Correct for TrackGhostProbability - Githubissues

umd-lhcb / lhcb-ntuples-gen

ntuples generation with DaVinci and in-house offline components

BSD 2-Clause "Simplified" License

1 stars 0 forks source link

Correct for TrackGhostProbability #112

Open yipengsun opened 2 years ago

yipengsun commented 2 years ago

As we discussed before, the run 2 data/MC TrackGhostProb have different distributions.

We are wondering how to correct for this inconsistency. The most simple thing is:

Use trackcalib2 (once it lands) and apply the efficiency ratio as a weight
- This shifts the distribution, so we anticipate a large RMS on the weights, thus killing stats on our MC

Review

As shown in Manuel's slides: Run2_tracking_problems

Other analyses

For Zishuo, his signal mode is B -> J/psi (-> mu mu) mu, and he has NO TrackGhostProb cut.

manuelfs commented 2 years ago

These were the problems that we reported in our talk to the semileptonic group

Run2_tracking_problems