The files in this directory are multi-Gaussian approximations to Sersic profiles,
as computed by the file optimize_mixture_profiles.py in TheTractor:

https://github.com/davidwhogg/TheTractor

For more information on why this is a useful idea, and how to select which one
to use, please see Hogg and Lang 2012:

http://arxiv.org/abs/1210.6563

The pickled objects themselves are numpy.ndarray objects of size 2K, in which the
first K elements are the amplitudes, the last K elements are the variances, and
K is the number of Gaussian components in the approximation.  The filenames are
of the form "<model>_K<number-of-gaussians>_MR<maximum-radius>".  The maximum radius
is the maximum radius used when optimizing the approximation, in units of the
profile's half-light radius.
