To run TbyT analysis you do 

1.  Create Directory TbyT_"old"_"new" and link there to "old" and "new" versions

 mkdir -p 2020/TbyT_SL_TFG 
 cd       2020/TbyT_SL_TFG
 ln -s    ../RC_opt_sl73_gcc485_SL  old
 ln -s    ../RC_opt_sl73_gcc485_TFG new 
 
2. Run there
         ~fisyak/bin/MakeTbyT.pl

3. Link/copy  

 ln -s  ~fisyak/macros/RunTbyT.C .

4. Submit jobs 

   star-submit ~fisyak/xml/jobs.xml

5. After jobs finished run

          root.exe ~fisyak/macros/TbyTPlots.C+

6. Copy *.svg (or *.png) files into your WWW directory structure 

       cp *.svg  SL_TFG/2020/

7.

       cp ~fisyak/WWW/star/TbyT/SL_TFG/index.php .

8. Change in index.php 'array' of directories to whatever you want.

9. See results  in your browser. x2
    For example, https://www.star.bnl.gov/~fisyak/star/TbyT/SL_TFG/


================================================================================
StTbyTMaker does match of track from two different (and independent) reconstructions and calculate efficiencies accordingly to 
Frederick James "Statistical Methods in Experimental Physics", 2nd Edition, 2.2.3. Example of the addition law : scannng efficiency. page 14
Let's 
   N1 no. pf events (tracks, ...) found by the first reconstruction
   N2 no. pf events (tracks, ...) found by the second reconstruction
and 
   N12 no. of events found by both recostruction.
Let'ss e1 is effciency of the first reconstruction, 
      e2 is effciency of the second reconstruction, 
If reconstructions are independent then 
    e12 = e1*e2   efficincy to find an event by both reconstructions

Let's N to be total number of reconstructable events in the sample, then

   N1 = e1 * N
   N2 = e2 * N
   N12 = e1*e2 *N

   e1 = N12/N2
   e2 = N12/N1
   N  = N12/(e1*e2)   


1. StTbyTMaker does match hits (clustes) reconstructed in the different reconstructions by proximity (TPC,  closest hits in pad and time bucket space with minimum distance delta_pad**2 + delta_timeucket**2  < 4).

2. StTbyTMaker does match tracks from the different reconstruction by requiring that the half of hits from shortest track is matched with hits from another reconstruction. 

3. Thus for each track we have 3 states:
     1. The only one <b>matched</b> track (reconstructed).
     2. More than 2 match tracks (the first one considered as <b>matched</b>, and all othees called <b>clone</b>).
     3. <b>ghost</b> track is found in this reconstruction.
     4. <b>lost</b>  track is not found in the reconstruction.