When performing a 'fit spline' what does the tolerance number actually
>represent. Sometimes when I look at the preview the fit spline will almost
>exactly match the curve, but the tolerance will be large (like 1mm). Anyway
>Im just curious what 'mathmatically speaking' is going on here.
I'd guess a least squares regression.
See:
formatting link
files under "15 Modeling of Data" at
formatting link
HTH