andrey-popov / PEC-tuples

CMSSW package to save basic physics objects in a simple ROOT format
1 stars 3 forks source link

Check if fix in ShiftedJetProducerT is still needed #79

Closed andrey-popov closed 8 years ago

andrey-popov commented 8 years ago

With the recent changes in MET in 76X, the fix of ShiftedJetProducerT is probably not needed. In any case, need to make sure that there are no merge commits. Links to two versions of the affected file: mine with the fix, from MET folks.

andrey-popov commented 8 years ago

The problem is still there. In a clean set-up without my fix get the following error:

----- Begin Fatal Exception 08-Apr-2016 16:20:31 CEST-----------------------
An exception of category 'BadToken' occurred while
   [0] Processing run: 1 lumi: 27631 event: 5503358
   [1] Running path 'elPath'
   [2] Calling event method for module PATCandViewCountMultiFilter/'countGoodJets'
   [3] Calling produce method for unscheduled module ShiftedPATJetProducer/'analysisPatJetsScaleUp'
Exception Message:
A get using a EDGetToken with the C++ type 'reco::JetCorrector' was made using an uninitialized token.
 Please check that the variable is being initialized from a 'consumes' call.
----- End Fatal Exception -------------------------------------------------
andrey-popov commented 8 years ago

Although the problematic file has been changed a bit by MET updates, my patch still works.

andrey-popov commented 8 years ago

The bug is still there in CMSSW_8_0_17, see here.