reco signature in [HGCAL] Deprecate MultiCluster #33488 (comment)
• This PR replaces the usage of HGCal MultiCluster with Trackster, thus cleaning the HGCal reconstruction step from producing the now deprecate multicluster's
  The only differences in the event output of the Jenkins tests just witness the removal of the multiclusters from the reco output, with no changes anywhere else, as expected; changes observed in a few DQM plots are at the level of the numerical precision

[lecriste]

@cms-sw/upgrade-l2 can you please review this PR?

[perrotta]

@cms-sw/upgrade-l2 can you please review this PR?

@hgcal-dpg since you requested to have this PR also assigned to you please don't forget to sign (if you think so, of course)

[rovere]

@perrotta thanks for the kind reminder. In fact, I'm not yet fully convinced that this PR will introduce no regression. A separate discussion is ongoing with the developer/proponent.
As soon as I'm confident, I'll approve the PR.

[rovere]

+1

[srimanob]

+Upgrade

This PR is to review the eventcontent for HGCal and to update names used in validation sequence. No change is expected.

[cmsbuild]

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. @silviodonato, @dpiparo, @qliphy (and backports should be raised in the release meeting by the corresponding L2)

[qliphy]

+1