Improvements for TSourceCalibration

[VinzenzBildstein]

This is mainly a note for myself, but if you have other things that you think should get implemented, please feel free to add to this!

• Add option to go back and forth between the view of the individual source spectra and the peaks found in them, and the calibration with the residuals.
• Add option to define separate regions of the spectrum to run the peak finding algorithm over. By allowing to use separate regions to find peaks, the threshold could stay relatively high. This means one could find all the weaker high energy lines without getting bogged down by tons of weak lines in the lower energy part of the spectrum (which is an issue right now).

[VinzenzBildstein]

• Check for any histogram name if a matching source file can be found and use that as default choice.
• In that case (and maybe in general) add check to histogram name to remove any /.
• Maybe check all .sou files and add all to the predefined drop-down menu?

[VinzenzBildstein]

Add at least an option to not have the non-linearities included in the cal file