Test of rerun auau200GeV 200ProductionMinBias sample.

It has some fixes for the /StDaqLib/EVP/ssd_reader.cxx and /StDaqLib/SSD/SSD_Reader.cxx

cuts of primary tracks:

• |z|<10 cm (vertex)
• TPC fitted > 20
• pT > 0.1
• |eta| in SSD acceptance

part I : on 500 files

TotNevent = 167031 Event Analyzed = 118219

Fig.1 : dca xy vs 1/P

Fig.2 : dca z vs. 1/P

Fig.3 Comparison of dca xy resolution with previous production (black : old, red :new) 

Fig.4 Comparison of dca z resolution with previous production (black : old, red :new)

Part II : Origin of the "funky" shape for lower momentum.

The reason is because I used FitSlicesY() in a given range : |dca_{xy}| < 1

The following plot shows the dca_{xy} for tracks with TPc+SSD vs. 1/P

(This plot is for all the tracks)

However, I have to take into account the whole dca range : |dca_{xy}|<3 (as shown in the next plot)

(this plot is for positive tracks only)

it means that up to a given momentum, a fit with a gaussian is not meaningf

to do/investigate :

1. increased tracks quality (Number of fit points, eta)
2. fit double gaussian for |dca_{xy}| <3 

 Part II : cut on chi2 of primaries tracks

The following plot show the dca_{xy} vs the chi2 of primaries tracks (x-axis) for 0,1,2,3 and 4 silicon hits

From this plot, I put a cut on chi^{2} of primaries tracks : chi^{2} < 4.

There are no other cuts added and the existing cuts I used remain the same.

Fig2.a : dca xy resolution vs. 1/P 

Fig2.b : dca z resolution vs.1/P

So it seems that cutting on chi^{2} of primaries tracks take the dca resolution in both direction to the level we expect.