Cut optimisation
Strategy
We aim to maximize the total statistical power for measurement of sin(2beta_s) following the
talk under these constraints
- avoid cuts that will directly distort proper time acceptance, such as muon/kaon IP cuts, opening angle cut, B IP significance cut.
- avoid cuts that will significantly distort angular acceptance, such as Kaon Pt cut and strong muon Pt cut.
- provide enough KK mass range (up to 1.1 GeV) for S-P interference study
- provide enough B mass side band for background study
Optimization will not be performed for varibales that will be used in data fit model: t, sigma_t, helicity angles, tag decision, per event mistag, Bs mass, Jpsi mass and phi mass.
Tools
Crop documentation (a little outdated) can be found
here and the latest version of crop is available in tupletools from (on the departmental computers) /phys/linux/s0127440/public/tupletools
Work done and work to do
Data ntuples have been produced for signal, inclusive bb events. Inclusive Jpsi events and B->JpsiX events are being processed.
Plan to optimize using Inclusive Jpsi events and B->JpsiX events. How to use the inclusive events is a problem due to its small statistics. Techniques like fast smearing should be investigated.
Need to include event local purity in the event weight. For simplicity we can chose to consider the dependence of local purity on proper time only and ignore other observables in fit model. This is necessary for correct treatment of prompt background. If this works, then consider the dependence on Bs mass, Jpsi mass and phi mass.
One idea to do this is to parameterize distribution of each of these variables for signal and each background component . Further assume different variables are not correlated. Then the local purity can be worked out using the number of signals and number of events of each background component. For instance, if the number of signals, prompt background events and number of long lived background events are known, and the local purity can be computed.