Fit should be a computation
Created by: kociak
Fits should be computations for single spectrum. The idea would be the fit to update when changing the interval region and/or when changing the spectrum (assuming it is a "pick" computed spectrum). One direct interest would be to find dynamically the good starting coefficients for further and longer, SI-based, spectra. @yvesauad if you are OK I can make a first trial as I think I know more of less how computations work