Apologies if this is not the correct forum for my question.
I have a branch in a TTree containing the mass of a diphoton object. Is there a way to find the smallest interval which contains 68% of the events, without performing a fit (i.e not the value of the sigma parameter of the gaussian fit to the distribution)?
On first thought, nothing comes to mind that would do this for you. A first approach could be to calculate a 68% quantile by calculating the upper and lower 16% quantiles. This is not what you want, but a first step. Now you can start to optimise the lower quantile until you get the desired quantile.