Finding the loglikihood

Roofit and RooStats
1

Dear ROOTers,
I have fitted my distribution via Vavilov distribution:
I want to find out the loglikehood value for this distribution.
Please help.
Thanks in advance.

2

Hi,

If you have fitted your histogram using TH1::Fit with option “L” for a Poisson likelihood fit, the log-likelihood value is :

TFitResultPtr result = h1->Fit(myFunction,"L S");
cout << "negative log-likelihood   = " << result->MinFcnValue() << endl;

Lorenzo

3

Thanks :slight_smile:. I got the logliklihood value. But the fit becomes worse after using this “L” option.
Could you tell me why this is so? I have attached the both plot before and after using the “L” option
h->Fit(f1,“L S”);

Thanks in advance.
hitx_19_after.ps (14.4 KB)
hitx_19_before.ps (9.42 KB)

4

Hi,

Your result looks strange to me. Can you please post your histogram and a minimal macro defining the fit function and doing the fit ?

Best Regards

Lorenzo