We want to have different peakshapes for the escape peaks and maybe for other low-statistics peaks. That's why I made the following changes:

1. Peakwise fit function: Calibration peak fit function of th228 lines fit_peaks() now also accepts a vector of fit_func.
2. Fit function stored in fit result
3. I renamed the fit functions, to avoid confusion (too many f_fit , fit_func , ...). If you feel creative and have better suggestions, feel free to write them below.
   Available fit functions:

• gamma_def: "default" gamma peakshape with gaussian signal, low-energy tail, and background (flat + step)
• gamma_tails: default gamma peakshape + high-energy tail
• gamma_bckSlope: default gamma peakshape + linear background slope
• gamma_bckExp: default gamma peakshape + exponential background
• gamma_bckFlat: default gamma peakshape - step background (only flat component!)
• gamma_tails_bckFlat: default gamma peakshape + high-energy tail - step background (only flat component!)

Example: with escape peaks with high-energy tail but w/o Compton step

myfuncs = [fill(:gamma_def, 3)..., :gamma_tails_bckFlat, :gamma_def, :gamma_tails_bckFlat, :gamma_def]
result_fit, report_fit = fit_peaks(result_simple.peakhists, result_simple.peakstats, th228_names; 
                         e_unit=result_simple.unit, calib_type=:th228, fit_func = myfuncs)

@theHenks we could write the fit_func for the th228_names in jldataprod config, as we discussed. Any additional suggestions?
@fhagemann, @verenaaur I don't think that the A/E functions have to be adapted to this change, right?