More options for prealignment

As discussed in issue #65 (closed), the Prealignment module is extended by providing two more options for the translational shift. Previously, the method used the mean of the correlation histogram to calculate the shift. Now, also the mean coming from a Gaussian fit to the peak or the maximum of the histogram can be used. For the Gaussian fit, one can choose the fitting range by hand. By default, the "old method" is used.

When I tested the three methods on CLICTD data from the previous test-beam, I found that the Gaussian fit and the maximum work equally well and both work better than the mean. When one looks at the Gaussian fitted to the histogram (which is by default displayed in the output of the Prealignment module) it's clear that a Gaussian does not describe the data very well but generally, the peak is fitted well enough to get its location and therefore do a reasonable translational shift, at least in my tests.

Feedback? :)

This fixes #65 (closed)