constructLogNormalPriors: A function to construct a list of independent priors on...

In apear9/SSNdesign: An R Package for Optimal and Adaptive Sampling Designs on Stream Networks

Description

The main design function in SSNdesign is called optimiseSSNDesign, and it has an argument specifying a list of independent priors on the covariance parameters of a fitted glmssn object.

Usage

constructLogNormalPriors(glmssn, std)

Arguments

glmssn	A fitted glmssn object, preferably with optimOutput$hessian.
std	A numeric vector. This vector specifies the standard deviation of the log-normal priors. See Details for more information.

Details

The argument std can be of length 1, in which case all covariance parameters (parsill, range) will have the same value for standard deviation. Alternatively, it can be of length two, in which case the first element will be used for the partial sill parameters and the second for range parameters. Alternatively, it can be of the same length as glmssn$estimates$theta in which the elements will be matched to their corresponding parameters.

Value

A list whose elements are functions parameterised in terms of x, the number of Monte Carlo draws to be taken from the priors when evaluating the expected utility.

Examples

# Set seed
set.seed(1)

# Create SSN
s <- createSSN(10, binomialDesign(10), 
path = paste(tempdir(), "example00.ssn", sep = "/"), importToR = TRUE)
# And distance matrix
createDistMat(s)

# Simulate data
s <- SimulateOnSSN(s, getSSNdata.frame(s), formula = ~ 1, coefficients = c(1),
 CorModels = c("Spherical.tailup"), CorParms = c(1, 2, 0.1), 
 addfunccol = "addfunccol")$ssn.object

# Fit model
m <- glmssn(s, formula = Sim_Values ~ 1, CorModels = c("Spherical.tailup"),
 addfunccol = "addfunccol")

# Construct log-normal priors
p <- constructLogNormalPriors(m)

apear9/SSNdesign documentation built on Feb. 19, 2020, 4:29 a.m.