Friday, December 16, 2005
B-Physics步驟
B-Physics Todo:
- 但其中要如何分辦Proton/AProton, Lam/ALam?
- 要如何找出ALam/Lam的child?
要作B-Physics,採用以下步驟:
1. 利用EvtGen,或是QQ98,作出DecayMode(B+--->Proton ALam Pi+ Pi-)的MC events,並將之存成。
2. 利用GsimRec,將這些事件丟入Gsim,模擬成真正粒子會碰撞的狀態,產生mdst檔。
3. 然後用Belle Library,將mdst的內容,選出在Mb-dE的box內,可能的B粒子,並將這些事件,output成hbk,用PAW將之視覺化,看出其分布是否有signal訊號。
* 填入ntuple的方法:
bTuple->column("ProtonEnergy",ProtonIter->e());
bTuple->column("ProtonPx",ProtonIter->px());
bTuple->column("ProtonPy",ProtonIter->py());
bTuple->column("ProtonPz",ProtonIter->pz());
bTuple->column("ProtonP",ProtonIter->mag());
bTuple->column("LamEnergy",LamIter->e());
bTuple->column("LamPx",LamIter->px());
bTuple->column("LamPy",LamIter->py());
bTuple->column("LamPz",LamIter->pz());
bTuple->column("LamP",LamIter->mag());
bTuple->column("PIplsEnergy",PIplsIter->e());
bTuple->column("PIplsPx",PIplsIter->px());
bTuple->column("PIplsPy",PIplsIter->py());
bTuple->column("PIplsPz",PIplsIter->pz());
bTuple->column("PIplsP",PIplsIter->mag());
bTuple->column("PImnsEnergy",PImnsIter->e());
bTuple->column("PImnsPx",PImnsIter->px());
bTuple->column("PImnsPy",PImnsIter->py());
bTuple->column("PImnsPz",PImnsIter->pz());
bTuple->column("PImnsP",PImnsIter->mag());
bTuple->dumpData();
Let me briefly summarize the possible steps in the P L pi pi analysis.
1. particle candidates selection:
a. First, you should select proton from the Mdst_charged bank by
applying atcpid(p-pi) > 0.6 , and atcpid (p-k) >0.6 and save them as a
particle vector.
b. Pion is also from the Mdst_charged bank by applying atcpid(pi-k) > 0.6.
c. For Lambda, you should select it from Mdst_Vee2 bank (kind=2 for
Lambda, kind =3 for anti- Lambda). Specially, you need the goodLambda
cut.
2. Particle re-combined.
Further, you have to combine these four particles to form B meson and
then save all of these particles' information in a ntuple. For
example, the Mb/dE distribution, all daughter particles' (p L pi pi )
four momentums, daughter particles' charge, pidcut you used before,
the shape variables ( for shape variable, there is also a function
named shape in my analysis code and you also have to include the
shape.*)
You can find some clues by refering to my code.
After finishing the analysis code, use it to process your signal Monte
Carlo sample to see the Mb/dE peak from PAW, because if your analysis
code and MC sample are both correct, there should be a peak in both Mb
and dE distribution.