added beam_height_ft_doviak, which uses a slightly different formular than
collier.
The advantage of this formula is that it is invertible for theta (which may
be of interest for visibility studies). Also, it is deriveable using simple
trigonometry, whereas the derivation of the Collier-formula is rather obscure.

Added literature reference to beam_height_ft

fixed wrong treatment of large and small overflow sectors.

added calls to interp_atten and bisectReferenceAttenuation.
Everything not tested. bisectReferenceAttenuation is also missing a reduction
of b (for those rare cases where this is necessary), but this might also
be addressed by another function before the bisection.

removed some of the mind-bogglingness of multidimensionality by reshaping the
original data to something three-dimensional ('the rest', azimuth, range).

added interp_atten. Function to interpolate reference attenuation to invalid
beams. Not yet tested.

made sector_filter work with arbitrary sector sizes (=dynamic origin).

Added testing routines for atten module

minor change in tutorial

resized some tutorial images and changed some tutorial text

removed the oversized tutorial images from the readDX tutorial

updated the io- and the vis-tutorial
georef_example.py: multiply the radar-range array by 1000, due to SI-convention

changes in wrl.ipol.interpolate_polar:
1. the mask array is optional, if the data array is masked
2. in terms of recommended efficiency ONLY masked values will be interpolated
3. in consequence for recalculation of linear interpolation (values at the rim can't...

added convenient interpolation function for polar data:
ipol.interpolate_polar

Added first (example of a) testing file.

Added preliminary code for the new (and almost complete) implementation of the
Kraemer algorithm.

Renamed some variables in atten.py for better understanding:
k -> pia (because it integrates all foregoing k's)
kn -> k (the explicit indication that it stands for the specific attenuation of
bin n is now (first rename) unessential

Attention for users of georef:
changed units for earth radius in polar2latlon so this function expects ranges in
SI-units([m])
This affects the functions polar2polyvert, polar2centroids and
projected_bincoords_from_radarspecs
Further inout parameter re (earth radius) is removed in functions, which do not...

Attention for users of georef.projected_bincoords_from_radarspecs:
changed units for range-array from [km] to [m] (SI-units)
will change that in the near future for the other functions, too

typo in the reference of ipol.linear documentation