Basis for Estimated Marginal Means for Compound Poisson-Normal Models

Provides the required components for computing estimated marginal means (EMMs) from a fitted Compound Poisson-Normal (CPN) model. This method is used by the emmeans package to interface with models of class "cpn".

Usage

emm_basis.cpn(object, trms, xlev, grid, ...)

Arguments

object: A fitted model object of class "cpn" returned by a call to e.g., cpn5().
trms: Terms object extracted from the model formula, typically provided by emmeans.
xlev: A list of levels for factors in the reference grid.
grid: A data frame representing the reference grid for which EMMs are to be computed.
...: Additional arguments (currently unused).

Value

A list with the components required by emmeans:

X: Model matrix for the reference grid.
bhat: Estimated regression coefficients.
V: Variance-covariance matrix of the coefficients.
nbasis: A matrix indicating no penalized basis functions (empty in this case).
dffun: Function returning degrees of freedom for inference (fixed at Inf).
dfargs: Arguments for dffun.
misc: List of transformation and variance functions used in EMM computations.