Code for spherical orbit superposition models of discrete kinematics.

To get it to compile it should just be a matter of

  % python3 setup.py install --user

but you might need to say "python" instead of "python3" and you'll
probably have to a install a c++ compiler, the GNU scientific (GSL)
and hdf5 libraries, plus pybind11.  On a mac that's probably most
easily dealt with by using homebrew (https://brew.sh).

Once that's working the main python driver program is
test_generate_and_liks.py.  It generates fake data and calculates the
likelihood histogram for each fake datapoint.



The files ELLzBlock.h and ELLzBlock.cc are the gory parts that you'll
have to wrestle with most, although there is a lot of extraneous stuff
there that you won't use.  Here's what most relevant:

 - The proberties of each block/tile are contained in the first few
   lines of ELLzBlock.h.  But see also the "BlockHolder" class later
   in the same file.

 - The likelihood calculation is carried out by the "fillup" method of
   the BlockHolder class.  The walk along the line of sight looking
   for intersections is from line 392 of ELLzBlock.cc