compPI: Computation of prediction intervals

In szugat/predfat: What the package does (short line)

Description

Computation of prediction intervals

Usage

compPI(xNew, x0, L, L_max = 35L, theta, lambda, gradient, type, alpha, I,
  withSolve = TRUE)

Arguments

xNew	values of influential variable
x0	observed values of influential variable
L	integer value indicating the number of events to be predicted (used when xNew not given)
L_max	integer value indicating the maximum number of events in Poisson process when xNew is not given
theta	numeric vector of length four with link function's parameters
gradient	gradient of link function w.r.t. parameters
type	if link or gradient are not given, there is collection of implented link functions and gradients, see linkfun and gradLambda
alpha	confidence niveau in (0,1)
I	information matrix
withSolve	logical value indicating if inverting of information matrix with solve() or ginv() from MASS
link	link function for the negative (!) logarithmized mean of the exponential distribution

Value

list with prediction interval bounds and corresponding quantiles of hypoexponential

szugat/predfat documentation built on May 31, 2019, 12:50 a.m.