Simulate Survival Data from Piecewise Exponential Distribution

This function generates simulated survival data following a piecewise exponential distribution with specified time intervals and hazard rates.

Usage

simPE(
  n,
  time_points,
  hazard_rates,
  max_time,
  censoring_prob = 0.1,
  seed = NULL
)

Arguments

n: Sample size (number of subjects to simulate)
time_points: Numeric vector of time points defining intervals
hazard_rates: Numeric vector of hazard rates for each interval
max_time: Maximum follow-up time (administrative censoring)
censoring_prob: Probability of random censoring (default 0.1)
seed: Random seed for reproducibility (default NULL)

Value

Data frame with columns:

time: Observed time (either event time or censoring time)
status: Event indicator (1 = event, 0 = censored)

Details

The function simulates survival times using inverse transform sampling for piecewise exponential distribution. Random censoring is applied uniformly between 0 and the simulated survival time.

Examples

# Simulate data with 4 intervals
time_points <- c(0, 6, 12, 18, 24)
hazard_rates <- c(0.05, 0.08, 0.12, 0.15)

sim_data <- simPE(
  n = 200,
  time_points = time_points,
  hazard_rates = hazard_rates,
  max_time = 24,
  censoring_prob = 0.15,
  seed = 123
)

# Check event rate
table(sim_data$status)
#> 
#>   0   1 
#>  15 185 

# Reproducible results
sim_data1 <- simPE(n = 100, time_points, hazard_rates, 24, seed = 456)
sim_data2 <- simPE(n = 100, time_points, hazard_rates, 24, seed = 456)
identical(sim_data1, sim_data2)  # Should be TRUE
#> [1] TRUE