Updates to Alignment

This MR

• corrects a mistake in calculating the alignment shifts for each iteration

• improves printout dramatically (units etc)

• moves to using local residuals for minimization for single detectors

• adds the possibility to prune tracks before selection for alignment. This includes

  • cut on the track's chi2ndof value
  • cut on the number of associated clusters

  both only really make sense for singe detector alignment (residual method)

• updates all alignment plots to be actually useful

Missing:

• document new features and parameters.
• cross-check dY/X etc profiles.