 To refactor: - Simplify the `fitLine` and `curveCalibration` groups to get more or less the 4 plots above - Automate peak detection: get list of emission energies from a database (e.g. pymca) and use estimation of zero and gain - External links to the raw data? - improve the `calibCoeffs` results structure 