I am trying to fit data using Roofit(using Roogaussian and Roochebychev). There is a statistically fluctuating point very close to my signal peak, which is affecting my good fit and estimating the number of signal events. I would like to exclude that particular bin while fitting. Can anyone suggest me how can i go about it?
Did you look at setting a fit range on the RooRealVar? You can have 2 disjoint ranges on the right and left of this bin. That should do the trick (although one may argue of the validity of the test you’re trying to perform or its meaning; I’d be carefull when presenting this result).
Note: if you do an unbinned fit, this “bin” have only an existence when you plot the data but the unbinned data points are used in the fit (so properly speaking it’d not be a bin you’re excluding but a range of values).