[Scilab-users] advice on a pseudo-"bandwidth" calculation

[Samuel Gougeon]

Hi Paul,

I am not sure to well understand the purpose and constrains.
Why not fitting experimental points with a model? -- yes, the gaussian 
law could be a good start, or a lorentzian one, or other, depending on 
the type of observed phenomena.
Then, you would get a set of modeling parameters (max amplitude, mean 
position, standard deviation), and you could compare fitting parameters 
from different sets of points, among which the dispersion of points.
This would be a standard way to proceed.

Samuel