Revisit the definition of leptons

BrieucF commented 9 years ago

[x] Add loose lepton, medium lepton (for electrons), tight lepton collections, ... Rather then adding plenty of new branch in the nTuple, we could keep simply the index pointing to the collection from the central framework, keeping the info if it is an electron or a muon.
[x] It could also be good to keep for each new collection the number of object we have in the event (nLooseLep, etc.). NB : the category definition has to be synchronized with the lepton collection we plan to use, we could maybe base it on loose leptons to begin with (so that we do not kill too many events).
[x] Add also different isolation levels

OlivierBondu commented 9 years ago

@blinkseb : to address the second point, is there a way around with the structs stored from PR #22 ? Or do we need to store these manually ?

we can request hh_leptons.isID_L, or check @hh_leptons.size(), but I could not find a way to get @hh_leptons.isID_L.size()

blinkseb commented 9 years ago

A shot in the dark here. Does hh_leptons.@isID_L.size() work ? Le 8 oct. 2015 19:03, "Olivier" notifications@github.com a écrit :

@blinkseb https://github.com/blinkseb : to address the second point, is there a way around with the structs stored from PR #22 https://github.com/cp3-llbb/HHAnalysis/pull/22 ? Or do we need to store these manually ?

we can request hh_leptons.isID_L, or check @hh_leptons.size(), but I could not find a way to get @hh_leptons.isID_L.size()

— Reply to this email directly or view it on GitHub https://github.com/cp3-llbb/HHAnalysis/issues/18#issuecomment-146624749.

OlivierBondu commented 9 years ago

It does :smile: !

blinkseb commented 9 years ago

\o/

2015-10-09 9:27 GMT+02:00 Olivier notifications@github.com:

It does [image: :smile:] !

— Reply to this email directly or view it on GitHub https://github.com/cp3-llbb/HHAnalysis/issues/18#issuecomment-146778588.

Sébastien Brochet

swertz commented 9 years ago

So, you have a vector of structures, and this gives you the length of the vector which have isID_L to true?

OlivierBondu commented 9 years ago

yep, aka nLeptonLoose in this case

blinkseb commented 9 years ago

I don't think it's true. hh_leptons.@isID_L.size() should returns the size of the isID_L vector, what ever it's inside.

OlivierBondu commented 9 years ago

correct, did not look correctly at the content, yes, hh_leptons.@isID_L.size() works, but return the same as hh_leptons.isID_L:

if we want to loop over the content of the trees, all the info is here
but we don't have access to this info easily for Draw

so we need to store these multiplicities independently. Is my understanding correct ?

blinkseb commented 9 years ago

Ok that's weird here, I'll do some test, but my understanding so far was:

hh_leptons.@isID_L.size() returns the number of entries inside the isID_L array
hh_leptons.isID_L returns the content of the array, so since it's bool, only true or false for each entry.

If you need to access the number of entries with hh_leptons.isID_L == true I don't think there's an easy way to do that in ROOT at Draw time, but maybe there's. Time to restudy the doc :fearful:

OlivierBondu commented 9 years ago

after discussing with @swertz : maybe the easiest and cleanest way would be, in addition to storing hh_leptons.isID_L, to store another vector hh_leptons_L, containing the indices of the loose leptons.

This way we would both have an easy access to multiplicities, and an easy looping access over loose leptons only (not the most optimal for the size, but not the worst either)

cp3-llbb / HHAnalysis

Revisit the definition of leptons #18