tests/testthat/test_cif_aiptw_pseudo.r

In RobinDenz1/adjustedCurves: Confounder-Adjusted Survival Curves and Cumulative Incidence Functions

set.seed(42)

sim_dat <- readRDS(system.file("testdata",
                               "d_sim_crisk_n_50.Rds",
                               package="adjustedCurves"))
sim_dat$group <- as.factor(sim_dat$group)

# outcome model
outcome_vars <- c("x2", "x3", "x4", "x6")

# treatment model
treat_mod <- glm(group ~ x1 + x2 + x3 + x4 + x5 + x6, data=sim_dat,
                 family="binomial")
ps_score <- stats::predict(treat_mod, newdata=sim_dat, type="response")

test_that("2 treatments, no conf_int, no boot", {
  adj <- adjustedcif(data=sim_dat,
                     variable="group",
                     ev_time="time",
                     event="event",
                     method="aiptw_pseudo",
                     conf_int=FALSE,
                     outcome_vars=outcome_vars,
                     treatment_model=treat_mod,
                     type_time="bs",
                     cause=1)
  expect_s3_class(adj, "adjustedcif")
  expect_true(is.numeric(adj$adj$cif))
  expect_equal(levels(adj$adj$group), levels(sim_dat$group))
})

test_that("2 treatments, with conf_int, no boot", {
  adj <- adjustedcif(data=sim_dat,
                     variable="group",
                     ev_time="time",
                     event="event",
                     method="aiptw_pseudo",
                     conf_int=TRUE,
                     outcome_vars=outcome_vars,
                     treatment_model=treat_mod,
                     type_time="bs",
                     cause=1)
  expect_s3_class(adj, "adjustedcif")
  expect_true(is.numeric(adj$adj$cif))
  expect_equal(levels(adj$adj$group), levels(sim_dat$group))
})

test_that("2 treatments, no conf_int, with boot", {
  adj <- adjustedcif(data=sim_dat,
                     variable="group",
                     ev_time="time",
                     event="event",
                     method="aiptw_pseudo",
                     conf_int=FALSE,
                     bootstrap=TRUE,
                     n_boot=2,
                     outcome_vars=outcome_vars,
                     treatment_model=treat_mod,
                     type_time="bs",
                     cause=1)
  expect_s3_class(adj, "adjustedcif")
  expect_true(is.numeric(adj$adj$cif))
  expect_equal(levels(adj$adj$group), levels(sim_dat$group))
})

test_that("2 treatments, using propensity score directly", {
  adj <- adjustedcif(data=sim_dat,
                     variable="group",
                     ev_time="time",
                     event="event",
                     method="aiptw_pseudo",
                     conf_int=FALSE,
                     outcome_vars=outcome_vars,
                     treatment_model=ps_score,
                     type_time="bs",
                     cause=1)
  expect_s3_class(adj, "adjustedcif")
  expect_true(is.numeric(adj$adj$cif))
  expect_equal(levels(adj$adj$group), levels(sim_dat$group))
})

test_that("2 treatments, no conf_int, no boot, with times, factor time", {
  adj <- adjustedcif(data=sim_dat,
                     variable="group",
                     ev_time="time",
                     event="event",
                     method="aiptw_pseudo",
                     conf_int=FALSE,
                     bootstrap=FALSE,
                     n_boot=2,
                     outcome_vars=outcome_vars,
                     treatment_model=treat_mod,
                     times=c(0.5, 0.8, 1),
                     type_time="factor",
                     cause=1)
  expect_s3_class(adj, "adjustedcif")
  expect_true(is.numeric(adj$adj$cif))
  expect_equal(levels(adj$adj$group), levels(sim_dat$group))
})

test_that("2 treatments, no conf_int, no boot, with times, ns time", {
  adj <- adjustedcif(data=sim_dat,
                     variable="group",
                     ev_time="time",
                     event="event",
                     method="aiptw_pseudo",
                     conf_int=FALSE,
                     bootstrap=FALSE,
                     n_boot=2,
                     outcome_vars=outcome_vars,
                     treatment_model=treat_mod,
                     times=c(0.5, 0.8, 1),
                     type_time="ns",
                     spline_df=2,
                     cause=1)
  expect_s3_class(adj, "adjustedcif")
  expect_true(is.numeric(adj$adj$cif))
  expect_equal(levels(adj$adj$group), levels(sim_dat$group))
})

# new data
sim_dat <- readRDS(system.file("testdata",
                               "d_sim_crisk_n_50.Rds",
                               package="adjustedCurves"))
sim_dat$group[sim_dat$group==1] <- sample(c(1, 2),
                                        size=nrow(sim_dat[sim_dat$group==1, ]),
                                        replace=TRUE)
sim_dat$group <- as.factor(sim_dat$group)

# treatment model
treat_mod <- quiet(nnet::multinom(group ~ x1 + x2, data=sim_dat))

test_that("> 2 treatments, no conf_int, no boot", {
  adj <- adjustedcif(data=sim_dat,
                     variable="group",
                     ev_time="time",
                     event="event",
                     method="aiptw_pseudo",
                     conf_int=FALSE,
                     outcome_vars=outcome_vars,
                     treatment_model=treat_mod,
                     type_time="bs",
                     cause=1)
  expect_s3_class(adj, "adjustedcif")
  expect_true(is.numeric(adj$adj$cif))
  expect_equal(levels(adj$adj$group), levels(sim_dat$group))
})

test_that("> 2 treatments, with conf_int, no boot", {
  adj <- adjustedcif(data=sim_dat,
                     variable="group",
                     ev_time="time",
                     event="event",
                     method="aiptw_pseudo",
                     conf_int=TRUE,
                     outcome_vars=outcome_vars,
                     treatment_model=treat_mod,
                     type_time="bs",
                     cause=1)
  expect_s3_class(adj, "adjustedcif")
  expect_true(is.numeric(adj$adj$cif))
  expect_equal(levels(adj$adj$group), levels(sim_dat$group))
})

test_that("> 2 treatments, no conf_int, with boot", {
  adj <- adjustedcif(data=sim_dat,
                     variable="group",
                     ev_time="time",
                     event="event",
                     method="aiptw_pseudo",
                     conf_int=FALSE,
                     bootstrap=TRUE,
                     n_boot=2,
                     outcome_vars=outcome_vars,
                     treatment_model=treat_mod,
                     type_time="bs",
                     cause=1)
  expect_s3_class(adj, "adjustedcif")
  expect_true(is.numeric(adj$adj$cif))
  expect_equal(levels(adj$adj$group), levels(sim_dat$group))
})

test_that("> 2 treatments, no conf_int, no boot, with times, factor time", {
  adj <- adjustedcif(data=sim_dat,
                     variable="group",
                     ev_time="time",
                     event="event",
                     method="aiptw_pseudo",
                     conf_int=FALSE,
                     bootstrap=FALSE,
                     n_boot=2,
                     outcome_vars=outcome_vars,
                     treatment_model=treat_mod,
                     times=c(0.5, 0.8, 1),
                     type_time="factor",
                     cause=1)
  expect_s3_class(adj, "adjustedcif")
  expect_true(is.numeric(adj$adj$cif))
  expect_equal(levels(adj$adj$group), levels(sim_dat$group))
})