This is a port of the changes added to matgwinc in
gwinc/matgwinc@a83e5053b64e6e8d3176f8fc29df631972fec451

The new pygwinc quantum curve diverges slightly from the matgwinc
version at the highest frequencies, for reasons that are not clear
to me...

e.g. ifo['Sub.Element']

newtonian noise calculation was fixed in both py and mat gwinc, so
update matgwinc calculation cache in tests.

this was just a typo

use consistent naming

(Gitlab wants it to be called CONTRIBUTING.md)

Implements the changes recommended by Jan Harms in https://git.ligo.org/rana-adhikari/CryogenicLIGO/issues/45
Fixes #18

Add better description about IO, and clarify some of the argument help.

Add ability to load a noise budget previously saved in an HDF5 file.

If an HDF5 file is passed to the --save option (indicated by '.hdf5' extension)
then save the data to an HDF5 file without plotting.

Functions to save/load noise traces and IFO model parameters to/from HDF5
files.  Requires the h5py module.

free up the "util" namespace for other purposes not related to noise
calculation

namespace conflict with the gwinc plot keyword argument.

automatic creation of the commad line interface script

So that distribution packages and tarballs don't misrepresent what the
project is.

fix name, version and included files

no functional change

Use the new CBCWaveform interface in the insprial_range package for specifying
the waveform and parameters, both in the CLI and test.