The least-squares goodness-of-fit parameter S for reflection data
               classified as 'observed' (see _reflns.observed_criterion) after
               the final cycle of refinement. Ideally, account should be taken
               of parameters restrained in the least-squares refinement.
               See also the definition of _refine.ls_restrained_S_obs.
                   ( sum|w |Yobs - Ycalc|2| )1/2
               S = ( ---------------------------- )
                   (       Nref - Nparam      )
               Yobs   = the observed coefficients
                          (see _refine.ls_structure_factor_coef)
               Ycalc  = the calculated coefficients
                          (see _refine.ls_structure_factor_coef)
               w        = the least-squares reflection weight
                          [1/(e.s.d. squared)]
               Nref   = the number of reflections used in the refinement
               Nparam = the number of refined parameters
               sum is taken over the specified reflections