idmloo: R function to estimate n illness-death models for...

In camsabathe/pseudoICD: Pseudo-observations for interval-censored data from an illness-death model

Description

R function to estimate n illness-death models for leave-one-out data

Usage

idmloo(object, data, parallel = FALSE, cpus = NULL, type = NULL, ...)

Arguments

object	an illness-death model estimate from SmoothHazard package
data	data.frame used to estimate object
parallel	Logical determinating parallel (=TRUE) or sequential (=FALSE) execution
cpus	Numerical amount of CPUs requested for the cluster. If not set, values are total cores of the machines-1.
type	Type of cluster. Can be 'SOCK', 'MPI', 'PVM' or 'NWS'. Default is 'SOCK' if parallel=TRUE.
...	others parameters

Value

a list of length N+1 with one indice for leave-one-out of the data and the last the original estimate of object

Examples

data("Paq1000")
fit.splines <-  idm(formula02=Hist(time=t,event=death)~1,
formula01=Hist(time=list(l,r),event=dementia)~1,
formula12=Hist(time=t,event=death)~1,
method="Splines", kappa=c(200000,200000,20000),
data=Paq1000)
fit.loo.sp=idmloo(object=fit.splines, data=Paq1000, parallel = FALSE, cpus = NULL, type = NULL)

camsabathe/pseudoICD documentation built on Nov. 11, 2019, 5:27 p.m.