weixin_40005542
weixin_40005542
2020-11-22 08:28

Use correct beamspot for 2016 MC + update MC tracker alignment scenario for 2016 pre-VFP era

PR description:

The MC that has been used to simulate 2016 data so far assumes a beamspot that is appropriate for describing the beamspot of data taken in 2015 with 25 ns bunch spacing. Previously the 2016 beamspot simulation was never changed to avoid inconsistencies between 2016 MC that had already been generated and any newly generated MC. There is no such consistency requirement for the UL as it will involve the regeneration of all GEN-SIM and so we can use the correct beamspot simulation. See the news from the 12 December 2019 PPD meeting for details.

The global tag updates change the BeamSpotObjectsRcd tag to BeamSpotObjects_Realistic25ns_13TeV2016Collisions_v1_mc for 2016 MC scenarios. The vertex smearing used by 2016 workflows has also been updated for consistency.

It also updates the MC tracker alignment scenario for the 2016 pre-VFP era.

Global tag diffs https://cms-conddb.cern.ch/cmsDbBrowser/diff/Prod/gts/110X_mcRun2_design_v3/110X_mcRun2_design_v4 https://cms-conddb.cern.ch/cmsDbBrowser/diff/Prod/gts/110X_mcRun2_asymptotic_preVFP_v1/110X_mcRun2_asymptotic_preVFP_v3 https://cms-conddb.cern.ch/cmsDbBrowser/diff/Prod/gts/110X_mcRun2_asymptotic_v5/110X_mcRun2_asymptotic_v6 https://cms-conddb.cern.ch/cmsDbBrowser/diff/Prod/gts/110X_mcRun2cosmics_startup_deco_v4/110X_mcRun2cosmics_startup_deco_v5

PR validation:

The beamspot parameters are explicitly listed on slide 5 of the news from the 12 December 2019 PPD meeting together with fitted beamspot parameters from data and MC. In addition, a technical test was performed: runTheMatrix.py -l 7.22,limited --ibeos. Note that no RelVal workflows use either auto:run2_design or auto:run2_mc_l1stage1 (though the addOn tests use auto:run2_mc_l1stage1) so a similar technical test cannot be performed for these GTs.

if this PR is a backport please specify the original PR:

This PR is not a backport but will be backported to 11_0_X and 10_6_X.

该提问来源于开源项目:cms-sw/cmssw

  • 点赞
  • 写回答
  • 关注问题
  • 收藏
  • 复制链接分享
  • 邀请回答

35条回答

  • weixin_39668527 weixin_39668527 5月前

    +1

    点赞 评论 复制链接分享
  • weixin_39554434 weixin_39554434 5月前

    +1

    点赞 评论 复制链接分享
  • weixin_39978863 weixin_39978863 5月前

    +1

    点赞 评论 复制链接分享
  • weixin_39993623 weixin_39993623 5月前

    +upgrade

    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

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

    点赞 评论 复制链接分享
  • weixin_39672979 weixin_39672979 5月前

    +1

    点赞 评论 复制链接分享
  • weixin_39807541 weixin_39807541 5月前

    thanks for this. I have few comments: * indeed there is no RelVal wf which runs auto:run2_mc_l1stage1 but there is one addOn test which consumes it: https://github.com/cms-sw/cmssw/blob/0d3768d63689ca8b98c9a08c206566e5eebf2bdb/Utilities/ReleaseScripts/scripts/addOnTests.py#L92 As the vertex smearing there is NominalCollision2015 which in turn points to : https://github.com/cms-sw/cmssw/blob/001ecd07d2fceae0f173a3ea169831f3171b03bb/Configuration/StandardSequences/python/VtxSmeared.py#L47 i.e. https://github.com/cms-sw/cmssw/blob/001ecd07d2fceae0f173a3ea169831f3171b03bb/IOMC/EventVertexGenerators/python/VtxSmearedParameters_cfi.py#L459-L468 I am wondering if that should not be changed as well. Maybe -Grunewald can clarify what is the preference on the TSG side. * as for auto:run2_design (as it is also not exercised in any RelVal) we might take the occasion to move it to consume a really ideal BeamSpot as it is already done for the upgrade 2017 and 2018 workflows which point to Ideal_Centered_SLHC_v3. The dump is the following: image * the vertex smearing Realistic50ns13TeVCollision is used in combination with the autoCond key run2_mc in at least one unit test (Tau embedding): https://github.com/cms-sw/cmssw/blob/001ecd07d2fceae0f173a3ea169831f3171b03bb/TauAnalysis/MCEmbeddingTools/test/runtests.sh#L37 After this update the consistency in that test will be broken. I am wondering if there are other cases in cmssw. * last but not the least, the Realistic50ns13TeVCollision is used as default key https://github.com/cms-sw/cmssw/blob/001ecd07d2fceae0f173a3ea169831f3171b03bb/Configuration/StandardSequences/python/VtxSmeared.py#L68 which in my understanding is used in a workflwow by the ConfigBuilder in case the Beam Spot is NOT specified, see https://github.com/cms-sw/cmssw/blob/7e27750861f9d80a2254a83af0d1f9f94f7f803f/Configuration/Applications/python/ConfigBuilder.py#L995 I am wondering how many cases consistency will be broken....

    点赞 评论 复制链接分享
  • weixin_40005542 weixin_40005542 5月前

    Thanks for your comments.

    thanks for this. I have few comments:

    * indeed there is no RelVal wf which runs `auto:run2_mc_l1stage1` but there is one `addOn` test which consumes it: https://github.com/cms-sw/cmssw/blob/0d3768d63689ca8b98c9a08c206566e5eebf2bdb/Utilities/ReleaseScripts/scripts/addOnTests.py#L92
        As the vertex smearing there is `NominalCollision2015` which in turn points to : https://github.com/cms-sw/cmssw/blob/001ecd07d2fceae0f173a3ea169831f3171b03bb/Configuration/StandardSequences/python/VtxSmeared.py#L47
       i.e. https://github.com/cms-sw/cmssw/blob/001ecd07d2fceae0f173a3ea169831f3171b03bb/IOMC/EventVertexGenerators/python/VtxSmearedParameters_cfi.py#L459-L468
       I am wondering if that should not be changed as well. Maybe -Grunewald can clarify what is the preference on the TSG side.
    

    In this case, I think that I should revert the change to the auto:run2_mc_l1stage1 GT but I will wait for a reply from -Grunewald.

    * as for `auto:run2_design` (as it is also not exercised in any RelVal) we might take the occasion to move it to consume a really ideal BeamSpot as it is already done for the upgrade 2017 and 2018 workflows which point to `Ideal_Centered_SLHC_v3`. The dump is the following:
      ![image](https://user-images.githubusercontent.com/5082376/70847234-6b0f1400-1e62-11ea-8ca8-fa86b2377757.png)
    

    OK. I will make this change, unless there are any objections.

    * the vertex smearing `Realistic50ns13TeVCollision` is used in combination with the autoCond key `run2_mc` in at least one unit test (Tau embedding):  https://github.com/cms-sw/cmssw/blob/001ecd07d2fceae0f173a3ea169831f3171b03bb/TauAnalysis/MCEmbeddingTools/test/runtests.sh#L37
    

    It looks like this is purely a unit test and so could be changed to use the Realistic25ns13TeV2016Collision vertex smearing for consistency with the other 2016 workflows. Would that be OK?

       After this update the consistency in that test will be broken. I am wondering if there are other cases in cmssw.
    
    * last but not the least, the `Realistic50ns13TeVCollision` is used as default key https://github.com/cms-sw/cmssw/blob/001ecd07d2fceae0f173a3ea169831f3171b03bb/Configuration/StandardSequences/python/VtxSmeared.py#L68
       which in my understanding is used in a workflwow by the ConfigBuilder in case the Beam Spot is **NOT** specified, see https://github.com/cms-sw/cmssw/blob/7e27750861f9d80a2254a83af0d1f9f94f7f803f/Configuration/Applications/python/ConfigBuilder.py#L995
    
      I am wondering how many cases consistency will be broken....
    

    Using a 2015 beamspot does not seem to be a very sensible default from a physics standpoint but if the beamspot is not specified explicitly (hopefully) that means that the choice of beamspot is not particularly relevant for that workflow and only consistency is needed. I will defer to PdmV experts on whether or not the default beamspot should be changed.

    点赞 评论 复制链接分享
  • weixin_39620629 weixin_39620629 5月前

    
    * the vertex smearing <code>Realistic50ns13TeVCollision is used in combination with the autoCond key run2_mc in at least one unit test (Tau embedding):  https://github.com/cms-sw/cmssw/blob/001ecd07d2fceae0f173a3ea169831f3171b03bb/TauAnalysis/MCEmbeddingTools/test/runtests.sh#L37
    

    It looks like this is purely a unit test and so could be changed to use the Realistic25ns13TeV2016Collision vertex smearing for consistency with the other 2016 workflows. Would that be OK?

    I think that it is OK. , could you confirm as embedding expert, please?

    点赞 评论 复制链接分享
  • weixin_39562340 weixin_39562340 5月前

    Yes, this should be okay

    点赞 评论 复制链接分享
  • weixin_39862847 weixin_39862847 5月前

    Hi,

    That specific addOnTest is a fastsim test, so you need to check with fastsim people...

    点赞 评论 复制链接分享
  • weixin_39687192 weixin_39687192 5月前

    do I understand correctly that this PR is needed for the UL reprocessing? So it should be backported into 10_6_X before the final patch is made?

    点赞 评论 复制链接分享
  • weixin_40005542 weixin_40005542 5月前

    Hi,

    That specific addOnTest is a fastsim test, so you need to check with fastsim people...

    There is a fastsim addOn test that uses the auto:run2_mc_l1stage1 GT, which is supposed to represent late-2015 data after the L1 Stage-1 upgrade was installed. (Actually, it has a mix of 2015 and 2016 conditions). This PR initially updated the beamspot and vertex smearing for the late-2015 workflows to use the 2016 beamspot but comment https://github.com/cms-sw/cmssw/pull/28624#issuecomment-565704469 suggested that this update was inappropriate. I intend to revert the change to this workflow unless you suggest otherwise, but please comment.

    点赞 评论 复制链接分享
  • weixin_40005542 weixin_40005542 5月前

    do I understand correctly that this PR is needed for the UL reprocessing? So it should be backported into 10_6_X before the final patch is made?

    Technically, I think that the GT and vertex smearing could be specified by hand, though I think that's error-prone. So it's probably better to have it backported.

    There is another update, a tracker MC APE update for the 2016 pre-VFP era that was approved at the AlCa meeting yesterday. Shall I add the APE update to this PR as otherwise the APE update would conflict with changes in this PR? Since the beamspot has been changed, the PR comparison tests are not particularly useful in any case.

    点赞 评论 复制链接分享
  • weixin_40005542 weixin_40005542 5月前
    • last but not the least, the Realistic50ns13TeVCollision is used as default key https://github.com/cms-sw/cmssw/blob/001ecd07d2fceae0f173a3ea169831f3171b03bb/Configuration/StandardSequences/python/VtxSmeared.py#L68 which in my understanding is used in a workflwow by the ConfigBuilder in case the Beam Spot is NOT specified, see https://github.com/cms-sw/cmssw/blob/7e27750861f9d80a2254a83af0d1f9f94f7f803f/Configuration/Applications/python/ConfigBuilder.py#L995

    I am wondering how many cases consistency will be broken.... ```

    Using a 2015 beamspot does not seem to be a very sensible default from a physics standpoint but if the beamspot is not specified explicitly (hopefully) that means that the choice of beamspot is not particularly relevant for that workflow and only consistency is needed. I will defer to PdmV experts on whether or not the default beamspot should be changed.

    , Could you comment on whether or not the default vertex smearing should be changed as well?

    点赞 评论 复制链接分享
  • weixin_39668527 weixin_39668527 5月前
    • last but not the least, the Realistic50ns13TeVCollision is used as default key https://github.com/cms-sw/cmssw/blob/001ecd07d2fceae0f173a3ea169831f3171b03bb/Configuration/StandardSequences/python/VtxSmeared.py#L68

    which in my understanding is used in a workflwow by the ConfigBuilder in case the Beam Spot is NOT specified, see https://github.com/cms-sw/cmssw/blob/7e27750861f9d80a2254a83af0d1f9f94f7f803f/Configuration/Applications/python/ConfigBuilder.py#L995

    I am wondering how many cases consistency will be broken....

    Using a 2015 beamspot does not seem to be a very sensible default from a physics standpoint but if the beamspot is not specified explicitly (hopefully) that means that the choice of beamspot is not particularly relevant for that workflow and only consistency is needed. I will defer to PdmV experts on whether or not the default beamspot should be changed.

    , Could you comment on whether or not the default vertex smearing should be changed as well?

    , I would leave the default as it is (not part of our development) and we modify corresponding beamspot for UL2016 inside relval code.

    点赞 评论 复制链接分享
  • weixin_39993623 weixin_39993623 5月前

    I agree with that it's best to isolate this kind of change to where it will be needed and used in practice (i.e. UL).

    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    The code-checks are being triggered in jenkins.

    点赞 评论 复制链接分享
  • weixin_40005542 weixin_40005542 5月前

    The changes in the last update are as follows:

    • fe3f449: Revert the change to auto:run2_mc_l1stage1, as suggested at https://github.com/cms-sw/cmssw/pull/28624#issuecomment-565704469.
    • 02a5e97: Update tau test as agreed at https://github.com/cms-sw/cmssw/pull/28624#issuecomment-565974290 and following discussion
    • 7a0d99e: Use ideal beamspot for design scenario, as suggested at https://github.com/cms-sw/cmssw/pull/28624#issuecomment-565704469.
    • a1be73e: Update tracker alignment for 2016 pre-VFP era. This is unrelated to the main content of this PR but since the change affects only one line and would conflict with this PR, I have included it here. I will update the title of the PR accordingly.
    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    +code-checks

    Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-28624/13211

    • This PR adds an extra 16KB to repository

    • There are other open Pull requests which might conflict with changes you have proposed:

    • File Configuration/AlCa/python/autoCond.py modified in PR(s): #28535
    • File Configuration/PyReleaseValidation/python/relval_steps.py modified in PR(s): #28429, #28214, #28470, #28535
    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    Pull request #28624 was updated. , , , , , , , , , , , , can you please check and sign again.

    点赞 评论 复制链接分享
  • weixin_40005542 weixin_40005542 5月前

    please test workflow 7.22

    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    The tests are being triggered in jenkins. Test Parameters: - MATRIX_EXTRAS = 7.22 https://cmssdt.cern.ch/jenkins/job/ib-run-pr-tests/4052/console Started: 2019/12/18 20:07

    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    +1 Tested at: a1be73e85d0f404351b504aa2f0b0d194567cc94 https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-a414fd/4052/summary.html CMSSW: CMSSW_11_1_X_2019-12-18-1100 SCRAM_ARCH: slc7_amd64_gcc820

    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    Comparison job queued.

    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    Comparison is ready https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-a414fd/4052/summary.html

    comparisons for the following workflows were not done due to missing matrix map: * /data/cmsbld/jenkins/workspace/compare-root-files-short-matrix/results/JR-comparison/PR-a414fd/7.22_Cosmics_UP16+Cosmics_UP16+DIGICOS_UP16+RECOCOS_UP16+ALCACOS_UP16+HARVESTCOS_UP16

    Comparison Summary: - No significant changes to the logs found - Reco comparison results: 6639 differences found in the comparisons - DQMHistoTests: Total files compared: 34 - DQMHistoTests: Total histograms compared: 2813777 - DQMHistoTests: Total failures: 20614 - DQMHistoTests: Total nulls: 0 - DQMHistoTests: Total successes: 2792822 - DQMHistoTests: Total skipped: 341 - DQMHistoTests: Total Missing objects: 0 - DQMHistoSizes: Histogram memory added: 0.654 KiB( 33 files compared) - DQMHistoSizes: changed ( 25202.0 ): 0.654 KiB SiStrip/MechanicalView - Checked 147 log files, 16 edm output root files, 34 DQM output files

    点赞 评论 复制链接分享
  • weixin_40005542 weixin_40005542 5月前

    +1

    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    The code-checks are being triggered in jenkins.

    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    +code-checks

    Logs: https://cmssdt.cern.ch/SDT/code-checks/cms-sw-PR-28624/13159

    • This PR adds an extra 48KB to repository

    • There are other open Pull requests which might conflict with changes you have proposed:

    • File Configuration/AlCa/python/autoCond.py modified in PR(s): #28535
    • File Configuration/PyReleaseValidation/python/relval_steps.py modified in PR(s): #28429, #28214, #28470, #28535
    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    A new Pull Request was created by for master.

    It involves the following packages:

    Configuration/AlCa Configuration/PyReleaseValidation

    , , , , , , , , , can you please review it and eventually sign? Thanks. , -Grunewald, , this is something you requested to watch as well. , , you are the release manager for this.

    cms-bot commands are listed here

    点赞 评论 复制链接分享
  • weixin_40005542 weixin_40005542 5月前

    please test workflow 7.22

    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    The tests are being triggered in jenkins. Test Parameters: - MATRIX_EXTRAS = 7.22 https://cmssdt.cern.ch/jenkins/job/ib-run-pr-tests/3972/console Started: 2019/12/13 22:18

    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    +1 Tested at: 7b0f12773c574b67c3b699292e1bf399ca04d04b https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-a414fd/3972/summary.html CMSSW: CMSSW_11_1_X_2019-12-13-1100 SCRAM_ARCH: slc7_amd64_gcc820

    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    Comparison job queued.

    点赞 评论 复制链接分享
  • weixin_39962889 weixin_39962889 5月前

    Comparison is ready https://cmssdt.cern.ch/SDT/jenkins-artifacts/pull-request-integration/PR-a414fd/3972/summary.html

    comparisons for the following workflows were not done due to missing matrix map: * /data/cmsbld/jenkins/workspace/compare-root-files-short-matrix/results/JR-comparison/PR-a414fd/7.22_Cosmics_UP16+Cosmics_UP16+DIGICOS_UP16+RECOCOS_UP16+ALCACOS_UP16+HARVESTCOS_UP16

    Comparison Summary: - No significant changes to the logs found - Reco comparison results: 6639 differences found in the comparisons - DQMHistoTests: Total files compared: 34 - DQMHistoTests: Total histograms compared: 2798405 - DQMHistoTests: Total failures: 20614 - DQMHistoTests: Total nulls: 0 - DQMHistoTests: Total successes: 2777450 - DQMHistoTests: Total skipped: 341 - DQMHistoTests: Total Missing objects: 0 - DQMHistoSizes: Histogram memory added: 0.654 KiB( 33 files compared) - DQMHistoSizes: changed ( 25202.0 ): 0.654 KiB SiStrip/MechanicalView - Checked 147 log files, 16 edm output root files, 34 DQM output files

    点赞 评论 复制链接分享