Added decfiles for cascade channels dangerous to B->Kee for RK at high q2
Decfiles for an RK measurement at high q2, these channels are the most dangerous cascade backgrounds for the electron channel.
Added the following decay files. Each has a cut on the invariant mass of the di-lepton candidates (some miss IDs), just to make the production more efficient for our purposes.
B0_Denu,K*enu,Kpi=DecProdCut,TightCut.dec
Bu_D*0enu,D0pi0,Kenu=DecProdCut,TightCut.dec
Bu_D*0pi,D0pi0,Kenu=DecProdCut,TightCut.dec
Bu_D0K,Kenu=DecProdCut,TightCut.dec
Bu_D0enu,Kenu=DecProdCut,TightCut.dec
Bu_D0pi,Kenu=DecProdCut,TightCut,2.dec
Merge request reports
Activity
added 1 commit
- d562ca0f - lowering cut to identify if the strict cut is breaking pipeline
Hey Michal (@kreps),
I have these 6 decfiles for some dangerous charm cascade backgrounds to B->Kee for the measurement of RK at high q^2.
The measurement will be performed in the q^2_{noBrem}>14 region and so I am only interested in the highest q^2 events from these channels. Therefore, I have developed decfiles for this request which include a generator level q^2 cut (actually a cut on q=GMASS('e-'==GID,'e+'==GID)) – as I only want to fully process events that will be useful. This q^2 cut is extremely efficient on some of these channels. So much so that Gauss takes a significant amount of time to generate an event that passes the cut.
I hope to use ReDecay to improve the efficiency here: I plan to ask for a redecay factor such that roughly 1 redecay passes the generator level cuts from each underlying event. However, for now, the q^2 cut causes the decfiles to fail the gitlab pipeline tests (which have timeouts).
How is best to proceed? I can loosen the q^2 cut which will bring down the simulation time per event but make the request less efficient for our direct needs, or I can proceed with the full cut with your permission?
(the most recent commit has a very loose q^2 cut in the decfiles, modified to demonstrate the decfiles are working correctly)
Cheers, Alex
What you want here is EvtGenDecayWithCutTool, which will redecay your signal until it passes cut without touching the rest of the event. In this cut, you will do a mass cut and then you add a normal cut tool to do the rest of the cuts which are not Lorentz invariant (like acceptance). There are some decay files already using this, so have a look.
added 1 commit
- d060759d - correct 12183005 LoKi cut tool selects particle(s) that it does not apply cuts to
-
Thanks for your advice Michal (@kreps), I am now using EvtGenDecayWithCutTool. Everything is now working, the decfiles are now fast and I believe they are ready to be merged.
Thanks a lot
mentioned in commit 542dcff1