This is basically the same issue as #31 but on PhenotypicTarget When building a phenotypic target from a set of Gaussians, the interval between 2 points was computed as 1/nb_points aka 1/sampling when it should have been 1/nb_intervals aka 1/(sampling-1)