Updates to documentation with xarray tutorial data

[roxyboy]

A PR to update the documentation and address #147 . I added an example from #144.

[review-notebook-app]

Check out this pull request on 

See visual diffs & provide feedback on Jupyter Notebooks.

---

Powered by ReviewNB

[roxyboy]

I think we should make a new release if people are ok with the changes in the documentation.

[roxyboy]

@lanougue @rabernat Any thoughts on this PR on making the documentation more helpful? I added an example using the xarray tutorial data.

[lanougue]

Thanks @roxyboy for this work! A personnal global comment on this documentation would be that it seems to prove that numpy.fft could be enough if used carrefully (in combination with apply_ufunc for example).
On the contrary I would promote xrft power arguing that it takes care of phase, amplitude, spacing, number of points, ... automatically. From my point of view, this should also be done by emphasizing the defaults of the numpy.fft instead of trying to show that they can be handled by a careful user.
xrft is not only a very useful wrapper for Fourier Transforms with xarrays, but also correctly takes care of all the scaling and phasing inherent to Fourier Transforms.
Thanks again for the energy you put in xrft! :)