1. Preparing samples

•  samples should be prepared using frames from the same night which will be tested in order to have best results
•  using picalib program first generate list of start with catlogue magnitudos and later use getpart2 program to make fitc files
   containing little images of selected stars ( later refered to as samples ). Program getpart2 must be executed with option -with_radec
  in order to generate samples with all nessesary information stored in fits header and list of stars must have the following format :
                   X     Y        RA                    DEC      MAG
                  1232 31 19h11m06.77s -15.88688110 7.68
  this means output file from picalib program must be parsed with awk and proper columns must be retrived to list file
• in order to clean pixels not belonging to main sample cluster use the following options for getpart2 program :
                   -clear_not_cluster=3
  value 3 represents treshold ( 3 sigma above bacground level )
• samples should be placed in the following directory structure :
                   /opt/pi/dev/pisys/daq/ndir/data/Samples
                  [msok@msoklap Samples]$ ls
                   k2a  k2b
• set parameter CCD_IMAGES_SAMPLE_DIR in ccd_pipeline0.cfg and ccd_pipeline1.cfg files to different values :
                   ccd_pipeline0.cfg - CCD_IMAGES_SAMPLE_DIR=/opt/pi/dev/pisys/daq/ndir/data/Samples/k2a
                   ccd_pipeline1.cfg - CCD_IMAGES_SAMPLE_DIR=/opt/pi/dev/pisys/daq/ndir/data/Samples/k2b
• in subdirectories k2a and k2b there should be file names list ( same as value of CCD_SAMPLE_LIST_BASENAME in your cfg file )
• file list should contain the following information ( just example ) :
                    msok@msoklap k2a]$ cat list
                    8.00 8.00/list
                    9.00 9.00/list
                    7.52 7.52/list
  and this says that samples of magnitudo 8.00 are stored in subdirectory 8.00 of directory k2a , and list of fits files for this sample
  is in files 8.00/list
• ready scripts for this purpose is le_sample! ( or le_analyse! )
  
  

2.  Running tests of efficiency

• by default seed is constant for all tests, if one wants to change it parameter :
                       CCD_CHANGE_SEED=1
  in such case seed will be taken from current time, currently there is no possibility to set seed specific seed value
• in order to set specific seed , set :
                        CCD_EXTERNAL_SEED=434637437
• in order to specify (X,Y) positions for samples to be put on image in file , set parameter :
                        CCD_SAMPLE_XY_FROM_LIST=sample_xy.txt
  and fill sample_xy.txt file in the following way :
                        270 134
                        1230 780
  where first column is X and second is Y, such file can be generated with program genxy
  
• in case sample is put with (X,Y) check and frame check , it is possible that for random position (x,y) no sample will be find , in such case program retries with nRetry times with new random position
  default is 5 and can be changed with parameter CCD_SAMPLE_PUT_RETRY, if set to 1 - means no sample will be put if not succeeded at first random position
  

3. Sampe matching procedure in Nparamstest / Nparamstest2C :

• sample is put on random position (X,Y), this information is written to genevents_0.log / genevents_1.log
  
• Lap_New := MAX( Laplace of pixels in sample )
  
• Lap_New and Lap_prev is filled in the following way , all pixels in square +-R pixels from (X,Y) are assumed as belonging to sample
• in case any of pixels in +-R around (X,Y) is has Laplace > Lap_New then Lap_New := Laplace
• Lap_prev is filled as value of last pixels from +-R around (X,Y) which passed Tn cut, so in fact this is the one with max(y) - which was analysied as the last one
• later descicion if sample was identified or not is undertaken according to check if any of identified events fullfiles this contidion :
                       ( X_event - X ) <= R AND ( Y_event - Y ) <= R
  such event is assumed as generated and added to list of identified generated events
  

<>