PIOStest_BiSurvCopula: PIOS test for misspecification of copula functions in...

In michellezhou2009/IRtests: Information ratio test and pseudo "in-and-out-of-sample" likelihood ratio test

PIOS test for misspecification of copula functions in semi-parametric survival copula model for bivariate right-censored data

Description

PIOS test for misspecification of copula functions in semi-parametric survival copula model for bivariate right-censored data

Usage

PIOStest_BiSurvCopula(
  x1,
  x2,
  d1,
  d2,
  copula.fam,
  control = list(yes.exact = FALSE, yes.boot = TRUE, nboot = 500, seed1 = 1234, same.cen
    = TRUE)
)

Arguments

x1	a vector, the first response
x2	a vector, the second response
d1	a vector of indicators whether each observation in x1 is fully observed: 1 indicates the observation is fully observed, and 0 indicates the observation is censored
d2	a vector of indicators whether each observation in x2 is fully observed: 1 indicates the observation is fully observed, and 0 indicates the observation is censored
copula.fam	a character indicating which one of the following copula families: "clayton", "frank", "joe","gumbel", and "normal"
control	a list of the following components: yes.exact, yes.boot, nboot, seed1, and same.cen. yes.exact is a logical value indicating whether to calculate the exact test statistic; if yes.exact=FALSE (default value), the approximate test statistic is calculated. yes.boot is a logical value indicating whether to implement the bootstrap procedure. nboot is the number of bootstrap samples. seed1 is the seed for generating the bootstrap samples. same.cen is a logical value indicating whether the censoring time is same for both event time.

Value

a list of the following components: theta.est, the PMLE of the copula parameter, PIOS, PIOS test statistic, PIOS.boot, the bootstrapped resamples of the PIOS test statistics if yes.boot=TRUE, and pval, the one-sided and two-sided p-values if yes.boot=TRUE.

michellezhou2009/IRtests documentation built on Aug. 19, 2023, 11:24 p.m.