pars_tau: The SPDE construction uses parameters kappa and tau, while...

In jkbest2/spatq: Use spatially varying catchability to standardize fishery-dependent data

The SPDE construction uses parameters kappa and tau, while the usual Matern parameters are sigma and rho. The kappa parameter is sqrt(8) / rho.

Description

The SPDE construction uses parameters kappa and tau, while the usual Matern parameters are sigma and rho. These functions allow converting between the parameterizations.

Usage

pars_kappa(rho)

pars_tau(sig, rho)

pars_rho(kappa)

pars_sig(tau, kappa)

Arguments

rho	Correlation range parameter
sig	Marginal standard deviation
kappa	Correlation decay parameter
tau	Ratio of precision and correlation range

Details

The tau parameter rescales the precision matrix to a given marginal standard deviation, while kappa controls the correlation decay rate. In particular,

• tau = 2 * sqrt(pi) * sig * kappa,

• kappa = sqrt(8) / rho,

• sig = tau / (2 * sqrt(pi) * kappa), and

• rho = sqrt(8) / kappa.

All assume that the Matern smoothness parameter is 1 and the domain is two-dimensional

Value

Parameter value

Functions

• pars_kappa: Calculate kappa from rho

• pars_rho: Calculate the correlation range

• pars_sig: Calculate the marginal standard deviation

Author(s)

John Best

jkbest2/spatq documentation built on Sept. 22, 2022, 3:22 a.m.