On 11/10/09 23:04:50, Hans-Bernhard Bröker wrote:
> lyn...@fr... wrote:
>
>> I'm wondering if there is some way to get a kind of fit-error for
acsplines.
>
> No. Nor would it make much sense to try. An acspline isn't a fit ---
> its parameters have no meaning to the actual curve, they don't model
> anything.
Thanks for Your quick answer.
If the acspline paramaters have no meaning to the actual curve I wonder if
there is a kind of measure that tells how big the average difference between
the actual data points to according spots of the acsplines curve is.
Basicly taking the y-value of the acspline curve at each data x-value and
cumputing the difference to the data y-value. I just don't know how to get the
function of the acsplines. Since the computer nows where to plot the curve
points it must know the y-values. The ugliest way to get them would be to
count the y-pixels in the plot...
Is there any better way to do this?
Regards,
Lynx
|
