Skip to content

Drastic reduction of efficiency of muonic lines by using muon PID

For your information @mvesteri @poluekt @cmarinbe @decianm :

@lecarus reported an important issue during the last RD Run 3 migration meeting. Since the latest changes this week, he sees by using the muon lines in !1224 (merged) a reduction of the signal efficiency of muonic modes by 85%. So, even loose requireements on the muon PID (PID_MU > 0) have an enormous impact on the efficiency. He could manifest a reduction of 50% by using the ISMUON functor. ( #403 (closed) ) The electronic modes are not impacted. Therefore, the cause are the recent changes this week of muons, I guess they are coming from Rec!2605 (merged) .

@tfulghes crosschecked the behavior of the muon PID with the line Hlt2RD_BdToKstTauMu_KstToKPi_TauTo3Pi_OS_Line from !1399 (merged) using B->K* tau mu events. There, a muon PID of F.PID_MU>2 and the ISMUON functor is required. Before updating the stack he got

Hlt2RD_BdToKstTauMu_KstToKPi_TauTo3Pi_OS_LineDecision
    #=30000   Sum=209         Eff=|(0.6966667 +- 0.0480213)%|
VoidFilter/require_pvs                     #=30000   Sum=30000       Eff=|( 100.0000 +- 0.00000 )%|
TwoBodyCombiner/dilepton_from_tauls        #=30000   Sum=1338        Eff=|( 4.460000 +- 0.119179)%|
TwoBodyCombiner/rd_detached_kstar0s        #=2854    Sum=2085        Eff=|( 73.05536 +- 0.830492)%|
TwoBodyCombiner/beauty_to_xtaul#2           #=956     Sum=209         Eff=|( 21.86192 +- 1.33674 )%|

After updating his stack, he got

Hlt2RD_BdToKstTauMu_KstToKPi_TauTo3Pi_OS_LineDecision                        
    #=10000   Sum=0           Eff=|( 0.000000 +- 0.00000 )%|
VoidFilter/require_pvs                #=10000   Sum=10000       Eff=|( 100.0000 +- 0.00000 )%|
TwoBodyCombiner/dilepton_from_tauls   #=10000   Sum=0           Eff=|( 0.000000 +- 0.00000 )%|
TwoBodyCombiner/rd_detached_kstar0s   #=0       Sum=0           Eff=|(-100.0000 +- -100.000)%|
TwoBodyCombiner/beauty_to_xtaul       #=0       Sum=0           Eff=|(-100.0000 +- -100.000)%|

Probably similar behavior can be seen in other WGs. We think it is worth it to investigate this further. How should be proceed ? @htilquin and @msaur such that you stay informed, too.

Edited by Felicia Volle