Electron IP sign

[swertz]

In #479 the sign of the 2D IP (dxy) for muons was fixed, and if I understand correctly the issue was fixed upstream in PAT (here cms-sw#28753) by switching from using IPTools::signedTransverseImpactParameter() to reco::TrackBase::dxy().

However it seems that PAT electrons still use signedTransverseImpactParameter(): https://github.com/cms-nanoAOD/cmssw/blob/master/PhysicsTools/PatAlgos/plugins/PATElectronProducer.cc#L1150-L1159

Is there a reason why this wasn't changed for electrons?

[peruzzim]

to be discussed offline

[arizzi]

was it discussed?

[swertz]

Didn't have any reply, I'll send a reminder.

[swertz]

Fixed in cms-sw#30738 (+ backport to 106X). Not sure if a fix a la #479 for non-UL is also needed?

[mariadalfonso]

@swertz
Thanks for this check.
indeed a fix like this for the OLD miniAOD is beneficial especially if we go for the re-nano of the UL before doing the re-mini.

[mariadalfonso]

For miniAOD it's taken care starting from the reminiAOD of the UL.
For nano made from previously made mini, the PR linked above fixed the issue.