f.double.prime: Second derivative of the posterior distribution of a spatial GEV
with respect a location random effect
Description
This is the second derivative of the conditional posterior probability
of a sptial GEV model with respect to a given random effect in the
location parameter. It is used to form the proposal for a
Metropolis-Hastings update of the given parameter.
Usage
f.double.prime(tau, tau.hat, varsigma, xi, kappa, R)
Arguments
tau
The current value of the given random effect
tau.hat
The conditional mean of the random effect given the other random
effects and the Gaussian process parameters
varsigma
The conditional variance of the random effect given the Gaussian
process parameters
xi
The current shape parameter at this location
kappa
The current precision parameter at this location
R
The residual between the observed values and the linear part of the
location parameter
Value
A scalar that gives the value of the derivative. This is then used to
form the relevant proposal.