Program:      GAUSSCUBE

Purpose:      Making the full datacube of the results of a gaussfit.
              

Category:     MODELS, FITTING

File:         gausscube.f

Author:       F.J. Sicking

Keywords:

   INSET=     The set that contains the data. It is used to get the
              coordinate system. The axis along which the subsets are
              specified is used as 'gaussfit' axis.

   PARSET=    The set containing the parameters found by the gaussfit.
              It must be a set with the subsets defined along a 
              parameter axis. The parameters are the baselevel (may be
              absent), the normalization, the mean and the width of the
              gaussian components. The subsets of this set must have the
              same coordinate system and size as the subsets of inset.
              The number of subsets determines the number of gauss
              components and if the baselevel was fitted. As number of
              components, the integer division of the number of subsets and
              three is used. The remainder of this division indicates
              if baselevels are present or not. If it is one, they are, if
              zero, they are not, and otherwise an error message is returned. 
              The subsets are used as follows:
              (subset 1 :  baselevel. (It does not need to be in PARSET.
                           If it is not the baselevels are assumed to be
                           zero.))
              subset 2 :   normalization of the first component.
                           (If the baselevels are not present the
                           normalization will be taken from subset 1 etc.)
              subset 3 :   mean value of the first component.
              subset 4 :   width of the first component.
              subset 5 :   normalization of the second component. Etc.

   BOX=       Area of operation. [entire subset]

   OUTSET=    The set the routine uses for the output, the model datacube.

   AMPLITUDE= Specifies whether the normalization is the amplitude [y] or the
              intensity (n). (amplitude=intensity/(sqrt(2*pi)*sigma))

   FWHM=      Specifies if the width is the full width half maximum [y] or the
              dispersion (n). (FWHM=2.35482*sigma).

Description:  The general idea is making a data cube of a function that
              depends on: the physical coordinate along which the subsets
              are defined, the datavalues and npar parameters which may be
              dependent on position in the subsets. The most common case,
              for which this routine is made, is making the full data cube
              of the results of a gaussfit. This datacube then can be
              subtracted from the original data to get the residuals which
              can be studied to see if the gaussfit was succesful.

Updates:      Jul 23, 1992: FJS, Document created.
              Dec  1, 1992: VOG, Category added.
              Sep 10, 1993: FJS, Box option implemented.