survsplit: Split a survival data set at specified cut points

In trtswitch: Treatment Switching

Split a survival data set at specified cut points

Description

For a given survival dataset and specified cut times, each record is split into multiple subrecords at each cut time. The resulting dataset is in counting process format, with each subrecord containing a start time, stop time, and event status. This is adapted from the survsplit.c function from the survival package.

Usage

survsplit(tstart, tstop, cut)

Arguments

tstart	The starting time of the time interval for counting-process data.
tstop	The stopping time of the time interval for counting-process data.
cut	The vector of cut points.

Value

A data frame with the following variables:

• row: The row number of the observation in the input data (starting from 0).

• start: The starting time of the resulting subrecord.

• end: The ending time of the resulting subrecord.

• censor: Whether the subrecord lies strictly within a record in the input data (1 for all but the last interval and 0 for the last interval with cutpoint set equal to tstop).

• interval: The interval number derived from cut (starting from 0 if the interval lies to the left of the first cutpoint).

Author(s)

Kaifeng Lu, [e-mail]

Examples

survsplit(15, 60, c(10, 30, 40))

trtswitch documentation built on June 8, 2025, 1:45 p.m.