Advanced automatic multi-step fitting

[EmCeBeh]

Often, data, which is measured multiple times is offset in one parameter (e.g. two delayscans in their t0 or two lscans in L).

For this a really advanced evaluation, which could be integrated, would be a multi-step-fitting process:

1. Get two or more data sets which belong to the same physical measurement (e.g. two repeats of the same `delayscan)
2. fit data with function (e.g. `delayscan by a exponential or double exponential function)
3. get arbitrary parameter (e.g. t0)
4. offset all data by their t0 in the x-axis
5. bin data together to average them
6. fit the binned data again with the same function
   (t0 will not needed to be fit again now)