Supplement to 'Goodness-of-fit test for a parametric mixture cure model under partly interval-censored data'.
R code and data for analysis in Ziqi Geng, Jialiang Li, Yi Niu and Xiaogung Wang. Goodness-of-fit test for a parametric mixture cure model under partly interval-censored data. manuscript.
Here is a general explanation of what is contained in the folders.
gof_icc.R: Main R application for testing the null hypothesis.
gqz.R: Density, distribution function and random generation for the Gompertz distribution with parameters shape and scale.
f.R: Log-likelihood function.
dlrb_n.R: Generation of bootstrap samples.
lamda_nbwB.R: The bootstrap test statistic calculated by assuming Weibull-distributed original data.
lamda_nbgB.R: The bootstrap test statistic calculated by assuming Gompertz-distributed original data.
lamda_nblB.R: The bootstrap test statistic calculated by assuming lognormal-distributed original data.
Turnbull3.R: A slight transformation of 'Turnbull' function from 'ReIns' package for the purpose of vectorization.
sc.dat: A PIC version of the data (7.6% exact observations) similarly to Gao et al.(2017) by treating observations with interval lengths less than or equal to 0.962 years as exact observations at the geometric mid-point of the interval to obtain 17 exact observations.
actg.dat: Interval-censored observations with interval lengths less than 2 months are treated as exact observations at the geometric mid-point of the interval to obtain 46 exact observations.