gpr_posterior: GPR posterior
Description
a simple GPR posterior distribution, no parameter learning
Usage
gpr_posterior(x, y, x.targets, noise, kernelfunc, derivatives = FALSE)
Arguments
x.targets
target timepoints
noise
noise std, a single value or a vector
kernelfunc
a kernel function of (x1,x2) (returns a matrix)
derivatives
compute also derivatives
Value
a gpsimple-object with fields
- x
- timepoints
- mean
- GP mean
- cov
- covariance matrix
- noisestd
- vector of noise std's
- mll
- marginal log likelihood
- x.obs
- original observation times
- y.obs
- original observation values