search

cp3-llbb / Framework

Common framework for all cp3-llbb analyses
https://github.com/cp3-llbb/cp3-llbb.github.io/wiki/CP3-llbb(met):-Getting-Started
0 stars 13 forks source link

Upgrade to CMSSW 8.0.26p2 #246

Closed blinkseb closed 7 years ago

blinkseb commented 7 years ago

Title says everything. Please do not merge until EGM smearing is updated (the recipe will probably change)

Note: bad muons filter has been removed since user are now expected to run over reMiniAOD

swertz commented 7 years ago

Thanks! I expect the build failed because you're waiting for the new scales to regenerate the ref trees? But the link to jenkins seems dead...

blinkseb commented 7 years ago

The build should not failed since the recipe is basically the same. But:

blinkseb commented 7 years ago

please test

blinkseb commented 7 years ago

please test

blinkseb commented 7 years ago

please test

blinkseb commented 7 years ago

please test

blinkseb commented 7 years ago

please test one more time :man_dancing:

blinkseb commented 7 years ago

please please please test test test :rocket:

blinkseb commented 7 years ago

:champagne: :tada: :balloon:

blinkseb commented 7 years ago

It'd good also to implement recommendations from https://twiki.cern.ch/twiki/bin/view/CMSPublic/ReMiniAOD03Feb2017Notes

swertz commented 7 years ago

Seems like we do have to change the MET input tag if we want to propagate the EGM fix? => slimmedMETsMuEGClean

blinkseb commented 7 years ago

Pff indeed

swertz commented 7 years ago

Ah but we don't really care about that one: jets are not e/gamma corrected => it makes sense to use only the muon-corrected MET, which is the default collection?

swertz commented 7 years ago

For the other things did you mean the It would be extremely advisable to store the bool particleFlowEGammaGSFixed:dupECALClusters and if ecalMultiAndGSGlobalRecHitEB:hitsNotReplaced is empty or not?

blinkseb commented 7 years ago

Yep that ones

blinkseb commented 7 years ago

I'm looking also into updating MET filters

blinkseb commented 7 years ago

please test

I'm trying to see if using webhooks to jenkins works with the new container infrastructure (instead of plain old cron task)

And looks like it's working :+1:

blinkseb commented 7 years ago

The EGMSmearing twiki has been updated with a new recipe I implemented in ab4c4bb :tada:

There's something wrong with the recipes, it pulls all packages of CMSSW instead of just the modified one... investigating

Ok should be fixed, The recipe is indeed buggy (the remote has a merge commit for the regression based on 8.1.X or 9.0.X). Let's wait for the official announcement before reporting :)

swertz commented 7 years ago

Twiki for EGMSmearing has been updated, no need to fetch giuseppe's repo anymore? https://twiki.cern.ch/twiki/bin/view/CMS/EGMSmearer#ECAL_scale_and_resolution_correc

blinkseb commented 7 years ago

Let me try it :sushi:

blinkseb commented 7 years ago

Much better indeed, works out-of-the-box. I'll update the PR Done.

blinkseb commented 7 years ago 

*** Discussion title: Egamma ID

Hi,

No, this is *not* the case because we updated the E-p regression method.
Right now we use a semi-parametric method with different inputs.

I will use the numbers you measured and release a consistent version
later in the day today.

Best,

-- Rafael.

Not sure what this really means, but we should probably wait a bit :worried:

swertz commented 7 years ago

Indeed, what a mess...

OlivierBondu commented 7 years ago

As Hugues was saying earlier: there is 'mess' in 'CMS' after all :smile:

blinkseb commented 7 years ago

PR updated with latest smearing recipe from https://twiki.cern.ch/twiki/bin/viewauth/CMS/EGMRegression#Consistent_EGMSmearer

OlivierBondu commented 7 years ago

many thanks !