surv_split: Split time to interval for survival model - improved

In kklot/ktools: Miscellaneous tools

Split time to interval for survival model - improved

Description

This is similar to survSplit but allow the starting time and ending time to be equal (e.g., immediate death, automatic default in loan) and automatically include min and max of duration in the cut intervals

Usage

surv_split(
  x,
  duration,
  event,
  cuts,
  time_varying = NULL,
  label_episode = FALSE
)

Arguments

x	data frame
duration	exposure time
event	binary 0 and 1 coded
time_varying	not in use
label_episode	set to TRUE to have the episode labeled with actual cut intervals, otherwise episodes are integers
cut	the cut points on duration forming intervals [min_duration, first_value], (first_value, second_value], ..., (last_value, max_duration]

Details

The tested speed is 1.37 times to that of survSplit.

Examples

x <- data.frame(time = c(5, 10), event = c(1, 0))
surv_split(x, "time", "event", c(2,4,6))
surv_split(x, "time", "event", c(2,4,6), label_episode = TRUE)

kklot/ktools documentation built on Aug. 13, 2024, 7:08 p.m.