Ds -> f0 pi
For the 1-c fit, my cuts are:
- Mfit=2 for each pion
- Vertex : 0 < R < 0.006
- Vertex : |Z| < 0.09
- |Mpipi-0.97| < 0.02 This has changed from 0.04, and reduces background.
- L(pi)/{L(e)+L(pi)+L(K)} > 0.5 for the 2 pions from the f0
- L(pi)/{L(e)+L(pi)+L(K)} > 0.6 for the pion from the Ds.
- PROB(chi2, 1)>0.2
The likelihoods L(*) comes from Mike Kelsey's package, with the
consistency cut for the combined likelihood at 0.001.
I would have used the same vertex cut as in the other modes,
- Vertex : 0 < R < 0.01
- Vertex : |Z| < 0.12
but the tighter cuts suppress more background.
The cuts were optimized for the 2-c fit, and may not yield a good result for
the 1-c fit. Here is the 1-c fitmass with different bin sizes:
The background seems to fall linearly in the vicinity of the peak. I count 22
Ds above background. On the other hand, the 2-c fit indicates 37 Ds:
