run_ml does not support "training_frac = 1.0" to get best hyperparameter value

[joannacolovas]

run_ml does not support arguement "training_fract = 1.0", and gives following error message:
Error in check_training_frac(training_frac) :
training_frac must be a numeric between 0 and 1.
You provided: 1
training_frac = 1.0 using a random forest model to find best mtry hyperparameter value.

Function "check_training_frac" is exclusive of (0,1). (mikropml/R/checks.R line 131-140:

check_training_frac <- function(frac) {
  if (!is.numeric(frac) | (frac <= 0 | frac >= 1)) {
    stop(paste0(
      "`training_frac` must be a numeric between 0 and 1.\n",
      "    You provided: ", frac
    ))
  } else if (frac < 0.5) {
    warning("`training_frac` is less than 0.5. The training set will be smaller than the testing set.")
  }
}

Reproducible example

  # random forest model trying to set training_fract to 1.0 to have no test set to get best mtry value
  run_ml(otu_small, 
         "rf",
         outcome_colname = "dx",
         training_frac = 1.0,
         seed = 5, 
         kfold = 5, 
         cv_times = 100, 
         hyperparameters = list(mtry =  c(84, 100, 150, 200, 300)), 
         calculate_performance = FALSE)