LikI.cmprsk.Cholesky: Wrapper implementing likelihood function assuming...

In depCensoring: Statistical Methods for Survival Data with Dependent Censoring

Wrapper implementing likelihood function assuming independence between competing risks and censoring using Cholesky factorization.

Description

This function does the same as LikI.cmprsk (in fact, it even calls said function), but it parametrizes the covariance matrix using its Cholesky decomposition in order to guarantee positive definiteness. This function is never used, might not work and could be deleted.

Usage

LikI.cmprsk.Cholesky(
  par.chol,
  data,
  eoi.indicator.names,
  admin,
  conf,
  cf,
  eps = 0.001
)

Arguments

par.chol	Vector of all second step model parameters, consisting of the regression parameters, Cholesky decomposition of the variance-covariance matrix elements and transformation parameters.
data	Data frame resulting from the 'uniformize.data.R' function.
eoi.indicator.names	Vector of names of the censoring indicator columns pertaining to events of interest. Events of interest will be modeled allowing dependence between them, whereas all censoring events (corresponding to indicator columns not listed in eoi.indicator.names) will be treated as independent of every other event.
admin	Boolean value indicating whether the data contains administrative censoring.
conf	Boolean value indicating whether the data contains confounding and hence indicating the presence of Z and W.
cf	"Control function" to be used. This can either be the (i) estimated control function, (ii) the true control function, (iii) the instrumental variable, or (iv) nothing (cf = NULL). Option (ii) is used when comparing the two-step estimator to the oracle estimator, and option (iii) is used to compare the two-step estimator with the naive estimator.
eps	Minimum value for the diagonal elements in the covariance matrix. Default is eps = 0.001.

Value

Log-likelihood evaluation for the second step in the estimation procedure.

depCensoring documentation built on April 4, 2025, 1:52 a.m.