Linear Model of Coregionalization

[kyle-messier]

An important comparison we should likely have in the manuscript is the Linear Model of Coregionalization (LMC)

Can we discuss what it would take adopting our current code base for LMC including the Vecchia approx?

1. LMC could potentially be done using the univariate PrestoGP to get marginal models, then fit the LMC matrices. So potentially no updates to the package directly.
2. Allowing the package to fit some version of the LMC

Doing LMC may also be a good starting point for the $\theta's$ in PrestoGP.

[kyle-messier]

After doing a little more reading - I think I can do an easy wrapper around the PrestoGP univariate model, then optimize for the coregionalization matrix. So for now, no need to do anything with this. We'll see if we want to make it an easier option in v2.0