Branch time propagation now produces plausible output when propagating LAI. This branch is still python 2.7 I still need to merge changes into the master branch that is now in python 3.
This branch deals with 2 issues:
#6 Now have inflation as well as some other fixes to the LAI propagator such that it gets the correct uncertainties in the correct places.
#11 I have modified the innovations to be what is required by the Hessian correction.