tests/testthat/test-double_ml_data_cluster.R

In DoubleML: Double Machine Learning in R

context("Unit tests for DoubleMLCluster (Additional tests)")

test_that("Unit tests for DoubleMLData", {

  set.seed(1234)
  N = 25 # number of observations (first dimension)
  M = 25 # number of observations (second dimension)
  dim_x = 10 # dimension of x
  data_one_way = make_pliv_multiway_cluster_CKMS2021(N, M, dim_x,
    omega_X = c(0.25, 0),
    omega_epsilon = c(0.25, 0),
    omega_v = c(0.25, 0),
    omega_V = c(0.25, 0))
  data_one_way$cluster_cols = "cluster_var_i"

  data_model = data_one_way$data_model
  data_model$S = rbinom(nrow(data_model), prob = 0.5, 1)

  # z_cols but no s_col and x_cols
  dml_data = DoubleMLClusterData$new(data_model,
    y_col = "Y",
    d_cols = "D",
    z_cols = "Z",
    cluster_cols = "cluster_var_i")
  expect_null(dml_data$s_col)
  expect_data_table(dml_data$data_model)
  x_cols_exp = c(paste0("X", 1:10), "S")
  expect_identical(dml_data$x_cols, x_cols_exp)

  # z_cols and s_col, but no x_cols (L. 502)
  dml_data = DoubleMLClusterData$new(data_model,
    y_col = "Y",
    d_cols = "D",
    s_col = "S",
    cluster_cols = "cluster_var_i")
  expect_data_table(dml_data$data_model)
  x_cols_exp = c(paste0("X", 1:10), "Z")
  expect_identical(dml_data$x_cols, x_cols_exp)

  msg = paste("At least one variable/column is set as selection variable",
    "\\('s_col'\\) and as a cluster variable \\('cluster_cols'\\).")

  expect_error(DoubleMLClusterData$new(data_model,
    x_cols = c("X1", "X2"),
    y_col = "Y",
    d_cols = "D",
    z_cols = "Z",
    s_col = "S",
    cluster_cols = "S"),
  regexp = msg)
})