To-dos for next production (II)

[swertz]

List of (non-critical) stuff that might go into the next prod:

• Logical OR of single-lepton & di-lepton triggers ==> Not entirely clear what paths to use, and how to separate the primary datasets in orthogonal way. Keep on the list for 80X.
• Include HLT SFs directly into objects (depending on the above, might have to add new SFs to Framework)
• In the categories, separate the "HLT matching" cut in two cuts: one simply tellig if the path fired, and one telling if the selected leptons correspond to that path. I'm wondering if the 10% data/MC discrepancy we see might be related to this object matching (the "official" Top analysis doesn't see it, and I don't see any mention of trigger matching, see AN 2015/309).
• Include generated W info
• Keep list of CSVv2-ordered DiLepDiJets(Met) objects, even if no b-tagging is included, to be able to do stuff by requiring only 1 b-jet, for instance. ==> Remove the distinction jet/b-jet, simply add the "A" (all) b-tag working point. Also remove the tight b-tagging (probably we'll never use it).
• Correct the Z peak mass veto: 91+15=106, not 116...
• Exclude ECAL gap for electrons (1.566 < eta < 1.4442)
• Gen info: do the matching with before FSR leptons (and move the rest of gen info to before-FSR partons)
• Neutrino solver: pass W & top masses on-the-fly
• B-jet & electron eta cut to 2.4

[blinkseb]

Generated W info for the W smearing? It may be hard to add because I'm not sure the W is in the decay history for all the samples. If we consider it is, then we can clean the code checking if the b is coming from the top decay or if it's a radiation.

We can also maybe think of adding the scale factors directly inside the objects in addition to the HLT ones? Not sure it's useful, be it can reduce the size of the cut string :)

Anyway, thanks for the list :)

[swertz]

For the W, I had forgotten about this issue, but we can continue doing what I'm doing right now, which is to add the p4's of the lepton (before FSR) and the neutrino...

What I meant about the HLT is that right now we access the SFs from a file included in the plotter, not from the anlyzer's output, and I don't see an easy way to access them without adding some information inside the TTAnalysis::DiLepton object (but I might be mistaken, I'm not 100% clear on how the HLT SFs are accessed in the framework)...