Re: maximum likelihood fit
Rene Brun (Rene.Brun@cern.ch)
Wed, 19 Aug 1998 09:33:34 +0200
Walerter Opaska wrote:
>
> Hello
>
> I was wondering if someone could explain to me the best way fit a curve in
> Root. I have a 1D histrogram with a non-gaussian or non-exponential
> distribution ( actually a Breit-Wigner distribution) and need to do a
> maximun likelihood fit on the curve. Could someone please give me a hint
> on the easiest way to fit this curve?
>
You can find an example how to fit with a user-defined function
in the Root tutorial
http://root.cern.ch/root/html/examples/myfit.C.html
Simply replace the expression in the function fitf by the
Breit-Wigner formula (a good candidate to be moved into the list
of standard functions!).
The option to use the "Maximum likelihood fit" is currently
not active in TH1::Fit.
Rene Brun