test_that("", {
  
  # Simulate a DGP using simFM
  set.seed(02102025)
  no_of_observations <- 200
  no_of_variables <- 10
  no_of_factors <- 3
  trans_error_var_cov <- diag(1, no_of_factors)
  loading_matrix <- matrix(round(rnorm(no_of_variables * no_of_factors)), no_of_variables, no_of_factors)
  meas_error_mean <- rep(0, no_of_variables)
  meas_error_var_cov <- diag(1, no_of_variables)
  trans_var_coeff <- cbind(diag(0.5, no_of_factors), -diag(0.25, no_of_factors))
  factor_lag_order <- 2 
  simul_delay <- c(floor(rexp(10, 1)))
  quarterfy <- FALSE
  quarterly_variable_ratio  <- 0
  corr <- TRUE
  beta_param <- 2
  seed <- 01102025
  set.seed(seed)
  burn_in <- 999
  starting_date <- "1970-01-01"
  rescale <- TRUE
  check_stationarity <- TRUE
  stationarity_check_threshold <- 1e-10
  
  FM_date <- simFM(no_of_observations = no_of_observations, no_of_variables = no_of_variables, 
                   no_of_factors = no_of_factors, loading_matrix = loading_matrix, 
                   meas_error_mean = meas_error_mean, meas_error_var_cov = meas_error_var_cov,
                   trans_error_var_cov = trans_error_var_cov, trans_var_coeff = trans_var_coeff, 
                   factor_lag_order = factor_lag_order, delay = simul_delay, quarterfy = quarterfy, 
                   quarterly_variable_ratio  = quarterly_variable_ratio, corr = corr, 
                   beta_param = beta_param, seed = seed, burn_in = burn_in, starting_date = starting_date,
                   rescale = rescale, check_stationarity = check_stationarity, 
                   stationarity_check_threshold = stationarity_check_threshold)
  
  FM_no_date <- simFM(no_of_observations = no_of_observations, no_of_variables = no_of_variables, 
                      no_of_factors = no_of_factors, loading_matrix = loading_matrix, 
                      meas_error_mean = meas_error_mean, meas_error_var_cov = meas_error_var_cov,
                      trans_error_var_cov = trans_error_var_cov, trans_var_coeff = trans_var_coeff, 
                      factor_lag_order = factor_lag_order, delay = simul_delay, quarterfy = quarterfy, 
                      quarterly_variable_ratio  = quarterly_variable_ratio, corr = corr, 
                      beta_param = beta_param, seed = seed, burn_in = burn_in, starting_date = NULL,
                      rescale = rescale, check_stationarity = check_stationarity, 
                      stationarity_check_threshold = stationarity_check_threshold)
  
  frequency <- FM_no_date$frequency
  delay <- simul_delay
  no_of_factors <- 3 
  max_factor_lag_order  <- 10
  decorr_errors <- TRUE 
  lag_estim_criterion  <- "BIC"
  comp_null <- 1e-15
  check_rank <- FALSE
  log <- FALSE
  parallel <- FALSE
  fcast_horizon <- 10
  
  # Minimal example
  
  # Test single factor
  expect_silent(twoStepDenseDFM(data = FM_date$data, delay = delay, no_of_factors = 1))
  
  # Parcing zoo object
  expect_silent(twoStepDenseDFM(data = FM_date$data, delay = delay, no_of_factors = no_of_factors))
  
  # Parcing matrix object
  expect_silent(twoStepDenseDFM(data = FM_no_date$data, delay = delay, no_of_factors = no_of_factors))
  
  fit_dense_date <- twoStepDenseDFM(data = FM_date$data, delay = delay, no_of_factors = no_of_factors,
                                    max_factor_lag_order  = max_factor_lag_order, 
                                    decorr_errors = decorr_errors,  lag_estim_criterion  = lag_estim_criterion, 
                                    comp_null = comp_null,  check_rank = check_rank, log = log, 
                                    parallel = parallel, fcast_horizon = fcast_horizon)
  
  fit_dense_no_date <- twoStepDenseDFM(data = FM_no_date$data, delay = delay, no_of_factors = no_of_factors,
                                       max_factor_lag_order  = max_factor_lag_order, 
                                       decorr_errors = decorr_errors,  lag_estim_criterion  = lag_estim_criterion, 
                                       comp_null = comp_null,  check_rank = check_rank, log = log, 
                                       parallel = parallel, fcast_horizon = fcast_horizon)
  
  # Basic checks
  expect_type(fit_dense_date, "list")
  expect_type(fit_dense_no_date, "list")
  expect_true(all(c("data", "loading_matrix_estimate", "smoothed_factors", "smoothed_state_variance",
                    "factor_var_lag_order", "error_var_cov_cholesky_factor")
                  %in% names(fit_dense_date)))
  expect_true(all(c("data", "loading_matrix_estimate", "smoothed_factors", "smoothed_state_variance",
                    "factor_var_lag_order", "error_var_cov_cholesky_factor")
                  %in% names(fit_dense_no_date)))
  expect_equal(dim(fit_dense_date$smoothed_factors), c(no_of_observations + fcast_horizon, no_of_factors))
  expect_equal(dim(fit_dense_no_date$smoothed_factors), c(no_of_factors, no_of_observations + fcast_horizon))
  fit_dense_coredata <- t(coredata(fit_dense_date$smoothed_factors))
  expect_equal(fit_dense_coredata, fit_dense_no_date$smoothed_factors, tolerance = 1e-10)
  expect_equal(fit_dense_date$loading_matrix_estimate, fit_dense_no_date$loading_matrix_estimate, tolerance = 1e-10)
  expect_true(is.zoo(fit_dense_date$smoothed_factors))
  expect_true(!is.zoo(fit_dense_no_date$smoothed_factors))
  
  # test information criteria
  expect_silent(twoStepDenseDFM(data = FM_no_date$data, delay = delay, no_of_factors = no_of_factors,
                                max_factor_lag_order  = max_factor_lag_order, 
                                decorr_errors = decorr_errors,  lag_estim_criterion  = "AIC", 
                                comp_null = comp_null,  check_rank = check_rank, log = log, 
                                parallel = parallel, fcast_horizon = fcast_horizon))
  expect_silent(twoStepDenseDFM(data = FM_no_date$data, delay = delay, no_of_factors = no_of_factors,
                                max_factor_lag_order  = max_factor_lag_order, 
                                decorr_errors = decorr_errors,  lag_estim_criterion  = "HIC", 
                                comp_null = comp_null,  check_rank = check_rank, log = log, 
                                parallel = parallel, fcast_horizon = fcast_horizon))
  
  # Test misshandling
  expect_error(twoStepDenseDFM(data = matrix(NaN, no_of_factors, no_of_observations), delay = delay, no_of_factors = no_of_factors,
                               max_factor_lag_order  = max_factor_lag_order, 
                               decorr_errors = decorr_errors,  lag_estim_criterion  = "HIC", 
                               comp_null = comp_null,  check_rank = check_rank, log = log, 
                               parallel = parallel, fcast_horizon = fcast_horizon))
  
  expect_error(twoStepDenseDFM(, delay = 1, no_of_factors = no_of_factors,
                               max_factor_lag_order  = max_factor_lag_order, 
                               decorr_errors = decorr_errors,  lag_estim_criterion  = "HIC", 
                               comp_null = comp_null,  check_rank = check_rank, log = log, 
                               parallel = parallel, fcast_horizon = fcast_horizon))
  
  expect_error(twoStepDenseDFM(data = FM_no_date$data, delay = delay, no_of_factors = -1,
                               max_factor_lag_order  = max_factor_lag_order, 
                               decorr_errors = decorr_errors,  lag_estim_criterion  = "HIC", 
                               comp_null = comp_null,  check_rank = check_rank, log = log, 
                               parallel = parallel, fcast_horizon = fcast_horizon))
  
  expect_error(twoStepDenseDFM(data = FM_no_date$data, delay = delay, delay = delay, no_of_factors = no_of_factors,
                               max_factor_lag_order  = -1, 
                               decorr_errors = decorr_errors,  lag_estim_criterion  = "HIC", 
                               comp_null = comp_null,  check_rank = check_rank, log = log, 
                               parallel = parallel, fcast_horizon = fcast_horizon))
  
  expect_error(twoStepDenseDFM(data = FM_no_date$data, delay = delay, no_of_factors = no_of_factors,
                               max_factor_lag_order  = max_factor_lag_order, 
                               decorr_errors = "flalse",  lag_estim_criterion  = "HIC", 
                               comp_null = comp_null,  check_rank = check_rank, log = log, 
                               parallel = parallel, fcast_horizon = fcast_horizon))
  
  expect_error(twoStepDenseDFM(data = FM_no_date$data, delay = delay, no_of_factors = no_of_factors,
                               max_factor_lag_order  = max_factor_lag_order, 
                               decorr_errors = decorr_errors,  lag_estim_criterion  = -1, 
                               comp_null = comp_null,  check_rank = check_rank, log = log, 
                               parallel = parallel, fcast_horizon = fcast_horizon))
  
  expect_error(twoStepDenseDFM(data = FM_no_date$data, delay = delay, no_of_factors = no_of_factors,
                               max_factor_lag_order  = max_factor_lag_order, 
                               decorr_errors = decorr_errors,  lag_estim_criterion  = "HIC", 
                               comp_null = -1,  check_rank = check_rank, log = log, 
                               parallel = parallel, fcast_horizon = fcast_horizon))
  
  expect_error(twoStepDenseDFM(data = FM_no_date$data, delay = delay, no_of_factors = no_of_factors,
                               max_factor_lag_order  = max_factor_lag_order, 
                               decorr_errors = decorr_errors,  lag_estim_criterion  = "HIC", 
                               comp_null = comp_null,  check_rank = "flasche", log = log, 
                               parallel = parallel, fcast_horizon = fcast_horizon))
  
  expect_error(twoStepDenseDFM(data = FM_no_date$data, delay = delay, no_of_factors = no_of_factors,
                               max_factor_lag_order  = max_factor_lag_order, 
                               decorr_errors = decorr_errors,  lag_estim_criterion  = "HIC", 
                               comp_null = comp_null,  check_rank = check_rank, log = "Yes", 
                               parallel = parallel, fcast_horizon = fcast_horizon))
  
  expect_error(twoStepDenseDFM(data = FM_no_date$data, delay = delay, no_of_factors = no_of_factors,
                               max_factor_lag_order  = max_factor_lag_order, 
                               decorr_errors = decorr_errors,  lag_estim_criterion  = "HIC", 
                               comp_null = comp_null,  check_rank = check_rank, log = log, 
                               parallel = "No", fcast_horizon = fcast_horizon))
  
  expect_error(twoStepDenseDFM(data = FM_no_date$data, delay = delay, delay = delay, no_of_factors = no_of_factors,
                               max_factor_lag_order  = max_factor_lag_order, 
                               decorr_errors = decorr_errors,  lag_estim_criterion  = lag_estim_criterion, 
                               comp_null = comp_null,  check_rank = check_rank, log = log, 
                               parallel = parallel, fcast_horizon = -1))
  
})