Tracks (All)  
 
 NTRKS
 -----
 Number of tracks reconstructed per event.  The number of entries is the number of events
 analyzed by RecoCert. In Reco p14 this has a peak at 40 tracks. 

 Exchange 04-Aug-2003:
 Question (FOB): Anything else?
 Answer   (HTD): Oh yeah, when we get triggers due to calorimeter noise,
                 such as "Ring-of-fire", these will show up with few tracks 
                 and no vertices. 
 Exchange 10-Apr-2003:
 Question (HTD): Why the 1000 event peak at zero?
 Answer    (DA): Zero or Min-Bias Events make this spike.
 Comment  (HTD): 5% of the events!  I checked run 175135 to see how many events came
                 in on triggers called Min_Bias + Zero-Bias. I found 2.9%. If I
                 include mu1ptxatxx that goes up to 4.2%. 

 Chisq prob
 ----------
 This is calculated by RecoCert from the GTRK refit chi-sq and the number of degrees of
 freedom.  In p13.06 it is peaked at zero because the chi-sq per degree of freedom is
 in the neighborhood of 1.1.  It should be flat if error analysis is correct.

 
 log10(PT)
 ---------
 The log(pT) of all tracks in a log scale plot. It extends from 10^-1 to 10^2.5
 GeV/c.  It seems to peak at about -0.3. 


 PHI at DCA(0,0)
 ---------------
 Phi is normally defined with 0 degrees pointing east. The charged particle 
 tracks are curved so their direction in phi depends on the position along 
 the track. It is plotted here at the DCA (distance-of-closest-approach) to
 the vertex. This plot is expected to be flat.

 Exchange 04-Aug-2003
 Comment  (HTD): At this point we think it has to do with the offset of the
                 beam from (0,0). It might be interesting to see if the
                 shape changes when the beam moves.
 Exchange 11-Apr-2003
 Question (HTD): Why is it not flat?
 Answer    (DA): The CFT will be responsible for some dip from 36-180 degrees.  
                 We hope this effect goes away when we move the CFT timing today.  
                 The other structures look like they might be SMT overlaps ... (the 
                 phi regions where there are 2 pieces of Si in the same layer) but that 
                 is just a guess.
          15-May-2003
 Question  (HM): Could the dip reflect the beam offset?
 

 TANLAM at DCA(0,0)
 ------------------
 Lambda is pi/2-theta. Theta is the angle off the beampipe on the positive-z side.
 Thus tanlam=0 is eta=0.

 Exchange 15-Apr-2003
 Comment  (SC): For axial-only tracks tanlam has been set to some off-scale value.

 Z at DCA(0,0)
 -------------
 The Z-position of the track DCA in centimeters. Negative-Z is towards the north.
 The RMS is roughly 22 cm and the distribution is symmetric in my example plot.

 Exchange 10-Apr-2003
 Comment  (HTD): There is a spike at zero in Reco p13.06 but not in Reco p14.01.

 DCA (best vertex)
 ----------------
 Tracks distance-of-closest-approach.  There are three kinds of DCA. DCA(0,0) is 
 the closest approach to the z-hat axis.  There is a DCA to the primary vertex as 
 chosen by Reco. Then there is the DCA to the closest vertex to the track as 
 determined from dist./unc'y.

 Exchange 10-Apr-2003
 Question (HTD): What are the units?
 Answer    (SC): CM. 

 DCA pull (best vertex)
 ----------------------
 For each track plot DCA/resolution track parameters. Should have a width of 1.0
 for track except those with long-lives. It's sensitive to misalignment and to 
 tracks due to noise.