I have two histograms that should be very similar (typically a data/simulation plot). At first sight, the simulation distribution seems to have a shift difference with the data