David J. Harris
David J. Harris
I've posted a (somewhat indirect) way to add to the loss function here: https://github.com/blei-lab/edward/issues/529#issuecomment-341943910
I'm not really sure what's out there in terms of R GP packages, but I know that @goldingn has been thinking about writing a package for a while. My needs...
Nice. Good luck!
Appears to sort-of work, as of [this gist](https://gist.github.com/davharris/c322b8bd521adb2d6759). Should still test it
Isn't this already done? What else do I need to do before I can close it?
Q should have the same marginals as P but all its elements should be independent. Can get that by doing 1000 steps of regular importance sampling (i.e. sampling from the...
Do I need the annealing? Why not just sample from the marginal distribution estimated by mean field and see how that does?
This could be useful for calibrating its parameters (number of iterations and dampening). Also good for testing.
I think I can get all of the parameter estimates for the above with ``` fitContinuous(phy = phylo, dat = latent.trait, model = "OU", SE = NA) ```
Partly implemented in the simulation script that makes `fakedata`