formatting

R/ensemble_models.R

@@ -236,7 +236,7 @@ ensemble_models <- function(run_info,
         avail_arg_list <- list(
           "train_data" = prep_ensemble_tbl %>% dplyr::select(-Train_Test_ID),
           "model_type" = "ensemble",
-          "pca" = FALSE, 
+          "pca" = FALSE,
           "multistep" = FALSE
         )
 

R/feature_selection.R

@@ -31,7 +31,7 @@ run_feature_selection <- function(input_data,
 
     return(fs_list)
   }
-  
+
   # check for multiple time series
   if (length(unique(input_data$Combo)) > 1) {
     global <- TRUE
@@ -103,7 +103,7 @@ run_feature_selection <- function(input_data,
         ) %>%
         dplyr::select(Feature, Vote, Auto_Accept)
 
-      # don't run leave one feature out process for daily, weekly, or global model data
+      # don't run boruta for daily, weekly, or global model data
       boruta_results <- tibble::tibble()
     } else {
       if (!fast) { # full implementation

R/models.R

@@ -69,7 +69,7 @@ list_global_models <- function() {
 #' @noRd
 list_multivariate_models <- function() {
   list <- c(
-    "cubist", "glmnet", "mars", "svm-poly", "svm-rbf", "xgboost", 
+    "cubist", "glmnet", "mars", "svm-poly", "svm-rbf", "xgboost",
     "arima-boost", "arimax", "prophet-boost", "prophet-xregs",
     "nnetar-xregs"
   )

R/train_models.R

@@ -419,7 +419,7 @@ train_models <- function(run_info,
           dplyr::select(Model_Workflow)
 
         workflow <- workflow$Model_Workflow[[1]]
-        
+
         if (nrow(prep_data) > 500 & model == "xgboost") {
           # update xgboost model to use 'hist' tree method to speed up training
           workflow <- workflows::update_model(workflow,
@@ -429,8 +429,10 @@ train_models <- function(run_info,
 
         if (combo_hash == "All-Data") {
           # adjust column types to match original data
-          prep_data <- adjust_column_types(prep_data, 
-                                           workflows::extract_recipe(workflow, estimated = FALSE))
+          prep_data <- adjust_column_types(
+            prep_data,
+            workflows::extract_recipe(workflow, estimated = FALSE)
+          )
         }
 
         if (feature_selection & model %in% fs_model_list) {
@@ -1013,16 +1015,16 @@ adjust_column_types <- function(data, recipe) {
   expected_types <- recipe$var_info %>%
     dplyr::select(variable, type) %>%
     dplyr::mutate(type = purrr::map_chr(type, ~ .x[[1]]))
-  
+
   # Identify and coerce mismatched columns
   for (i in seq_len(nrow(expected_types))) {
     col_name <- expected_types$variable[i]
     expected_type <- expected_types$type[i]
-    
+
     # Check if column exists and type mismatch
     if (col_name %in% names(data)) {
       actual_type <- class(data[[col_name]])[1]
-      
+
       # Convert if types are different
       if (expected_type == "string" && actual_type != "character") {
         data[[col_name]] <- as.character(data[[col_name]])
@@ -1035,5 +1037,3 @@ adjust_column_types <- function(data, recipe) {
   }
   return(data)
 }
-
-