Improve plotting of fit results (in data_retrieval)

[153957]

The fit results are continuous functions, so it would be nicer to plot them as line plots.

Currently the x-points for plotting the fit data are the same as the x-points fo the data that was fit..
This can make the fit less clear in a scatter plot if the source data also have very few data points.

So, it would be better to create a range from x-min to x-max with a bunch of steps (several thousands should be more that enough) and then feed those into the fit function.
If that works we can see if it is better to change the plot style of the fit to normal lines, or to keep them as scatter/step..

For example (current results, low density of points at high pulseheights/integrals):