EGM .cc and python patch for new regression tags in the DB

[lgray]

The PR contains .cc updates and ESPrefers for the new regressions.
The global tags are in preparation.

In general we see moderate to substantial improvements in the photon/electron energy response in all eta regions.

You can see the overview of the training here:
https://indico.cern.ch/getFile.py/access?contribId=3&resId=0&materialId=slides&confId=293993

The result of private reRECO (black) on top of vanilla pre12 (blue) attached for electrons (Zee 50ns PU, 13 TeV):
--- ecal energy only

--- combined p4 measurement

and for photons (Hgg in 50ns PU, 13 TeV):

[cmsbuild]

A new Pull Request was created by @lgray (Lindsey Gray) for CMSSW_7_0_X.

EGM .cc and python patch for new regression tags in the DB

It involves the following packages:

RecoEcal/EgammaClusterAlgos
RecoEcal/EgammaClusterProducers
RecoEgamma/EgammaElectronAlgos
RecoEgamma/EgammaElectronProducers
RecoEgamma/EgammaPhotonAlgos
RecoEgamma/EgammaPhotonProducers
RecoEgamma/EgammaTools
RecoParticleFlow/PFProducer

@nclopezo, @cmsbuild, @anton-a, @thspeer, @slava77, @Degano can you please review it and eventually sign? Thanks.
You can sign-off by replying to this message having '+1' in the first line of your reply.
You can reject by replying to this message having '-1' in the first line of your reply.
@ktf you are the release manager for this.
You can merge this pull request by typing 'merge' in the first line of your comment.

[lgray]

@rcastello Here's the PR for the changes we talked about yesterday. If the new GTs can be made soon we can remove the ESPrefers.

[slava77]

Hi Lindsey,

What are the _[1-5] in the plots?

[lgray]

Slices in increasing eta, roughly 0.5 units of eta per slice.

[slava77]

working on it

[cmsbuild]

+1
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-2202/10/summary.html

[slava77]

Set 1 (based on CMSSW_7_0_X_2014-01-28-1400)

Noticeable improvements overall for real electrons

• 1 TeV ele gun
  
  
• 35 GeV electrons
  
• 35 GeV electrons DQM: the scale actually went down a bit (is it because this is mean not a peak?)
  
• 35 GeV photons
  
• ttbar PU the EE cluster count went down by more than 20% (it's unclear where this is coming from .. electrons in EE went down by ~4%, from a total of 70)
  

Lindsey, is the EE cluster count expected?

[lgray]

The EE cluster count difference is expected, was not expecting it to be so large, but I can tell you what the cause is.
The cut to retain a supercluster is based on the final superCluster.energy(), which changed since you are applying the new regression. In addition, since there is now a 4 GeV minimum pT cut in the mustache regression training, the training sample is significantly more clean from pileup and gives a better response estimate.

[slava77]

+1

tested 7db3dda in CMSSW_7_0_X_2014-01-28-1400/sign301

Changes are as expected
See comments above for more details

[cmsbuild]

This pull request is fully signed and it will be integrated in one of the next CMSSW_7_0_X IBs unless changes (tests are also fine). @ktf can you please take care of it?

[slava77]

@ktf @davidlange6
Giulio and David,
this is the egamma PR for 700 (pre13), mentioned in the planning meeting yesterday.
It uses ES prefer.
After it's integrated and after the GT is updated, we can get rid of the ESprefer for these payloads

[ktf]

Ok, merging this.