About 62M events within |Vz|<6cm from official production.

Common cuts
pT1>1.2, pT2>1.2 GeV/c
dca1>0.008, dca2>0.008
dca12<0.005

For HS reconstruction
cos(theta_hs)>0.995
costhetastar_hs<0.8 

For KF fitting,
prob_kf>0.6
costhetastar_kf<0.8 

Helix Swimming	KFParticle fit

For these particular sets of cuts,
HS method gives yield = 673 +/- 73  (9.2 sigma)
KF method gives yield = 624 +/- 61 ( 10.2 sigma)

I am not trying to prove that KF gives a better result. The best performance from each method will still depend on the cuts used. Here I would rather just offer an alternative reconstruction through KFParticle that may provide a cross-check or maybe better result comparing to the conventional HS method.