% @(#)mfit_roma.hlq	17.1.1.1 (ESO-IPG) 01/25/02 17:17:46 
\se                                                                             
SECTION./ROMAFOT
\es\co                                                                          
MFIT/ROMAFOT                                                26-OCT-1989 RHW
\oc\su                                                                          
MFIT/ROMAFOT frame [int_tab] [thres,sky] [sig,sat,tol,iter] [meth,[beta]]
	[fit_opt] [mean_opt] [mod_file]
	determine characteristics of stellar images by non-linear fitting
\us\pu
Purpose: Determine characteristics of stellar images by non-linear
         fitting. The integral of the PSF is computed over the pixel area.
\up\sy                                                                          
Syntax:  MFIT/ROMAFOT frame [int_tab] [thres,sky] [sig,sat,tol,iter] 
                      [meth,[beta]] [fit_opt] [mean_opt] [mod_file]
\ys\pa                                                                          
         frame =      input image frame
\ap\pa                                                                          
         int_tbl =    name of the intermediate table which contains the 
                      selected objects; default is the frame name with 
                      the extension `INT'.
\ap\pa
         thres,sky =  photometric threshold and sky background. Data below the
                      threshold will be ignored. Default values are those used 
                      in the commands SKY/ROMAFOT and indicated in FIND/ROMAFOT.
\ap\pa
         sig,sat,tol,iter = trial value for sigma (parameter for the width of
                      the Gaussian or Moffet curve), the saturation, the shift
                      tolerance, and the maximum number of iterations for each
                      each window. The defaults for sigma and saturation are 
                      the values used by previous commands (SKY/ROMAFOT); the
                      default shift tolerance is 2 pixels, and for the number of
                      iterations 50.
\ap\pa                                                                          
         meth,beta =  non-linear fitting method to be used. One can choose
                      between two fitting functions:\\
                      GI: a Gaussian function with inverse prop. weighting;\\
                      GU: a Gaussian function with uniform weighting;\\
                      MI,number: Moffat function with inverse proportional
                                 weighting and with beta factor number;\\
                      MU,number: Moffat function with uniform weighting and
                                 with beta factor number.\\
                      The default is MU,4.0.
\ap\pa                                                                          
         fit_opt =    four character logical string consisting of `Y'
                      and `N', indicating which parameters should kept 
                      fixed. The meaning of the string is as follows:\\
                      position 1: fix sigma;\\
                      position 2: fix sky background;\\
                      position 3: fix position;\\
                      position 4: allow tilted plane.\\
                      The default is `YNNN'.
\ap\pa                                                                          
         mean_opt =   character indicating if mean values of previous 
                      windows should be used as trial values for the next.
                      One can choose between:\\
                      S: mean values for sigma should be used;\\
                      B: mean values for sky background should be used;\\
                      A: mean values for sigma and sky background should
                         be used;\\
                      N: No mean values should be used as trials.\\
                      Default is `N'.
\ap\pa
         mod_file =   Data file containing the number of intervals over 
                      each pixel and according to the table created by 
                      MODEL/ROMAFOT. The file is assumed to have extension
                      `.dat'. Default is that PSF function is estimated over
                      a fixed 3x3 array over each pixel.
\ap\no                             
Note:     The difference between FIT/ROMAFOT and MFIT/ROMAFOT is that
          the former compares a given pixel to the value of the PSF at the
          centre of the pixel, the latter computes the integral of the PSF
          over the pixel area. The approximation used by FIT/ROMAFOT is
          good when the pixel size is small compared with the scale
          lenght of the point images.

          All the comments made for FIT/ROMAFOT hold for IFIT/ROMAFOT
\on\exs                                                                         
Examples:                                                                       
\ex
          MFIT/ROMAFOT GCLUST1 GCLINTER 
          Fit the windows in frame GCLUST1, stored in the intermediate table
          GCLINTER, using a maximum of 50 iteration (default). Start from 
          the beginning in the table (default), and use a Moffat function 
          with beta=4 (default). 
\xe \sxe