shorten tests

tests/testthat/test-multistep_horizon.R

@@ -1,95 +1,3 @@
-
-test_that("multistep_horizon yearly data", {
-
-  # Mock data setup
-  data <- timetk::m4_yearly %>%
-    dplyr::mutate(id = as.character(id)) %>%
-    dplyr::rename(Date = date) %>%
-    dplyr::filter(id == "Y1")
-
-  run_info <- set_run_info()
-
-  # prep data and models
-  prep_data(
-    run_info = run_info,
-    input_data = data,
-    combo_variables = c("id"),
-    target_variable = "value",
-    date_type = "year",
-    forecast_horizon = 4,
-    recipes_to_run = "R1",
-    multistep_horizon = TRUE
-  )
-
-  prep_models(
-    run_info = run_info,
-    back_test_scenarios = 4,
-    models_to_run = "xgboost",
-    run_ensemble_models = FALSE,
-    num_hyperparameters = 1, 
-    pca = TRUE
-  )
-
-  # train models
-  train_models(run_info = run_info)
-
-  # pull trained model
-  workflow_tbl <- get_trained_models(run_info = run_info) %>%
-    dplyr::select(Model_Fit)
-
-  model_length <- length(workflow_tbl$Model_Fit[[1]]$fit$fit$fit$models)
-
-  # Assertions
-  expect_equal(model_length, 4)
-})
-
-test_that("multistep_horizon quarterly data", {
-
-  # Mock data setup
-  data <- timetk::m4_quarterly %>%
-    dplyr::mutate(id = as.character(id)) %>%
-    dplyr::rename(Date = date) %>%
-    dplyr::filter(id == "Q10")
-
-  run_info <- set_run_info()
-
-  # prep data and models
-  prep_data(
-    run_info = run_info,
-    input_data = data,
-    combo_variables = c("id"),
-    target_variable = "value",
-    date_type = "quarter",
-    forecast_horizon = 6,
-    recipes_to_run = "R1",
-    multistep_horizon = TRUE
-  )
-
-  prep_models(
-    run_info = run_info,
-    back_test_scenarios = 6,
-    models_to_run = "mars",
-    run_ensemble_models = FALSE,
-    num_hyperparameters = 1, 
-    pca = TRUE
-  )
-
-  # train models
-  train_models(
-    run_info = run_info,
-    feature_selection = TRUE
-  )
-
-  # pull trained model
-  workflow_tbl <- get_trained_models(run_info = run_info) %>%
-    dplyr::select(Model_Fit)
-
-  model_length <- length(workflow_tbl$Model_Fit[[1]]$fit$fit$fit$models)
-
-  # Assertions
-  expect_equal(model_length, 5)
-})
-
 test_that("multistep_horizon monthly data", {
 
   # Mock data setup
@@ -136,98 +44,3 @@ test_that("multistep_horizon monthly data", {
   # Assertions
   expect_equal(model_length, 2)
 })
-
-test_that("multistep_horizon weekly data", {
-
-  # Mock data setup
-  data <- timetk::m4_weekly %>%
-    dplyr::mutate(id = as.character(id)) %>%
-    dplyr::rename(Date = date) %>%
-    dplyr::filter(
-      id == "W10",
-      Date >= "2014-01-01"
-    )
-
-  run_info <- set_run_info()
-
-  # prep data and models
-  prep_data(
-    run_info = run_info,
-    input_data = data,
-    combo_variables = c("id"),
-    target_variable = "value",
-    date_type = "week",
-    forecast_horizon = 4,
-    recipes_to_run = "R1",
-    multistep_horizon = TRUE
-  )
-
-  prep_models(
-    run_info = run_info,
-    back_test_scenarios = 4,
-    models_to_run = "glmnet",
-    run_ensemble_models = FALSE,
-    num_hyperparameters = 1, 
-    pca = TRUE
-  )
-
-  # train models
-  train_models(run_info = run_info)
-
-  # pull trained model
-  workflow_tbl <- get_trained_models(run_info = run_info) %>%
-    dplyr::select(Model_Fit)
-
-  model_length <- length(workflow_tbl$Model_Fit[[1]]$fit$fit$fit$models)
-
-  # Assertions
-  expect_equal(model_length, 1)
-})
-
-test_that("multistep_horizon daily data", {
-
-  # Mock data setup
-  data <- timetk::m4_daily %>%
-    dplyr::mutate(id = as.character(id)) %>%
-    dplyr::rename(Date = date) %>%
-    dplyr::filter(
-      id == "D10",
-      Date >= "2014-01-01"
-    )
-
-  run_info <- set_run_info()
-
-  # prep data and models
-  prep_data(
-    run_info = run_info,
-    input_data = data,
-    combo_variables = c("id"),
-    target_variable = "value",
-    date_type = "day",
-    forecast_horizon = 30,
-    recipes_to_run = "R1",
-    multistep_horizon = TRUE
-  )
-
-  prep_models(
-    run_info = run_info,
-    back_test_scenarios = 4,
-    back_test_spacing = 7,
-    models_to_run = "glmnet",
-    run_ensemble_models = FALSE,
-    num_hyperparameters = 1, 
-    pca = TRUE
-  )
-
-  # train models
-  train_models(run_info = run_info)
-
-  # pull trained model
-  workflow_tbl <- get_trained_models(run_info = run_info) %>%
-    dplyr::select(Model_Fit)
-
-  model_length <- length(workflow_tbl$Model_Fit[[1]]$fit$fit$fit$models)
-
-  # Assertions
-  expect_equal(model_length, 2)
-})