R: Non-Separable Space-Time model

RMgennsst {RandomFields}

R Documentation

Non-Separable Space-Time model

RMgennsst is a univariate stationary spaceisotropic covariance model on R^d whose corresponding covariance is given by

C(h,u)= φ( sqrt(h^\top ψ^{-1}(u) h)) / √(det(psi))

RMgennsst(phi, psi, dim_u, var, scale, Aniso, proj)

`phi`	is a normal mixture `RMmodel`, cf. `RFgetModelNames(monotone="normal mixture")`.
`psi`	is a d-variate variogram model `RMmodel`.
`dim_u`	the dimension of the component `u`
`var,scale,Aniso,proj`	optional arguments; same meaning for any `RMmodel`. If not passed, the above covariance function remains unmodified.

This model is used for space-time modelling where the spatial component is isotropic.

RMgennsst returns an object of class RMmodel.

Schlather, M. (2010) On some covariance models based on normal scale mixtures. Bernoulli, 16, 780-797.

[Package RandomFields version 3.3.4 Index]