R: Diggle-Rowlingson Raised Incidence Model

tribble {splancs}

R Documentation

Diggle-Rowlingson Raised Incidence Model

Description

Fits the Diggle-Rowlingson Raised Incidence Model.

Usage

tribble(ccflag, vars=NULL, alphas=NULL, betas=NULL, rho, 
 which=1:length(alphas), covars=NULL, thetas=NULL, 
 steps=NULL, reqmin=0.001, icount=50, hessian=NULL)

Arguments

`ccflag`	Case-control flag : a vector of ones and zeroes.
`vars`	A matrix where `vars[i,j]` is the distance squared from point `i` to source `j`.
`alphas`	Initial value of the `alpha` parameters.
`betas`	Initial value of the `beta` parameters.
`rho`	Initial value of the `rho` parameter.
`which`	Defines the mapping from sources to parameters.
`covars`	A matrix of covariates to be modelled as log-linear terms. The element `covars[i,j]` is the value of covariate `j` for case/control `i`.
`thetas`	Initial values of covariate parameters.
`steps`	Step sizes for the Nelder-Mead simplex algorithm.
`reqmin`	Tolerance for simplex algorithm
`icount`	Iteration count for simplex algorithm
`hessian`	by default NULL, any other value causes hessian to be computed and returned

Value

The return value is a list with many components, and class ribfit.

`alphas`	A vector of the alpha parameters at the maximum
`betas`	A vector of the beta values at the maximum
`rho`	The value of rho at the maximum
`logl`	The maximised log-likelihood
`null.logl`	The null log-likelihood
`call`	The function call to `tribble`

For further information see Diggle and Rowlingson (1993).

References