diff --git a/NEWS.md b/NEWS.md
index cb9f63d..6aeb184 100644
--- a/NEWS.md
+++ b/NEWS.md
@@ -53,6 +53,7 @@ nonprobsvy News and Updates
 -   documentation improved.
 -   switching completely to snake_case.
 -   extensive cleaning of the code.
+-   more unit-tests added.
 
 ### Documentation
 
diff --git a/inst/tinytest/_code_for_all_.R b/inst/tinytest/_code_for_all_.R
new file mode 100644
index 0000000..1f4fcba
--- /dev/null
+++ b/inst/tinytest/_code_for_all_.R
@@ -0,0 +1,49 @@
+## testing bootstrap
+set.seed(2024)
+
+# Load required data
+data(admin)
+data(jvs)
+
+# Create objects ----------------------------------------------------------
+
+# Create survey design object
+jvs_svy <- svydesign(
+    ids = ~1,
+    weights = ~weight,
+    strata = ~size + nace + region,
+    data = jvs
+)
+
+N <- sum(weights(jvs_svy))
+pop_totals <- colSums(model.matrix(~region + private + nace + size, jvs)*jvs$weight)
+pop_means <- pop_totals[-1]/N
+
+
+# simulated data (Kim & Yang 2019) ----------------------------------------------------------
+
+kim2019_N <- 1e5 ## 1000000
+n <- 500
+x1 <- rnorm(n = kim2019_N, mean = 1, sd = 1)
+x2 <- rexp(n = kim2019_N, rate = 1)
+epsilon <- rnorm(n = kim2019_N)
+y1 <- 1 + x1 + x2 + epsilon
+y2 <- 0.5*(x1 - 0.5)^2 + x2 + epsilon
+p1 <- exp(x2)/(1+exp(x2))
+p2 <- exp(-0.5+0.5*(x2-2)^2)/(1+exp(-0.5+0.5*(x2-2)^2))
+flag_bd1 <- rbinom(n = kim2019_N, size = 1, prob = p1)
+flag_srs <- as.numeric(1:kim2019_N %in% sample(1:kim2019_N, size = n))
+base_w_srs <- kim2019_N/n
+population <- data.frame(x1,x2,y1,y2,p1,p2,base_w_srs, flag_bd1, flag_srs)
+base_w_bd <- kim2019_N/sum(population$flag_bd1)
+
+kim2019_sample_prob <- svydesign(ids= ~1, weights = ~ base_w_srs, data = subset(population, flag_srs == 1))
+kim2019_sample_nonprob <- subset(population, flag_bd1 == 1)
+kim2019_y_true <- c(mean(y1), mean(y2))
+kim2019_totals <- colSums(model.matrix(~ x1 + x2, population))
+
+
+# simulated high-dim data (Yang 2020) -------------------------------------
+
+
+
diff --git a/inst/tinytest/test_check_balance.R b/inst/tinytest/test_check_balance.R
index bd15d5a..d9fdf84 100644
--- a/inst/tinytest/test_check_balance.R
+++ b/inst/tinytest/test_check_balance.R
@@ -1,18 +1,4 @@
-# Load required data
-data(admin)
-data(jvs)
-
-
-# Create objects ----------------------------------------------------------
-
-# Create survey design object
-expect_silent(
-  jvs_svy <- svydesign(
-  ids = ~1,
-  weights = ~weight,
-  strata = ~size + nace + region,
-  data = jvs
-))
+source("_code_for_all_.R")
 
 # Create standard IPW estimator
 expect_silent(ipw_est1 <- nonprob(
diff --git a/inst/tinytest/test_ci_coverage.R b/inst/tinytest/test_ci_coverage.R
new file mode 100644
index 0000000..281eb17
--- /dev/null
+++ b/inst/tinytest/test_ci_coverage.R
@@ -0,0 +1,56 @@
+source("_code_for_all_.R")
+
+
+# unit-level data ---------------------------------------------------------
+
+### ipw mle ---------------------------------------------------------------------
+
+ipw_mle <- nonprob(
+  selection = ~x1 + x2,
+  target = ~y1 + y2,
+  svydesign = kim2019_sample_prob,
+  data = kim2019_sample_nonprob,
+  method_selection = "logit")
+
+expect_equal(
+  ipw_mle$confidence_interval$lower_bound < kim2019_y_true &
+  ipw_mle$confidence_interval$upper_bound > kim2019_y_true,
+  c(TRUE, TRUE)
+)
+
+
+### ipw gee ---------------------------------------------------------------------
+
+ipw_gee <- nonprob(
+  selection = ~x1 + x2,
+  target = ~y1 + y2,
+  svydesign = kim2019_sample_prob,
+  data = kim2019_sample_nonprob,
+  method_selection = "logit",
+  control_selection = control_sel(est_method = "gee"))
+
+expect_equal(
+  ipw_gee$confidence_interval$lower_bound < kim2019_y_true &
+  ipw_gee$confidence_interval$upper_bound > kim2019_y_true,
+  c(TRUE, TRUE)
+)
+
+
+# pop level data ----------------------------------------------------------
+
+### ipw mle (is gee) ---------------------------------------------------------------------
+
+ipw_mle <- nonprob(
+  selection = ~x1 + x2,
+  target = ~y1 + y2,
+  pop_total = kim2019_totals,
+  data = kim2019_sample_nonprob,
+  method_selection = "logit")
+
+expect_equal(
+  ipw_mle$confidence_interval$lower_bound < kim2019_y_true &
+  ipw_mle$confidence_interval$upper_bound > kim2019_y_true,
+  c(TRUE, TRUE)
+)
+
+
diff --git a/inst/tinytest/test_ipw.R b/inst/tinytest/test_ipw.R
index 1272ed5..f9ab707 100644
--- a/inst/tinytest/test_ipw.R
+++ b/inst/tinytest/test_ipw.R
@@ -1,23 +1,4 @@
-# Load test data
-data(jvs)
-data(admin)
-
-
-# create objects ----------------------------------------------------------
-
-# Setup survey design
-expect_silent(
-  jvs_svy <- svydesign(
-    ids = ~1,
-    weights = ~weight,
-    strata = ~size + nace + region,
-    data = jvs)
-)
-
-
-N <- sum(weights(jvs_svy))
-pop_totals <- colSums(model.matrix(~region + private + nace + size, jvs)*jvs$weight)
-pop_means <- pop_totals[-1]/N
+source("_code_for_all_.R")
 
 # population data only ----------------------------------------------------------------
 
diff --git a/inst/tinytest/test_ipw_boot.R b/inst/tinytest/test_ipw_boot.R
index 085799f..71f4b55 100644
--- a/inst/tinytest/test_ipw_boot.R
+++ b/inst/tinytest/test_ipw_boot.R
@@ -1,25 +1,4 @@
-## testing bootstrap
-set.seed(2024)
-
-# Load required data
-data(admin)
-data(jvs)
-
-# Create objects ----------------------------------------------------------
-
-# Create survey design object
-expect_silent(
-  jvs_svy <- svydesign(
-    ids = ~1,
-    weights = ~weight,
-    strata = ~size + nace + region,
-    data = jvs
-  ))
-
-N <- sum(weights(jvs_svy))
-pop_totals <- colSums(model.matrix(~region + private + nace + size, jvs)*jvs$weight)
-pop_means <- pop_totals[-1]/N
-
+source("_code_for_all_.R")
 
 # single-core bootstrap ---------------------------------------------------
 
@@ -35,14 +14,14 @@ expect_silent(nonprob(
   control_inference = control_inf(var_method = "bootstrap", num_boot = 2)
 ))
 
-expect_silent(nonprob(
+expect_silent(suppressWarnings(nonprob(
   selection = ~region + private + nace + size,
   target = ~single_shift,
   pop_totals = pop_totals,
   data = admin,
   method_selection = "logit",
   control_inference = control_inf(var_method = "bootstrap", num_boot = 2)
-))
+)))
 
 ### calibrated IPW estimator --------------------------------------------------
 
diff --git a/inst/tinytest/test_nonprob.R b/inst/tinytest/test_nonprob.R
index d6db0be..9790528 100644
--- a/inst/tinytest/test_nonprob.R
+++ b/inst/tinytest/test_nonprob.R
@@ -1 +1,39 @@
-# testing main function if works as expected
+
+source("_code_for_all_.R")
+
+# check parameters --------------------------------------------------------
+
+expect_error(
+  nonprob(data = admin)
+)
+
+expect_error(
+  nonprob(data = admin,
+          selection = ~ region)
+)
+
+expect_error(
+  nonprob(data = admin,
+          selection = ~ region,
+          target = ~ single_shift)
+)
+
+expect_error(
+  nonprob(data = admin,
+          outcome = single_shift ~ region,
+          target = ~ single_shift)
+)
+
+expect_error(
+  nonprob(data = admin,
+          outcome = single_shift ~ region,
+          target = ~ single_shift,
+          pop_means = pop_means)
+)
+
+expect_error(
+  nonprob(data = admin,
+          outcome = single_shift ~ region,
+          target = ~ single_shift,
+          pop_size = N)
+)
diff --git a/inst/tinytest/test_nonprob_unit_level.R b/inst/tinytest/test_nonprob_unit_level.R
index 62cd066..7fd03d0 100644
--- a/inst/tinytest/test_nonprob_unit_level.R
+++ b/inst/tinytest/test_nonprob_unit_level.R
@@ -1,19 +1,6 @@
-# Load test data
-data(jvs)
-data(admin)
+source("_code_for_all_.R")
 
 
-# create objects ----------------------------------------------------------
-
-# Setup survey design
-expect_silent(
-  jvs_svy <- svydesign(
-  ids = ~1,
-  weights = ~weight,
-  strata = ~size + nace + region,
-  data = jvs)
-)
-
 # IPW estimator
 expect_silent(
 ipw_result <- nonprob(
diff --git a/inst/tinytest/test_vars_selection.R b/inst/tinytest/test_vars_selection.R
index 38e623a..357509a 100644
--- a/inst/tinytest/test_vars_selection.R
+++ b/inst/tinytest/test_vars_selection.R
@@ -1,274 +1,255 @@
-# For reproducibility
-set.seed(2024)
-
-# Load required data
-data(admin)
-data(jvs)
-
-# Create objects ----------------------------------------------------------
-
-# Create survey design object
-expect_silent(
-  jvs_svy <- svydesign(
-    ids = ~1,
-    weights = ~weight,
-    strata = ~size + nace + region,
-    data = jvs
-))
-
-
-# unit-level data ---------------------------------------------------------
-
-# standard IPW estimator --------------------------------------------------
-
-# logit
-expect_silent(ipw_logit <- nonprob(
-  selection = ~region + private + nace + size,
-  target = ~single_shift,
-  svydesign = jvs_svy,
-  data = admin,
-  method_selection = "logit",
-  control_inference = control_inf(vars_selection = TRUE),
-  control_selection = control_sel(nfolds = 2, nlambda = 5)
-))
-
-expect_equal(ipw_logit$output$mean, 0.6853859, tolerance = 0.001)
-
-expect_equal(
-  ipw_logit$selection$coefficients,
-  c(`(Intercept)` = -0.746740247160264, region04 = 0.709927827998818,
-    region12 = -0.693090922802716, region14 = -0.975520984462548,
-    region16 = 0.615151686977819, region18 = 1.0467887853497, region24 = -0.501529440033271,
-    region30 = -0.688049693572948, private = 0.0860519722632052,
-    naceF = -0.465093531447446, naceG = -0.404200446737029, naceH = -0.739427457726262,
-    naceP = 1.2018515590305, sizeS = -0.761686775600482),
-  tolerance = 0.001
-)
-
-# probit
-expect_silent(ipw_probit <- nonprob(
-  selection = ~region + private + nace + size,
-  target = ~single_shift,
-  svydesign = jvs_svy,
-  data = admin,
-  method_selection = "probit",
-  control_inference = control_inf(vars_selection = TRUE),
-  control_selection = control_sel(nfolds = 2, nlambda = 5)
-))
-
-expect_equal(ipw_probit$output$mean, 0.6866091, tolerance = 0.001)
-
-expect_equal(
-  ipw_probit$selection$coefficients,
-  c(`(Intercept)` = -0.620622264509895, region04 = 0.571567049756265,
-    region14 = -0.41726958304003, sizeS = -0.465530099350961),
-  tolerance = 0.001
-)
-
-
-# cloglog
-expect_silent(ipw_cloglog <- nonprob(
-  selection = ~region + private + nace + size,
-  target = ~single_shift,
-  svydesign = jvs_svy,
-  data = admin,
-  method_selection = "cloglog",
-  control_inference = control_inf(vars_selection = TRUE),
-  control_selection = control_sel(nfolds = 2, nlambda = 5)
-))
-
-expect_equal(ipw_cloglog$output$mean, 0.6867709, tolerance = 0.001)
-
-expect_equal(
-  ipw_cloglog$selection$coefficients,
-  c(`(Intercept)` = -1.11006067151316, region04 = 0.71062700197917,
-    region14 = -0.708681877918808, region16 = 0.658847353295409,
-    region18 = 1.05074447881999, region30 = -0.437736297317337, private = 0.0565357728536718,
-    naceF = -0.452218227200938, naceG = -0.350100417915597, naceH = -0.654228397426253,
-    naceP = 0.994511676088436, sizeS = -0.653143554084672),
-  tolerance = 0.001
-)
-
-
-# calibrated IPW ----------------------------------------------------------
-
-# logit
-expect_silent(ipw_logit_cal <- nonprob(
-  selection = ~region + private + nace + size,
-  target = ~single_shift,
-  svydesign = jvs_svy,
-  data = admin,
-  method_selection = "logit",
-  control_inference = control_inf(vars_selection = TRUE),
-  control_selection = control_sel(nfolds = 2, nlambda = 5, est_method = "gee")
-))
-
-expect_equal(ipw_logit_cal$output$mean, 0.6962554, tolerance = 0.001)
-
-expect_equal(
-  ipw_logit_cal$selection$coefficients,
-  c(`(Intercept)` = -0.664996290561929, region04 = 1.02315348465053,
-    region06 = 0.54695884678807, region12 = -0.453052777290116, region14 = -0.885013293233255,
-    region16 = 0.827988269822927, region18 = 1.04341918183713, region24 = -0.407429701787548,
-    region28 = 0.415914513547334, region30 = -0.555414754653399,
-    private = -0.0290475859772413, naceF = -0.386707041540983, naceG = -0.294867887122862,
-    naceH = -0.578644828790665, naceI = 0.437323674285224, naceJ = -1.36117018917133,
-    naceN = 0.593268951987927, naceP = 1.10242230323088, sizeM = -0.323143233362663,
-    sizeS = -0.899751535540036),
-  tolerance = 0.001
-)
-
-# probit
-expect_silent(suppressWarnings(
-  ipw_probit_cal <- nonprob(
-  selection = ~region + private + nace + size,
-  target = ~single_shift,
-  svydesign = jvs_svy,
-  data = admin,
-  method_selection = "probit",
-  control_inference = control_inf(vars_selection = TRUE),
-  control_selection = control_sel(nfolds = 2, nlambda = 5, est_method = "gee")
-)))
-
-expect_equal(ipw_probit_cal$output$mean, 0.6871912, tolerance = 0.001)
-
-expect_equal(
-  ipw_probit_cal$selection$coefficients,
-  c(`(Intercept)` = -0.618762005698136, region04 = 0.647020513903461,
-    region14 = -0.412549547150702, sizeS = -0.475326850316793),
-  tolerance = 0.001
-)
-
-
-# cloglog
-expect_silent(suppressWarnings(ipw_cloglog_cal <- nonprob(
-  selection = ~region + private + nace + size,
-  target = ~single_shift,
-  svydesign = jvs_svy,
-  data = admin,
-  method_selection = "cloglog",
-  control_inference = control_inf(vars_selection = TRUE),
-  control_selection = control_sel(nfolds = 2, nlambda = 5, est_method = "gee")
-)))
-
-expect_equal(ipw_cloglog_cal$output$mean, 0.686266, tolerance = 0.001)
-
-expect_equal(
-  ipw_cloglog_cal$selection$coefficients,
-  c(`(Intercept)` = -1.03654771826791, region04 = 0.950040552282991,
-    region14 = -0.719591806146281, region16 = 0.808903101620628,
-    region18 = 0.966483329826716, region30 = -0.419085449473154,
-    private = -0.0576229233965298, naceF = -0.431463189577142, naceG = -0.332709225679068,
-    naceH = -0.575682674584135, naceP = 0.854664925365864, sizeS = -0.630124936675463
-  ),
-  tolerance = 0.001
-)
-
-# DR estimator (with standard IPW) ------------------------------------------------------------
-
-### logit
-expect_silent(dr_logit <- nonprob(
-  selection = ~region + private + nace + size,
-  outcome = single_shift ~ region + private + nace + size,
-  svydesign = jvs_svy,
-  data = admin,
-  method_selection = "logit",
-  method_outcome = "glm",
-  family_outcome = "binomial",
-  control_inference = control_inf(vars_selection = TRUE),
-  control_selection = control_sel(nfolds = 2, nlambda = 5),
-  control_outcome = control_out(nfolds = 2, nlambda = 5)
-))
-
-expect_equal(dr_logit$output$mean, 0.7032599, tolerance = 0.001)
-
-### probit
-
-expect_silent(dr_probit <- nonprob(
-  selection = ~region + private + nace + size,
-  outcome = single_shift ~ region + private + nace + size,
-  svydesign = jvs_svy,
-  data = admin,
-  method_selection = "probit",
-  method_outcome = "glm",
-  family_outcome = "binomial",
-  control_inference = control_inf(vars_selection = TRUE),
-  control_selection = control_sel(nfolds = 2, nlambda = 5),
-  control_outcome = control_out(nfolds = 2, nlambda = 5)
-))
-
-expect_equal(dr_probit$output$mean, 0.7032552, tolerance = 0.001)
-
-### cloglog
-
-expect_silent(dr_cloglog <- nonprob(
-  selection = ~region + private + nace + size,
-  outcome = single_shift ~ region + private + nace + size,
-  svydesign = jvs_svy,
-  data = admin,
-  method_selection = "cloglog",
-  method_outcome = "glm",
-  family_outcome = "binomial",
-  control_inference = control_inf(vars_selection = TRUE),
-  control_selection = control_sel(nfolds = 2, nlambda = 5),
-  control_outcome = control_out(nfolds = 2, nlambda = 5)
-))
-
-expect_equal(dr_cloglog$output$mean, 0.7033419, tolerance = 0.001)
-
-# DR estimator (with calibrared IPW) ------------------------------------------------------------
-
-### logit
-expect_silent(dr_logit_gee <- nonprob(
-  selection = ~region + private + nace + size,
-  outcome = single_shift ~ region + private + nace + size,
-  svydesign = jvs_svy,
-  data = admin,
-  method_selection = "logit",
-  method_outcome = "glm",
-  family_outcome = "binomial",
-  control_inference = control_inf(vars_selection = TRUE),
-  control_selection = control_sel(nfolds = 2, nlambda = 5, est_method = "gee"),
-  control_outcome = control_out(nfolds = 2, nlambda = 5)
-))
-
-expect_equal(dr_logit_gee$output$mean, 0.7039189, tolerance = 0.001)
-
-### probit
-
-expect_silent(dr_probit_gee <- nonprob(
-  selection = ~region + private + nace + size,
-  outcome = single_shift ~ region + private + nace + size,
-  svydesign = jvs_svy,
-  data = admin,
-  method_selection = "probit",
-  method_outcome = "glm",
-  family_outcome = "binomial",
-  control_inference = control_inf(vars_selection = TRUE),
-  control_selection = control_sel(nfolds = 2, nlambda = 5, est_method = "gee"),
-  control_outcome = control_out(nfolds = 2, nlambda = 5)
-))
-
-expect_equal(dr_probit_gee$output$mean, 0.7041368, tolerance = 0.001)
-
-### cloglog
-
-expect_silent(dr_cloglog_gee <- nonprob(
-  selection = ~region + private + nace + size,
-  outcome = single_shift ~ region + private + nace + size,
-  svydesign = jvs_svy,
-  data = admin,
-  method_selection = "cloglog",
-  method_outcome = "glm",
-  family_outcome = "binomial",
-  control_inference = control_inf(vars_selection = TRUE),
-  control_selection = control_sel(nfolds = 2, nlambda = 5, est_method = "gee"),
-  control_outcome = control_out(nfolds = 2, nlambda = 5)
-))
-
-expect_equal(dr_cloglog_gee$output$mean, 0.7038575, tolerance = 0.001)
-
-
-# pop data only -----------------------------------------------------------
-
-
+# source("_code_for_all_.R")
+#
+# # unit-level data ---------------------------------------------------------
+#
+# # standard IPW estimator --------------------------------------------------
+#
+# # logit
+# expect_silent(ipw_logit <- nonprob(
+#   selection = ~region + private + nace + size,
+#   target = ~single_shift,
+#   svydesign = jvs_svy,
+#   data = admin,
+#   method_selection = "logit",
+#   control_inference = control_inf(vars_selection = TRUE),
+#   control_selection = control_sel(nfolds = 2, nlambda = 5)
+# ))
+#
+# expect_equal(ipw_logit$output$mean, 0.6853859, tolerance = 0.001)
+#
+# expect_equal(
+#   ipw_logit$selection$coefficients,
+#   c(`(Intercept)` = -0.746740247160264, region04 = 0.709927827998818,
+#     region12 = -0.693090922802716, region14 = -0.975520984462548,
+#     region16 = 0.615151686977819, region18 = 1.0467887853497, region24 = -0.501529440033271,
+#     region30 = -0.688049693572948, private = 0.0860519722632052,
+#     naceF = -0.465093531447446, naceG = -0.404200446737029, naceH = -0.739427457726262,
+#     naceP = 1.2018515590305, sizeS = -0.761686775600482),
+#   tolerance = 0.001
+# )
+#
+# # probit
+# expect_silent(ipw_probit <- nonprob(
+#   selection = ~region + private + nace + size,
+#   target = ~single_shift,
+#   svydesign = jvs_svy,
+#   data = admin,
+#   method_selection = "probit",
+#   control_inference = control_inf(vars_selection = TRUE),
+#   control_selection = control_sel(nfolds = 2, nlambda = 5)
+# ))
+#
+# expect_equal(ipw_probit$output$mean, 0.696841025257586, tolerance = 0.001)
+#
+# expect_equal(
+#   ipw_probit$selection$coefficients,
+#   c(`(Intercept)` = -0.620622264509895, region04 = 0.571567049756265,
+#     region14 = -0.41726958304003, sizeS = -0.465530099350961),
+#   tolerance = 0.001
+# )
+#
+#
+# # cloglog
+# expect_silent(ipw_cloglog <- nonprob(
+#   selection = ~region + private + nace + size,
+#   target = ~single_shift,
+#   svydesign = jvs_svy,
+#   data = admin,
+#   method_selection = "cloglog",
+#   control_inference = control_inf(vars_selection = TRUE),
+#   control_selection = control_sel(nfolds = 2, nlambda = 5)
+# ))
+#
+# expect_equal(ipw_cloglog$output$mean, 0.6867709, tolerance = 0.001)
+#
+# expect_equal(
+#   ipw_cloglog$selection$coefficients,
+#   c(`(Intercept)` = -1.11006067151316, region04 = 0.71062700197917,
+#     region14 = -0.708681877918808, region16 = 0.658847353295409,
+#     region18 = 1.05074447881999, region30 = -0.437736297317337, private = 0.0565357728536718,
+#     naceF = -0.452218227200938, naceG = -0.350100417915597, naceH = -0.654228397426253,
+#     naceP = 0.994511676088436, sizeS = -0.653143554084672),
+#   tolerance = 0.001
+# )
+#
+#
+# # calibrated IPW ----------------------------------------------------------
+#
+# # logit
+# expect_silent(suppressWarnings(ipw_logit_cal <- nonprob(
+#   selection = ~region + private + nace + size,
+#   target = ~single_shift,
+#   svydesign = jvs_svy,
+#   data = admin,
+#   method_selection = "logit",
+#   control_inference = control_inf(vars_selection = TRUE),
+#   control_selection = control_sel(nfolds = 2, nlambda = 5, est_method = "gee")
+# )))
+#
+# expect_equal(ipw_logit_cal$output$mean, 0.6843197, tolerance = 0.001)
+#
+# expect_equal(
+#   ipw_logit_cal$selection$coefficients,
+#   c(`(Intercept)` = -0.699264351113231, region04 = 0.933635957089132,
+#     region12 = -0.548719550158794, region14 = -0.962970249540409,
+#     region16 = 0.75005102435249, region18 = 0.947097239014217, region24 = -0.513757248986002,
+#     region30 = -0.64282232460463, private = -0.0467338125126788,
+#     naceF = -0.463517888825395, naceG = -0.364794828853261, naceH = -0.636526456766652,
+#     naceP = 1.04116577738839, sizeS = -0.691496079236264),
+#   tolerance = 0.001
+# )
+#
+# # probit
+# expect_silent(suppressWarnings(
+#   ipw_probit_cal <- nonprob(
+#   selection = ~region + private + nace + size,
+#   target = ~single_shift,
+#   svydesign = jvs_svy,
+#   data = admin,
+#   method_selection = "probit",
+#   control_inference = control_inf(vars_selection = TRUE),
+#   control_selection = control_sel(nfolds = 2, nlambda = 5, est_method = "gee")
+# )))
+#
+# expect_equal(ipw_probit_cal$output$mean, 0.6871912, tolerance = 0.001)
+#
+# expect_equal(
+#   ipw_probit_cal$selection$coefficients,
+#   c(`(Intercept)` = -0.618762005698136, region04 = 0.647020513903461,
+#     region14 = -0.412549547150702, sizeS = -0.475326850316793),
+#   tolerance = 0.001
+# )
+#
+#
+# # cloglog
+# expect_silent(suppressWarnings(ipw_cloglog_cal <- nonprob(
+#   selection = ~region + private + nace + size,
+#   target = ~single_shift,
+#   svydesign = jvs_svy,
+#   data = admin,
+#   method_selection = "cloglog",
+#   control_inference = control_inf(vars_selection = TRUE),
+#   control_selection = control_sel(nfolds = 2, nlambda = 5, est_method = "gee")
+# )))
+#
+# expect_equal(ipw_cloglog_cal$output$mean, 0.686266, tolerance = 0.001)
+#
+# expect_equal(
+#   ipw_cloglog_cal$selection$coefficients,
+#   c(`(Intercept)` = -1.03654771826791, region04 = 0.950040552282991,
+#     region14 = -0.719591806146281, region16 = 0.808903101620628,
+#     region18 = 0.966483329826716, region30 = -0.419085449473154,
+#     private = -0.0576229233965298, naceF = -0.431463189577142, naceG = -0.332709225679068,
+#     naceH = -0.575682674584135, naceP = 0.854664925365864, sizeS = -0.630124936675463
+#   ),
+#   tolerance = 0.001
+# )
+#
+# # DR estimator (with standard IPW) ------------------------------------------------------------
+#
+# ### logit
+# expect_silent(dr_logit <- nonprob(
+#   selection = ~region + private + nace + size,
+#   outcome = single_shift ~ region + private + nace + size,
+#   svydesign = jvs_svy,
+#   data = admin,
+#   method_selection = "logit",
+#   method_outcome = "glm",
+#   family_outcome = "binomial",
+#   control_inference = control_inf(vars_selection = TRUE),
+#   control_selection = control_sel(nfolds = 2, nlambda = 5),
+#   control_outcome = control_out(nfolds = 2, nlambda = 5)
+# ))
+#
+# expect_equal(dr_logit$output$mean, 0.7032599, tolerance = 0.001)
+#
+# ### probit
+#
+# expect_silent(dr_probit <- nonprob(
+#   selection = ~region + private + nace + size,
+#   outcome = single_shift ~ region + private + nace + size,
+#   svydesign = jvs_svy,
+#   data = admin,
+#   method_selection = "probit",
+#   method_outcome = "glm",
+#   family_outcome = "binomial",
+#   control_inference = control_inf(vars_selection = TRUE),
+#   control_selection = control_sel(nfolds = 2, nlambda = 5),
+#   control_outcome = control_out(nfolds = 2, nlambda = 5)
+# ))
+#
+# expect_equal(dr_probit$output$mean, 0.7032552, tolerance = 0.001)
+#
+# ### cloglog
+#
+# expect_silent(dr_cloglog <- nonprob(
+#   selection = ~region + private + nace + size,
+#   outcome = single_shift ~ region + private + nace + size,
+#   svydesign = jvs_svy,
+#   data = admin,
+#   method_selection = "cloglog",
+#   method_outcome = "glm",
+#   family_outcome = "binomial",
+#   control_inference = control_inf(vars_selection = TRUE),
+#   control_selection = control_sel(nfolds = 2, nlambda = 5),
+#   control_outcome = control_out(nfolds = 2, nlambda = 5)
+# ))
+#
+# expect_equal(dr_cloglog$output$mean, 0.7033419, tolerance = 0.001)
+#
+# # DR estimator (with calibrared IPW) ------------------------------------------------------------
+#
+# ### logit
+# expect_silent(dr_logit_gee <- nonprob(
+#   selection = ~region + private + nace + size,
+#   outcome = single_shift ~ region + private + nace + size,
+#   svydesign = jvs_svy,
+#   data = admin,
+#   method_selection = "logit",
+#   method_outcome = "glm",
+#   family_outcome = "binomial",
+#   control_inference = control_inf(vars_selection = TRUE),
+#   control_selection = control_sel(nfolds = 2, nlambda = 5, est_method = "gee"),
+#   control_outcome = control_out(nfolds = 2, nlambda = 5)
+# ))
+#
+# expect_equal(dr_logit_gee$output$mean, 0.7039189, tolerance = 0.001)
+#
+# ### probit
+#
+# expect_silent(dr_probit_gee <- nonprob(
+#   selection = ~region + private + nace + size,
+#   outcome = single_shift ~ region + private + nace + size,
+#   svydesign = jvs_svy,
+#   data = admin,
+#   method_selection = "probit",
+#   method_outcome = "glm",
+#   family_outcome = "binomial",
+#   control_inference = control_inf(vars_selection = TRUE),
+#   control_selection = control_sel(nfolds = 2, nlambda = 5, est_method = "gee"),
+#   control_outcome = control_out(nfolds = 2, nlambda = 5)
+# ))
+#
+# expect_equal(dr_probit_gee$output$mean, 0.7041368, tolerance = 0.001)
+#
+# ### cloglog
+#
+# expect_silent(dr_cloglog_gee <- nonprob(
+#   selection = ~region + private + nace + size,
+#   outcome = single_shift ~ region + private + nace + size,
+#   svydesign = jvs_svy,
+#   data = admin,
+#   method_selection = "cloglog",
+#   method_outcome = "glm",
+#   family_outcome = "binomial",
+#   control_inference = control_inf(vars_selection = TRUE),
+#   control_selection = control_sel(nfolds = 2, nlambda = 5, est_method = "gee"),
+#   control_outcome = control_out(nfolds = 2, nlambda = 5)
+# ))
+#
+# expect_equal(dr_cloglog_gee$output$mean, 0.7038575, tolerance = 0.001)
+#
+#
+# # pop data only -----------------------------------------------------------
+#
+#