Search All of the Math Forum:
Views expressed in these public forums are not endorsed by
Drexel University or The Math Forum.
|
|
|
|
Re: lsqcurvefit: What to do against it?
Posted:
Jan 15, 2013 8:51 AM
|
|
> > Set TolFun to a smaller value.
> >
> > Best wishes
> > Torsten.
>
> Dear Torsten,
> that worked, but the fit is still pretty bad.
> I have a curve running from 10^0 down to 10^(-4) on
> the y axis.
> The first 10 data points are fitted well for low
> x-values, but badly for higher x-values.
> I wanted to fit the log of my function, but received
> after ca. 4h calculations times just four misfitted
> point from 299 data points.
> I received during fitting:
>
> Warning: Maximum function count exceeded; singularity
> likely.
>
> What do I do with it?
> Is there a chance to introduce with lsqcurvefit some
> error boundaries for the data or some weighing,
> please?
Can you supply the data points and the fitting function ?
Best wishes
Torsten.
|
|
|
|
