| dirttee-package {dirttee} | R Documentation |
DIstributional Regression for Times To EvEnt
Description
This package includes regession methods for right-censored response variables. It allows for the estimation of distributional regression methods with semiparametric predictors, including, for example, nonlinear, spatial or random effects. The distribution of the response can be estimated with expectiles, quantiles and mode regression. Censored observations can be included with accelerated failure time models or inverse probability of censoring weights.
Author(s)
Alexander Seipp, Fabian Otto-Sobotka
Carl von Ossietzky University Oldenburg
https://uol.de/eub
Maintainer: Alexander Seipp <alexander.seipp@uni-oldenburg.de>
Special thanks for their help go to Lisa Eilers and Florian Berger!
Partially funded by the German Research Foundation (DFG) grant SO1313/1-1, project 'Distributional Regression for Time-to-Event Data'.
References
Seipp A, Uslar V, Weyhe D, Timmer A, Otto-Sobotka F. Weighted expectile regression for right-censored data. Statistics in Medicine. 2021;40(25):5501–5520. doi: 10.1002/sim.9137
Seipp A, Uslar V, Weyhe D, Timmer A, Otto-Sobotka F. Flexible Semiparametric Mode Regression for Time-to-Event Data. Statistical Methods in Medical Research. 2022;31(12):2352-2367. doi: 10.1177/09622802221122406
Examples
data(colcancer)
c100 <- colcancer[1:100,]
#mode regression
reg <- modreg(Surv(logfollowup, death) ~ sex + LNE, data = c100)
#expectile regression
fit_exp <- expectreg.aft(Surv(logfollowup, death) ~ LNE, data = c100,smooth="f")
fit_expipc <- expectreg.ipc(Surv(logfollowup, death) ~ sex + LNE, data = c100)
#quantile regression
qu1 <- qureg.aft(Surv(logfollowup, death) ~ sex + LNE, data=c100, smooth="fixed")