compVagueSd: compVagueSd - Compute sD's of vague priors

In tmcd82070/EoAR: Evidence of Absence Regression (EoAR)

Description

Compute the standard deviation of a normal distribution that is big enough to be considered a 'vague' prior. This is not straight forward when there are covariates, as here.

Usage

compVagueSd(Y, alpha.vec, beta.vec, X, range.multiplier = 100)

Arguments

Y	Vector of number of carcasses found, one element per year. If multiple sites are involved, elements of Y are the total (summed) number of targets found per season.
alpha.vec	vector of the alpha parameters of the Beta distributions
beta.vec	vector of the beta parameters of the Beta distributions
X	a design matrix upon which an approximation of inflated numbers of targets is regressed. Usually, this is meant to be the design matrix from the eoar function.
range.multiplier	a multipiler for the range of coefficient estimates to make the output standard deviations sufficiently vague. Increasing this number increases vagueness.

Value

List containing two components. $vagueSd is a vector, one per parameter, of standard deviations that should be large enough to call vague when used in a normal prior. $startA is a vector of potential starting values for the coefficients in the model.

tmcd82070/EoAR documentation built on July 13, 2021, 5:52 p.m.