The goal of simtte is to fast simulate simple and complex time-to-event models also known as survival analysis. It also allows user-defined time-to-event models via the awesome package mrgsolve.
You can install the released version of simtte from CRAN with:
install.packages("simtte")And the development version from GitHub with:
# install.packages("devtools")
devtools::install_github("csetraynor/simtte")This is a basic example which shows you how to solve a common problem:
library(simtte)
# Load M-splines data example
data("ms_data")
mu <- ms_data$mu
basis <- ms_data$basis
coefs <- ms_data$coefs
time <- ms_data$time
# Simulate prognostic index, linear predictor
lp <- matrix(runif(nrow(basis)), nrow = nrow(basis))
# simulate M-splines model
sim_tte(pi = lp, mu = mu, basis = basis, coefs = coefs, time = time, type = "ms", end_time = 100)
#> Compiling ms ... done.
#> # A tibble: 299 x 4
#> sim_time sim_status ID lp
#> <dbl> <dbl> <int> <dbl>
#> 1 1.03 1 1 0.406
#> 2 2.59 1 2 0.696
#> 3 6.73 0 3 0.840
#> 4 3.50 1 4 0.973
#> 5 6.73 0 5 0.935
#> 6 0.745 1 6 0.290
#> 7 0.562 1 7 0.594
#> 8 2.24 1 8 0.299
#> 9 6.73 0 9 0.706
#> 10 6.73 0 10 0.735
#> # … with 289 more rows
# simulate Weibull model
sim_tte(pi = lp, mu = -1, coefs = 1.1, time = time, type = "weibull", obs.only = F, obs.aug = T, delta = 0.05, end_time = 100)
#> Compiling weibull ... done.
#> # A tibble: 299 x 4
#> sim_time sim_status ID lp
#> <dbl> <dbl> <int> <dbl>
#> 1 4.3 1 1 0.406
#> 2 0.45 1 2 0.696
#> 3 0.75 1 3 0.840
#> 4 0.1 1 4 0.973
#> 5 1.5 1 5 0.935
#> 6 0.75 1 6 0.290
#> 7 1.65 1 7 0.594
#> 8 1.6 1 8 0.299
#> 9 1.45 1 9 0.706
#> 10 1.45 1 10 0.735
#> # … with 289 more rows