• The idea is to look for another channel for the D⁺ : D⁺ --> K⁰_spi⁺

• The main difference/advantage with D⁰->Kpi is that we can have a clear sample of K⁰_s and somehow avoid the Pid identification that leads to background

First we have to reconstruct K⁰_s : previous studies (here and here) using TCFIT clearly show that a cut on the signed decay length significance of (pipi) pairs remove background (as opposed as a cut on length only) 

• cuts for K0s

Fig 1 : all (pipi) pairs w/o cuts

Fig 2 : same (pipi) pairs with cuts based on signed decay length and track significance 

S_L = decay length / error of decay length

S_t = dca_XY / error of dca_XY

Blue and green curves are ~ the same means dca_XY(Pion)*dca_XY(Kaon) has no effect after S_L and S_t cuts

• update : 06/17 : rerun with loose cuts to determine the best [unbiased] cuts to apply for quick/efficient K⁰_s reco.

This means , no cut on :

1. silicon hits
2. S_L and S_t
3. probability, chi² of Fit

 The cuts applied are :

1. triggerId = 200001,200003 and 200013
2. first vertex only
3. |ZVertex|<10 cm and its resolution sigma_ZVERTEX < 0.1 cm
4. Loop over primary tracks then take the global associated
5. NFit/NPoss > 0.51 (from TPC)
6. |eta|<1.2
7. dEdxTrackLength >40cm
8. |nsigma_pion|<2
9. 0.2 < M_pipi < 0.8
10. |y_pipi| <1
11. sum of Pt daughters > 1 GeV/c

• results

from left to right : no cut ; cuts based on S_t of tracks (used to removed tracks from PV) ; + cut based on S_L

The conclusion is that there is no peak seen ; the reason is explained with the following plots

It shows the probability of the fit ; there is a *huge* peak below Probability < 0.0001 (I used probability > 0.01 for D⁰)

Then redoing the same exercise with a cut on probability > 0.0001 [and 0.01] (and still trying to keep other cuts as low as possible to avoid bias)

Plots with loose cuts (in the middle) do not show again a invariant mass peak ; the next try was to increase S_t from 1 to 3 

we see that for probability > 0.0001 and St > 3 (upper middle), a tiny peak can be seen --> this set of cut will be the default to run over the run 7