Extended likelihood fit with one yield constant

Roofit and RooStats
1

Hi,

I am trying to perform an extended likelihood fit involving three PDFs. Earlier in my analysis, I have calculated the yeild for one components and I now want to set this constant in the extended fit. The fit should then return me only the yields for the other two PDFs. When I do this, I find that the total yield (sum of yields for the 3 PDFs) is 78.76, whereas the number of events I fit to is 78. Is this something I should be worried about?

If I redo the fit but this time float all three yields then the total yield is exactly 78, as I would expect.

Thanks,
Greig

2

Hi,

Can you post (or mail me) the complete example? I’d have a take a closer look at it.

Wouter