R/coef.coxph_mpl_dc.R

In survivalMPLdc: Penalised Likelihood for Survival Analysis with Dependent Censoring

#' @title Extract regression coefficients of a coxph_mpl_dc Object
#'
#' @description Extract the matrix of regression coefficients with their corresponding standard errors,
#' \eqn{z}-statistics and \eqn{p}-values of the model part of interest of a \code{coxph_mpl_dc} object
#'
#' @method coef coxph_mpl_dc
#'
#' @param object an object inheriting from class \code{\link{coxph_mpl_dc}}
#' @param parameter the set of parameters of interest.  Indicate \code{parameters="beta"} for the regression parameter of beta
#' and \code{parameters="phi"} for the regression parameter of phi
#' @param ... other arguments
#'
#'
#' @details When the input is of class \code{coxph_mpl_dc} and \code{parameters=="beta"},
#' the matrix of beta estimates with corresponding standar errors, \eqn{z}-statistics and \eqn{p}-values are reported. When the input is of class \code{coxph_mpl_dc} and \code{parameters=="phi"},
#' the matrix of phi estimates with corresponding standar errors, \eqn{z}-statistics and \eqn{p}-values are reported.
#'
#' @return \item{est}{a matrix of coefficients with standard errors, z-statistics and corresponding p-values}
#'
#' @author Jing Xu, Jun Ma, Thomas Fung
#'
#' @references Brodaty H, Connors M, Xu J, Woodward M, Ames D. (2014). \emph{"Predictors of institutionalization
#' in dementia: a three year longitudinal study"}. Journal of Alzheimers Disease 40, 221-226.
#'
#' Xu J, Ma J, Connors MH, Brodaty H. (2018). \emph{"Proportional hazard model estimation under dependent censoring using copulas and penalized likelihood"}.
#' Statistics in Medicine 37, 2238–2251.
#'
#'
#' @seealso \code{\link{plot.coxph_mpl_dc}}, \code{\link{coxph_mpl_dc.control}}, \code{\link{coxph_mpl_dc}}
#'
#' @importFrom copula indepCopula claytonCopula gumbelCopula frankCopula iTau rCopula
#' @importFrom survival coxph Surv
#' @importFrom splines2 iSpline mSpline
#'
#' @examples
#' \donttest{
#'
#'  ##-- Copula types
#'  copula3 <- 'frank'
#'
#' ##-- A real example
#' ##-- One dataset from Prospective Research in Memory Clinics (PRIME) study
#' ##-- Refer to article Brodaty et al (2014),
#' ##   the predictors of institutionalization of dementia patients over 3-year study period
#'
#' data(PRIME)
#'
#' surv<-as.matrix(PRIME[,1:3]) #time, event and dependent censoring indicators
#' cova<-as.matrix(PRIME[, -c(1:3)]) #covariates
#' colMeans(surv[,2:3])  #the proportions of event and dependent censoring
#'
#' n<-dim(PRIME)[1];print(n)
#' p<-dim(PRIME)[2]-3;print(p)
#' names(PRIME)
#'
#' ##--MPL estimate Cox proportional hazard model for institutionalization under independent censoring
#' control <- coxph_mpl_dc.control(ordSp = 4,
#'                                 binCount = 200, tie = 'Yes',
#'                                 tau = 0.5, copula = copula3,
#'                                 pent = 'penalty_mspl', smpart = 'REML',
#'                                 penc = 'penalty_mspl', smparc = 'REML',
#'                                 cat.smpar = 'No' )
#'
#' coxMPLests_tau <- coxph_mpl_dc(surv=surv, cova=cova, control=control, )
#' MPL_beta<-coef(object = coxMPLests_tau, parameter = "beta",)
#' MPL_phi<-coef(object = coxMPLests_tau, parameter = "phi",)
#'
#' }
#'
#' @export

coef.coxph_mpl_dc<-function(object, parameter,...)
{
out<-object$mpl_beta_phi_zp
  if(parameter=="beta")
  {
   est <- out[ 1 : ( dim(out)[1]/2 ) , ]
  }else if(parameter=="phi"){
    est <- out[ ( dim(out)[1]/2 + 1 ) : ( dim(out)[1] ), ]
  }else (stop("Error! Define a proper parameter name"))
return(est)
}