I perform an histogram fit of a mass distribution (mES in BaBar).
I have fixed the shape of the background (an Argus function) and want to fit its normalization to an histo of mES but while ignoring the signal peaking region.
The purpose is to count the number of signal Nsig event without having to modelize its mass distribution,
Nsig = NTotal - NArgus
Hence, how to ignore a range for histo fit and get a suitable normalization ?