zero_corr_max_peak makes strange empty samples at the end of the profile

[erikmannerfelt]

zero_corr_max_peak leaves/appends strange values at the end of a profile when run. I assume that this has to do with that a new array is constructed with only zeros:

rsgpr/src/gpr.rs

Line 620 in dc73fa0

let mut new_data = Array2::from_elem(

.. and then it may not be trimmed at the end correctly.

Here's the end of a radar profile without zero-corr:

And here it is with zero_corr_max_peak:

The fix may be as simple as just trimming the data correctly. I also think there's a zero-copy way of doing this instead of instantiating an entirely new array. Finally, a test should be added to make sure this does not happen!