Replication data for: Repeated Events Survival Models: The Conditional Frailty Model - Janet M. Box-Steffensmeier Dataverse

Use the check boxes next to the file name to download multiple files. Data files will be downloaded in their default format. You can also download all the files in a category by checking the box next to the category name. You will be prompted to save a single archive file. Study files that have restricted access will not be downloaded.

Select all files

1. Documentation

repeatedevents.pdf
Adobe PDF - 218 KB - 30 downloads

Download

Original article for this study

simulation_readme.rtf
MS Word - 7 KB - 23 downloads

Download

Document describing the files in the simulation zip file

2b. Data - Simulation Files

covarnorm.R
application/octet-stream - 698 bytes - 13 downloads

Download

creates a function called covarnorm that generates the heterogeneity in the data. The heterogeneity is normally distributed when covarnorm is called. The length of the vector is determined by the first argument passed to the function. The second and third arguments are the mean (default 0) and standard deviation (default 1) of the normally distributed random variables

covar.R
application/octet-stream - 929 bytes - 13 downloads

Download

creates a function called covar that generates heterogeneity in the data. It produces a vector of uniformly distributed random variables. The length of the vector is determined by the first argument, the last two arguments give the min and max values of the randomly generated variables. The defaults are -2 and 2, but alternative values can be passed to the function. A uniform variable on -2 to 2 has a variance of 1

klambda.R
application/octet-stream - 817 bytes - 14 downloads

Download

creates the function klambda. This generates the event dependence in the data. The function is passed a vector of risks; the maximum number of events to be generated; and a flag signifying whether event dependence exists or alternatively whether the risk values returned are independent. If the flag is one, then the risks are generated to be independent. The form of the event dependence is: klambda: for(i in 1:k){ timesbyk[,i]<-rexp(n,risk+((i-1)*risk)) }

simdatanorm.R
application/octet-stream - 12 KB - 12 downloads

Download

creates a function called simdatanorm that generates data with normally distributed random effects. The function is passed covarnorm rather than covar and also klambda

simdata.R
application/octet-stream - 14 KB - 12 downloads

Download

creates a function called simdata that generates data with uniformly distributed random effects. The function requires many arguments, including the functions covar and klambda, as well as the sample size, treatment effect, baseline hazard... The function returns a dataframe that has data in both gap and elapsed time

Simulation Files.zip
application/x-zip-compressed - 12 KB - 14 downloads

Download

Simulation Files for "Repeated Events Survival Models"

The Ohio State University