Inclusion of time-varying covariates in cure survival models with an application in fertility studies
P. Lambert and V. Bremhorst 
J. R. Statist. Soc. A, 183 (2020), 333 -- 354


REQUIREMENT : 
-----------

Please, cite the paper when publishing analyses using this code. 


REMARKS : 
-------

1) The code is written in R with some calls to Fortran routines. The R code and the compiled (.so) file of the fortran routines are provided.
   
2) After downloading the data and defining the working directory, you simply have to execute the R code to obtain the results of the application section.

- The R code is well commented. No difficulties should be encountered when using it.
- For using the code on your own data, please note that this version of the code needs at least one covariate in each regression part.

3) Be aware that depending on the sample size, the computation time can take some time... Be patient

4) The code can only be run on Linux.

 
The data : 
--------

The data used in this paper are available at : 
https://www.diw.de/de/diw_01.c.361286.de/archiv.html

Data description 
----------------

ID : Subject ID 
BEGIN : Time of the follow-up in a given covariate configuration 
END : Time of the follow-up at the end of a given covariate configuration 
EVENT : Event indicator - Birth or no birth at the time who the covariate configuration change 
EDU02 : Mother's education level
PEDU02 : Partner's education level 
AGEKID1C : Age of the mother at first birth (continuous) 
AGEKID1 : Age of the mother at first birth (into category)
PERIOD : Calendar period 

ONLY FOR THIRD BIRTH

SEXCOMP : Sex composition of the first two children 
SPACE : Time elapsed between the first two children of the mother.


Philippe Lambert
Institut de Recherche en Sciences Sociales
Mthodes Quantitatives en Sciences Sociales
Universit de Lige
Place des Orateurs
3Quartier Agora
Btiment B31
B-4000 Lige
Belgium

E-mail: p.lambert@uliege.be

