Re: [Help-gsl] nonlinear fit

[Francesco Abbate]

2011/3/8 Irina Stefan <address@hidden>:
> Hi,
>
> This is probably a very easy question, but I'm not very comfortable with gsl, 
> so here it goes.
> I'm writing a routine to fit a Gaussian to a set of data points. I'm pretty 
> much modifying the example from the gsl manual. However, I want to perform an 
> unweighted fit and I'm not sure what function to use as f_i.

Hi,

this is actually an easy question :-)

If you want to perform an unweighted fit just set f_i to Y_i - y_i,
i.e., to the difference between the obsered value and the predicted
one. Just don't forget to do the same thing for f_i and the Jacobian.

In the example they divides by sigma_i so you have just to omit the
division by sigma_i both for f_i and the jacobian.

-- 
Francesco