set.par.ps: Define the model coefficients of a geostatistical linear...

In PrevMap: Geostatistical Modelling of Spatially Referenced Prevalence Data

Description

set.par.ps defines the model coefficients of a geostatistical linear model with preferentially sampled locations. The output of this function can be used to: 1) define the parameters of the importance sampling distribution in lm.ps.MCML; 2) the starting values of the optimization algorithm in lm.ps.MCML.

Usage

set.par.ps(p = 1, q = 1, intensity, response, preferentiality.par)

Arguments

p	number of covariates used in the response variable model, including the intercept. Default is p=1.
q	number of covariates used in the log-Guassian Cox process model, including the intercept. Default is q=1.
intensity	a vector of parameters of the log-Gaussian Cox process model. These must be provided in the following order: regression coefficients of the explanatory variables; variance and scale of the spatial correlation for the isotropic Gaussian process. In the case of a model with a mix of preferentially and non-preferentially sampled locations, the order of the regression coefficients should be the following: regression coefficients for the linear predictor with preferential sampling; regression coefficients for the linear predictor with non-preferential samples.
response	a vector of parameters of the response variable model. These must be provided in the following order: regression coefficients of the explanatory variables; variance and scale of the spatial correlation for the isotropic Gaussian process; and variance of the nugget effect.
preferentiality.par	value of the preferentiality paramter.

Value

a list of coefficients of class coef.PrevMap.ps.

Author(s)

Emanuele Giorgi [e-mail]

PrevMap documentation built on Oct. 7, 2021, 5:07 p.m.