Ntupliser fails on 2016 MC

Using !216 but I don't think it's related. I get the failure below for pre and post VFP. The same script works for 2017_UL and 2018_UL.

Command:

das-cmssw el8 cmsRun "$CORE_BASE/Ntupliser/python/Ntupliser_cfg.py" \
      inputFiles="file:/eos/cms/store/group/cat/datasets/recipes/2016preVFP_UL/TTTo2L2Nu_10k_mini.root" \
      outputFile="file:M0_mkNtuples.root" maxEvents=-1 configFile=MetaInfo.json

Config:

{
    "preseed": 1739751423,
    "git": {
        "commit": "22c6c808f60d7d5",
        "repo": "/afs/cern.ch/user/l/lmoureau/work/CAT/analysis_recipes/frameworks/dasanalysissystem/DasAnalysisSystem/Darwin",
        "reproducible": true,
        "complete": false
    },
    "software": {
        "Boost": "1_82",
        "Cpp": "C++20",
        "Darwin": "22c6c808f60d7d5ab872aea90d4aaa2655f97309",
        "ROOT": "6.28/04",
        "gpp": "13.1.0",
        "libgit2": "1.1.1"
    },
    "flags": {
        "isMC": true,
        "R": 4,
        "options": [
            "muons",
            "flavour",
            "jets",
            "photons",
            "triggers"
        ],
        "year": 2016,
        "labels": [
            "HIPM",
            "CHS"
        ]
    }
}

Output:

106X_mcRun2_asymptotic_preVFP_v11
No MET filters
**************************************************************
b tagging needs to be run on uncorrected jets. Hence, the JECs
will first be undone for 'updatedPatJetsRetrained' and then applied to
'updatedPatJetsTransientCorrectedRetrained'.
**************************************************************
%MSG-w CondDBESSource:  (NoModuleName) 17-Feb-2025 01:17:10 CET pre-events
Plugin for Record PGeometricDetExtraRcd has not been found.
%MSG
17-Feb-2025 01:17:10 CET  Initiating request to open file file:/eos/cms/store/group/cat/datasets/recipes/2016preVFP_UL/TTTo2L2Nu_10k_mini.root
17-Feb-2025 01:17:16 CET  Successfully opened file file:/eos/cms/store/group/cat/datasets/recipes/2016preVFP_UL/TTTo2L2Nu_10k_mini.root
/afs/cern.ch/work/l/lmoureau/CAT/analysis_recipes/frameworks/dasanalysissystem/DasAnalysisSystem/CMSSW_12_4_0/src/Core/Ntupliser/plugins/Ntupliser.cc:Ntupliser
/afs/cern.ch/work/l/lmoureau/CAT/analysis_recipes/frameworks/dasanalysissystem/DasAnalysisSystem/CMSSW_12_4_0/src/Core/Ntupliser/plugins/Ntupliser.cc:beginJob
Rivet.AnalysisHandler: INFO  Using named weights
%MSG-w L3Absolute not found:  PATJetUpdater:updatedPatJetsRetrained  17-Feb-2025 01:17:41 CET Run: 1 Event: 1194001
L2L3Residual and L3Absolute are not part of the jetCorrFactors
of module patJetCorrFactorsRetrained. Jets will remain uncorrected.
%MSG
%MSG-w L3Absolute not found:   PATJetUpdater:updatedPatJetsTransientCorrectedRetrained  17-Feb-2025 01:17:41 CET Run: 1 Event: 1194001
L2L3Residual and L3Absolute are not part of the jetCorrFactors
of module patJetCorrFactorsTransientCorrectedRetrained. Jets will remain uncorrected.
%MSG
----- Begin Fatal Exception 17-Feb-2025 01:17:42 CET-----------------------
An exception of category 'Ntupliser' occurred while
   [0] Processing  Event run: 1 lumi: 1195 event: 1194001 stream: 0
   [1] Running path 'p'
   [2] Calling method for module Ntupliser/'ntupliser'
Exception Message:
No dedicated collection exists for HLTriggerFinalPath
----- End Fatal Exception -------------------------------------------------
    687.069813      0.517409      1.000000        322200
 
17-Feb-2025 01:17:42 CET  Closed file file:/eos/cms/store/group/cat/datasets/recipes/2016preVFP_UL/TTTo2L2Nu_10k_mini.root

assigned to @paconnor

added changepatch statusto do typesoftware labels

changed milestone to %Zμμγ workshop Feb 2025

I have encountered the same error before. The cause of the error was using a trigger that was not in the trigger menu corresponding to the current dataset. I’m not sure about the situation with using triggers for MC or using your own ROOT file, but I think you can check if the trigger list provided in the Core/Ntupliser/python/Ntupliser_cfg.py can be used.

added DasAnalysisSystem#3 as parent epic

mentioned in merge request !233 (merged)

@lmoureau You can try again with !233 (merged). In general, we need to externalise the list of input triggers from Ntupliser_cfg.py (a common fix should be found for the present issue and for #134).

https://gitlab.cern.ch/cms-analysis/general/analysis_recipes/-/pipelines/11621732

----- Begin Fatal Exception 03-Apr-2025 08:22:36 UTC-----------------------
An exception of category 'Ntupliser' occurred while
   [0] Processing  Event run: 1 lumi: 28822 event: 28821008 stream: 0
   [1] Running path 'p'
   [2] Calling method for module Ntupliser/'ntupliser'
Exception Message:
HLT_Mu17_TrkIsoVVL_Mu8_TrkIsoVVL_DZ_Mass8_v could not be found in the trigger menu:
HLTriggerFirstPath HLT_AK8PFJet360_TrimMass30_v7 HLT_AK8PFJet400_TrimMass30_v1 HLT_AK8PFHT750_TrimMass50_v1 .........

The correct combination is HLT_Mu17_TrkIsoVVL_Mu8_TrkIsoVVL(_DZ)_v* OR HLT_Mu17_TrkIsoVVL_TkMu8_TrkIsoVVL(_DZ)_v* OR HLT_TkMu17_TrkIsoVVL_TkMu8_TrkIsoVVL_(DZ)_v* (https://twiki.cern.ch/twiki/bin/view/CMS/MuonHLT2016#Recommended_trigger_paths_for_20)

added statusongoing label and removed statusto do label

Affects cms-analysis/general/analysis_recipes!21 (merged)

changed milestone to %N-tupliser configuration

I'm looking at this, and I think deleting this block would make things much easier:

        if (name == "HLTriggerFinalPath"){
            string triggermenu;
            for (unsigned int itrig = 0; itrig < triggerResults->size(); ++itrig) triggermenu += names.triggerName(itrig) + " ";
            throw cms::Exception("Ntupliser") << p.triggerNames_[k] << " could not be found in the trigger menu:\n" << triggermenu;
        }

Then we plug the full list of triggers for all eras into triggerNames_ and they always appear at the same index -- but always false when not present in the input.