Test AdaBoost conversion to GBRForest

[lgray]

Move GsfElectronAlgo / Producers to use edm::GlobalCache object + GBRForest.

This pull request features a few AdaBoost BDTs that are converted to GBRForest, using the functionality from #10275.

No physics changes expected beyond jitter.

cms-data PR here: cms-data/RecoEgamma-ElectronIdentification#3

[cmsbuild]

A new Pull Request was created by @lgray (Lindsey Gray) for CMSSW_7_6_X.

Test AdaBoost conversion to GBRForest

It involves the following packages:

CondFormats/EgammaObjects
RecoEgamma/EgammaElectronAlgos
RecoEgamma/EgammaElectronProducers
RecoEgamma/ElectronIdentification
RecoParticleFlow/PFClusterTools
RecoParticleFlow/PFProducer
RecoParticleFlow/PFTracking

@diguida, @cvuosalo, @cerminar, @cmsbuild, @slava77, @ggovi, @mmusich can you please review it and eventually sign? Thanks.
@apfeiffer1, @Sam-Harper, @mmarionncern, @bachtis this is something you requested to watch as well.
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.
If you are a L2 or a release manager you can ask for tests by saying 'please test' in the first line of a comment.
@Degano you are the release manager for this.
You can merge this pull request by typing 'merge' in the first line of your comment.

[bendavid]

are you sure you have GetClassifier vs GetAdaBoostClassifier in the right places?

[bendavid]

please test

[cmsbuild]

The tests are being triggered in jenkins.

[cmsbuild]

-1
Tested at: 39126d5
When I ran the RelVals I found an error in the following worklfows:
1306.0 step3

runTheMatrix-results/1306.0_SingleMuPt1_UP15+SingleMuPt1_UP15+DIGIUP15+RECOUP15+HARVESTUP15+MINIAODMCUP15/step3_SingleMuPt1_UP15+SingleMuPt1_UP15+DIGIUP15+RECOUP15+HARVESTUP15+MINIAODMCUP15.log

you can see the results of the tests here:
https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-10327/6437/summary.html

[cvuosalo]

Memory use measurement in progress...

[cvuosalo]

+1

For #10327 d6b8b0f

Converting GsfElectronAlgo / Producers to use edm::GlobalCache object + GBRForest. There should be no significant change in monitored quantities, though there may be some jitter.

The code changes are satisfactory, and Jenkins tests against baseline CMSSW_7_6_X_2015-07-27-2300 show no significant differences, as expected.

Memory usage for workflow 25202.0_TTbar_13 with 70 events and no DQM or Validation against the same baseline shows a 70 MB reduction in RSS:

Baseline
Max VSIZ 1868.8 on evt 57 ; max RSS 1631.89 on evt 57
Time av 15.0774 s/evt max 37.3317 s on evt 57
M1 Time av 14.7675 s/evt max 37.3317 s on evt 57

PR 10327
Max VSIZ 1849.55 on evt 57 ; max RSS 1559.58 on evt 57
Time av 14.9711 s/evt max 41.7561 s on evt 1
M1 Time av 14.5829 s/evt max 36.7071 s on evt 57

[cmsbuild]

This pull request is fully signed and it will be integrated in one of the next CMSSW_7_6_X IBs (tests are also fine). This pull request requires discussion in the ORP meeting before it's merged. @davidlange6, @Degano, @smuzaffar

[davidlange6]

+1