In a multiple fit we end up with a table of parameter names all called something like f0.f0.f0.A0, etc, One or more of the “f0” bits refers to the data set (in the future we may be fitting across more than one dimension of runs, periods and groups/spectra). Another one (or more) refers to the different parts of a composite function applied to each spectrum. It’s confusing working out what is what, especially if some parameters, or sometimes some of the “f0” prefixes, are missing because of global ties.
Could we have friendly names such as HIFI23437.LaserOn.Forward.ExpDecay.A0 instead? Depending on the interface it might be reasonable for the user to give names to the runs, e.g. “sample”, “reference”, “background”. Groups and periods usually have names anyway, if none can be found then use “G0”,“G1”,“G2” and “P0”,“P1”, and if there’s more than one of the same function then revert to numbering: “ExpDecay1”,“ExpDecay2”.