CQCS: Censored quantile correlation based screening

The cqcs package provides the functions to compute the CQC (censored quantile correlation) proposed by Peng Su, Ruoxuan Zhang and Maozai Tian.

We have included a simulation workflow (simu and analysis) in vignettes.

Key functions

There are two key functions in the package.

function cqcor is used to caculate the CQC between two variables, which can be used to screen the variables.

cqcor(y, d, x, tau) 
# y: response variable, 
# d: censoring indicator, 
# x: covariate, 
# tau: quantile level

function cqcs is used to screen the variables based on CQC.

cqcs(y, d, x, tau = 0.5, len = 100) 
# y: response variable, 
# d: censoring indicator, 
# x: covariates, 
# tau: quantile level, 
# len: screening length

Installation

You can install the package using:

remotes::install_github("PengSU517/cqcs")