tests/testthat/test_shrinkDSM.R

In shrinkDSM: Efficient Bayesian Inference for Dynamic Survival Models with Shrinkage

library(shrinkDSM)
set.seed(42)
data("gastric")
intervals <- divisionpoints(gastric$time, gastric$status, 2)
test_that("divisionpoints", {
    expect_equal(divisionpoints(gastric$time, gastric$status, 1), unique(gastric$time[gastric$status==1]))
    expect_equal(intervals, unique(gastric$time[gastric$status==1])[c(FALSE, TRUE)])
    expect_equal(divisionpoints(gastric$time, gastric$status, 3), unique(gastric$time[gastric$status==1])[c(FALSE, FALSE, TRUE)])
})

mod <- shrinkDSM(time ~ radiation, gastric,
            delta = gastric$status, S = intervals, niter = 5)
mod2 <- shrinkDSM(time ~ radiation, gastric,
            delta = gastric$status, S = intervals,
            mod_type = "triple", niter = 5)
            # Change some of the hyperparameters
mod3 <- shrinkDSM(time ~ radiation, gastric,
            delta = gastric$status, S = intervals,
            mod_type = "triple",
            hyperprior_param = list(beta_a_xi = 5,
            alpha_a_xi = 10), niter = 5) 

mods <- list(mod, mod2, mod3)
# Estimate baseline model
test_that("Main shrinkDSM function",{
    for(idx in 1:3){
        m <- mods[[idx]]
        expect_s3_class(m, "shrinkDSM")
        expect_visible(m)
    }
})

test_that("print", {
    for(idx in 1:3){
        m <- mods[[idx]]
        expect_match(capture.output(print(m))[1], "Object containing a fitted DSM model with:") 
    }
})

test_that("Summary", {
    for(idx in 1:3){
        m <- mods[[idx]]
        expect_match(capture.output(summary(m))[2], "Summary of 3 MCMC draws after burn-in of 2")
    }
})

test_that("Plot parameters", {
# Plot piecewise constant, time-varying parameter
    for(idx in 1:3){
        m <- mods[[idx]]
        expect_invisible(plot(m))
        expect_invisible(plot(m$beta$beta_radiation))
        expect_invisible(plot(m$beta[[1]]))
        expect_invisible(plot(m$beta[["beta_radiation"]]))
    }
})

test_that("Prediction and predictive plots", {
    newdata <- data.frame(radiation = c(0, 1))
    for(idx in 1:3){
        m <- mods[[idx]]        
        p <- predict(m, newdata)
        expect_s3_class(p, "shrinkDSM_pred")
        expect_visible(p)
        expect_invisible(plot(p))
    }
})