Learn R Programming

depCensoring (version 0.1.7)

Statistical Methods for Survival Data with Dependent Censoring

Description

Several statistical methods for analyzing survival data under various forms of dependent censoring are implemented in the package. In addition to accounting for dependent censoring, it offers tools to adjust for unmeasured confounding factors. The implemented approaches allow users to estimate the dependency between survival time and dependent censoring time, based solely on observed survival data. For more details on the methods, refer to Deresa and Van Keilegom (2021) , Czado and Van Keilegom (2023) , Crommen et al. (2024) , Deresa and Van Keilegom (2024) , Rutten et al. (2024+) and Ding and Van Keilegom (2024).

Copy Link

Version

Install

install.packages('depCensoring')

Monthly Downloads

217

Version

0.1.7

License

GPL-3

Maintainer

Negera Wakgari Deresa

Last Published

March 11th, 2025

Functions in depCensoring (0.1.7)

Compute the Gn matrix in step 3b of Bei (2024).

First step log-likelihood function for Z binary.

Link function (AFT model)

Loglikehood function under independent censoring

Lambda_inverse_Cox_wb

Inverse of link function (Cox model)

Second likelihood function needed to fit the independence model in the second step of the estimation procedure.

Family of box functions

Analogue to KMS_AUX4_MSpoints(...) in MATLAB code of Bei (2024).

Family of continuous/discrete instrumental function

M-step in the EAM algorithm described in KMS19.

Link function (Cox model)

Fit a semiparametric transformation model for dependent censoring

Obtain the correlation matrix of the moment functions

Expected improvement

Lambda_inverse_AFT_ll

Inverse of link function (AFT model)

E-step in the EAM algorithm as described in KMS19.

Family of spline instrumental functions

Likelihood function under dependent censoring

Likelihood.Profile.Solve

Solve the profiled likelihood function

Likelihood.Semiparametric

Calculate the semiparametric version of profiled likelihood function

Search function

Score equations of finite parameters

Inverse Yeo-Johnson transformation function

Second step log-likelihood function under independence assumption.

Calculate the goodness-of-fit test statistic

Semiparametric Estimation of the Survival Function under Dependent Censoring

Compute the variance-covariance matrix of the moment functions.

Estimate finite parameters based on score equations

LikI.cmprsk.Cholesky

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

control.arguments

Prepare initial values within the control arguments

Cumulative hazard function of survival time under independent censoring

Transform Cholesky decomposition to covariance matrix parameter element.

Estimate a nonparametric transformation function

Nonparametric bootstrap approach for the dependent censoring model

Nonparametric bootstrap approach for the independent censoring model

Function to simulate (Y,Delta) from the copula based model for (T,C).

check.args.pisurv

Check argument consistency.

Estimating equation for Ht1

Cumulative hazard function of survival time under dependent censoring

Change H to long format

Transform Cholesky decomposition to covariance matrix

Yeo-Johnson transformation function

Estimated survival function based on copula-graphic estimator (Archimedean copula only)

copdist.Archimedean

The distribution function of the Archimedean copula

Competing risk likelihood function.

Derivative of link function (AFT model)

get.anchor.points

Get anchor points on which to base the instrumental functions

Estimated survival function based on Kaplan-Meier estimator

Vector of sample average of each moment function \((\bar{m}_n(\theta))\).

feasible_point_search

Method for finding initial points of the EAM algorithm

Fit Dependent Censoring Models

get.cond.moment.evals

Compute the conditional moment evaluations

Derivative of transform Cholesky decomposition to covariance matrix element.

Derivative of link function (Cox model)

dat.sim.reg.comp.risks

Data generation function for competing risks data

Clear plotting window

do.optimization.Mstep

Optimize the expected improvement

Draw initial set of starting values for optimizing the expected improvement.

Derivative of transform Cholesky decomposition to covariance matrix.

Insert row into a matrix at a given row index

Obtain the matrix of partial derivatives of the sample variances.

Convert the Kendall's tau into the copula parameter

generator.Archimedean

The generator function of the Archimedean copula

Fit Independent Censoring Models

loglike.indep.unconstrained

Log-likelihood function for the independence copula.

Compute the critical value of the test statistic.

loglike.gumbel.unconstrained

Log-likelihood function for the Gumbel copula.

plot_addpte.eval

Draw evaluated points.

Draw points to be evaluated

get.test.statistic

Obtain the test statistic by minimizing the S-function over the feasible region \(\beta(r)\).

get.instrumental.function.evals

Evaluate each instrumental function at each of the observations.

get.deriv.mom.func

Matrix of derivatives of conditional moment functions

loglike.frank.unconstrained

Log-likelihood function for the Frank copula.

Faster implementation of vector of moment functions.

variance.cmprsk

Compute the variance of the estimates.

Estimate the control function

estimate.cmprsk

Estimate the competing risks model of Rutten, Willems et al. (20XX).

Rudimentary, bidirectional 1D grid search algorithm.

get.extra.Estep.points

Get extra evaluation points for E-step

Faster implementation to obtain the tensor of the evaluations of the derivatives of the moment functions at each observation.

Loss function to compute Delta(beta).

'Loss function' of the test statistic.

Logarithmic transformation function.

loglike.gaussian.unconstrained

Log-likelihood function for the Gaussian copula.

loglike.clayton.unconstrained

Log-likelihood function for the Clayton copula.

Obtain the value of the density function

Obtain the value of the distribution function

Second step log-likelihood function.

First step log-likelihood function for Z continuous

Grid search algorithm for finding the identified set

Likelihood.Parametric

Calculate the likelihood function for the fully parametric joint distribution

gs.interpolation

Return the next point to evaluate when doing interpolation search

Estimation of a parametric dependent censoring model without covariates.

Parameters.Constraints

Generate constraints of parameters

Likelihood.Profile.Kernel

Calculate the profiled likelihood function with kernel smoothing

normalize.covariates

Normalize the covariates of a data set to lie in the unit interval by scaling based on the ranges of the covariates.

test.point_Bei_MT

Perform the test of Bei (2024) simultaneously for multiple time points.

Vector of sample average of each moment function \((\bar{m}_n(\theta))\).

Return the next point to evaluate when doing regular grid search

power_transform

Power transformation function.

uniformize.data

Standardize data format

Maximum likelihood estimator for a given parametric distribution

set.hyperparameters

Define the hyperparameters used for finding the identified interval

Convert the copula parameter the Kendall's tau

SurvMLE.Likelihood

Likelihood for a given parametric distribution

The h-function of the copula

Summary of depCensoringFit object

Combine bounds based on majority vote.

boot.nonparTrans

Nonparametric bootstrap approach for a Semiparametric transformation model under dependent censpring

Return the next point to evaluate when doing binary search

Obtain next point for feasible point search.

set.EAM.hyperparameters

Set default hyperparameters for EAM algorithm

get.starting.values

Main function for obtaining the starting values of the expected improvement maximization step.

set.GS.hyperparameters

Set default hyperparameters for grid search algorithm

normalize.covariates2

Normalize the covariates of a data set to lie in the unit interval by transforming based on PCA.

Obtain the adjustment value of truncation

optimlikelihood

Fit the dependent censoring models.

Perform the test of Bei (2024) for a given point

summary.indepFit

Summary of indepCensoringFit object

likF.cmprsk.Cholesky

Wrapper implementing likelihood function using Cholesky factorization.

likIFG.cmprsk.Cholesky

Full likelihood (including estimation of control function).

Estimate the model of Willems et al. (2024+).

Main function to run the EAM algorithm

Check convergence of the EAM algorithm.

Obtain the diagonal matrix of sample variances of moment functions

Distance between vectors

Compute bivariate survival probability

Chronometer object

Compute phi function

Derivative of the Yeo-Johnson transformation function

A-step in the EAM algorithm described in KMS19

Bspline.unit.interval

Evaluate the specified B-spline, defined on the unit interval

Family of discrete/continuous instrumental functions, in the case of many covariates.

Calculate the kernel function