LowPtElectrons: convert Seed BDTs from XML to ROOT file format

[bainbrid]

PR description:

This PR converts the formats (from XML to ROOT) of two files containing the weights of two BDT models used by the low pT electron seeding step in the reconstruction chain.

The effect is to reduce the file size by a factor two for both files, and to reduce the memory consumption and CPU time when parsing the weights files.

This was done in response to the issues raised here: #34707

A similar change was made previously, but on the weights file for the BDT model used by the low pT electron ID module (as opposed to the seeding step, as above), which solved the issues raised here: #28780

This PR depends on cms-data/RecoEgamma-ElectronIdentification#22

PR validation:

Local tests were made with the wf 1304.182.

if this PR is a backport please specify the original PR and why you need to backport that PR:

No back port required.

[cmsbuild]

+code-checks

Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-34908/24716

• This PR adds an extra 12KB to repository

[cmsbuild]

A new Pull Request was created by @bainbrid for master.

It involves the following packages:

• RecoEgamma/EgammaElectronProducers (reconstruction)

@perrotta, @jpata, @cmsbuild, @slava77 can you please review it and eventually sign? Thanks.
@Sam-Harper, @jainshilpi, @rovere, @lgray, @sobhatta, @lecriste, @afiqaize, @varuns23, @ram1123 this is something you requested to watch as well.
@perrotta, @dpiparo, @qliphy you are the release manager for this.

cms-bot commands are listed here

[perrotta]

test parameters:

• pull_request = LowPtElectrons: convert Seed BDTs from XML to ROOT file format cms-data/RecoEgamma-ElectronIdentification#22

[perrotta]

please test

[bainbrid]

Hi @perrotta, does this require "enable profiling"? (#34633 (comment))

Or do you test first?

[perrotta]

please abort

[perrotta]

Hi @perrotta, does this require "enable profiling"? (#34633 (comment))

Or do you test first?

You are right, thank you

[perrotta]

enable profiling

[perrotta]

please test

[cmsbuild]

+1

Summary: https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-dcc05a/17827/summary.html
COMMIT: b50f20a
CMSSW: CMSSW_12_1_X_2021-08-17-1100/slc7_amd64_gcc900
Additional Tests: PROFILING
User test area: For local testing, you can use /cvmfs/cms-ci.cern.ch/week0/cms-sw/cmssw/34908/17827/install.sh to create a dev area with all the needed externals and cmssw changes.

Comparison Summary

Summary:

• No significant changes to the logs found
• Reco comparison results: 6 differences found in the comparisons
• DQMHistoTests: Total files compared: 39
• DQMHistoTests: Total histograms compared: 3000352
• DQMHistoTests: Total failures: 6
• DQMHistoTests: Total nulls: 0
• DQMHistoTests: Total successes: 3000324
• DQMHistoTests: Total skipped: 22
• DQMHistoTests: Total Missing objects: 0
• DQMHistoSizes: Histogram memory added: 0.0 KiB( 38 files compared)
• Checked 165 log files, 37 edm output root files, 39 DQM output files
• TriggerResults: no differences found

[slava77]

+reconstruction

for #34908 b50f20a

• code changes are in line with the PR description
• jenkins tests pass and comparisons with the baseline show show no differences
• the igprof results in baseline https://cmssdt.cern.ch/SDT/cgi-bin/igprof-navigator/CMSSW_12_1_X_2021-08-17-1100/slc7_amd64_gcc900/profiling/11834.21/igprofCPU_step3 vs this PR https://cmssdt.cern.ch/SDT/cgi-bin/igprof-navigator/CMSSW_12_1_X_2021-08-17-1100/slc7_amd64_gcc900/profiling/11834.21/PR-dcc05a/17827/igprofCPU_step3 show the expected reduction in CPU cost in 11834.21:
  • edm::Maker::makeModule changed from 22.70/5.16% to 6.11/1.46%
  • all of this reduction is in edm::WorkerMaker<LowPtGsfElectronSeedProducer>::makeModule where the call to createGBRForest(edm::FileInPath const&) has reduced from 16.48 to 0.24.
• I expect that also the memory spike during unpacking and parsing the xml files is removed as well (there's no standard profiling measurement to support this claim though)

this PR should be merged together with a cmsdist corresponding to cms-data/RecoEgamma-ElectronIdentification#22

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

[perrotta]

+1

• To go together with LowPtElectrons: convert Seed BDTs from XML to ROOT file format cms-data/RecoEgamma-ElectronIdentification#22