TruthPrior-class {EcoEnsemble} | R Documentation |
A class to hold the priors for the truth model in the ensemble framework
Description
An TruthPrior
object encapsulates the prior information for the short-term discrepancies of the shared discrepancy of the ensemble model.
Details
The truth \mathbf{y}^{(t)}
is modelled as a random walk such that
\mathbf{y}^{(t+1)} \sim N(\mathbf{y}^{(t)}, \Lambda_y).
The covariance matrix \Lambda_y
is parameterised by an inverse Wishart distribution (contained in the rw_covariance
slot) and the initial value is modelled as drawn from a normal distribution.
Slots
d
A
numeric
giving the number of variables of interest in the ensemble model.initial_mean
A
numeric
giving the standard deviation of the normal prior on the initial mean value of the random walk. This is the same standard deviation for each variable of interest.initial_var
A
list
of length2
containing the shape and scale parameters (respectively) for the gamma priors on the variance of the initial value of the truth.rw_covariance
A
list
of length2
containing the inverse-Wishart parameters for the covariance of the random walk of the truth.