R/BinModelVariables.R

In citronmeliss/predictionpackr: Predictionpackr

#' Bin model variables into scores
#'
#' This function bins model variables according to cutpoints.
#' @param data The study data frame. No default.
#' @param model_variables The variables to be included in the model. As character vector. No default.
#' @param cut_points The cut points used to bin the model variables. As list of numeric vectors. No default.
#' @param levels The scores to be assigned to each bin. As list of character vectors. No default.
#' @export

BinModelVariables <- function(data, model_variables,
                              cut_points, levels)
{

  ## Subset variables for model
  model_df <- data[, model_variables]

  ## Bin variables according to cut points
  binned_variables <- sapply(model_variables,
                             function (var){
                               cut_var <- cut(model_df[,var],
                                              cut_points[[var]],
                                              labels = levels[[var]],
                                              include.lowest = TRUE)
                               num_var <- as.numeric(as.character(cut_var))
                               return (num_var)
                             }
  )
  return (binned_variables)
}