Maybe we can set range before the projection to remove the unwanted bins
the concern is that in this way the first bin will be lost, and the pdf is not normalized anymore.
Currently we are using Iterator method to loop over the bins to calculate the NLL, and it corresponds to the same bin in dataset and pdf, so dataset and pdf should be of same dimension. This requires we do the same thing for both dataset and pdf.
Maybe we can set range before the projection to remove the unwanted bins
the concern is that in this way the first bin will be lost, and the pdf is not normalized anymore.
Currently we are using Iterator method to loop over the bins to calculate the NLL, and it corresponds to the same bin in dataset and pdf, so dataset and pdf should be of same dimension. This requires we do the same thing for both dataset and pdf.