Allow previous runs of a specific sample to be ignored for statistics

Dear sirs,

i have 2 recommendations (if not implemented on pro version):

The idea is to have the possibility to engage or not previous measured samples, why? Because in certain situations we don't want to consider the several times that the same program runned as different parts with same reference, that is, in the case of an iterative measure and tunning of an assembly we just want to consider the last one for the general statistics...