Interpretation of posterior predictive checks for a Gaussian Process

Questions v5
, ,
4

essentially the same as gp.Marginal - I got a very similar result

Yup that’s right, pm.MvNormal is pretty much all gp.Marginal does when fitting.

I see, you have 300 data points so that’s the size of the GP, not unusual at all. If convergence is good then you’re most likely good to go

1 Like