GSI Forum: Analysis » Probability ditribution of mass constraint fitting in oct14 vs. mar15

Home » PANDA » PandaRoot » Analysis » Probability ditribution of mass constraint fitting in oct14 vs. mar15

Show: Today's Messages :: Polls :: Message Navigator

Probability ditribution of mass constraint fitting in oct14 vs. mar15 [message #18355]

Tue, 07 July 2015 15:25

Lu Cao
Messages: 77
Registered: February 2013

continuous participant

From: *ikp.kfa-juelich.de

Dear all,

I simulated 2k evt of this decay chain:
pbarp->Ds+ Ds-; Ds- -> K+K-pi-; Ds+ -> eta e+ nu_e; eta -> pi+ pi- pi0; pi0->2 gamma.

Inorder to reconstruct eta, I do mass constraint fitting on (pi+, pi- pi0) after the vtx fitting of combined charged daughters. Recently I find the probability distribution of the eta mass constraint fitting seems incorrect, where most of prob are skewed to 1 and chi2 is very small, i.e. the mean value of chi2 is 0.0003. The interesting thing is this doesn't happen in the mass constraint fittings of Ds- (pure charged daughters) and pi0 (pure neutral daughters). This problem is found with the release mar15 (#27529). I test the identical code with release oct14 (#26514), and find this problem doesn't arise there (see attached histograms). The code I used for testing is attached.

Question: Why it happens and how can I fix it?
Many thanks in advance.

Best regards,
Lu
index.php?t=getfile&id=8453&private=0