R/mod_model_xgb_utils.R

In predictoR: Predictive Data Analysis System

#' Var importance XGBoosting
#'
#' @param modelo a random forest model.
#' @param error a character specifying the type of importance.
#'
#' @author Joseline Quiros <joseline.quiros@promidat.com>
#' @return echarts4r plot
#' @export e_xgb_importance
#' @import echarts4r
#' @import traineR
#' @examples
#' model <- traineR::train.xgboost(Species ~ ., data = iris, nrounds = 20)
#' e_xgb_importance(model)
#' 
e_xgb_importance <- function(modelo, error = "Gain") {
  x   <- modelo$feature_names
  aux <- xgboost::xgb.importance(feature_names = x, model = modelo)
  
  aux <- data.frame(aux)
  aux <- na.omit(aux)
  if(is.null(aux[[error]])) {
    error <- "Weight"
    aux <- aux |>
      dplyr::group_by(Feature) |>
      dplyr::summarise(Weight = sum(Weight), .groups = 'drop') |>
      data.frame()
  }
  aux[[error]] <- abs(aux[[error]])
  aux <- aux[order(aux[[error]], decreasing = T), ]
  color <- gg_color_hue(nrow(aux))
  aux[["color"]] <- color
  
  res <- aux |> e_charts(Feature) |> e_bar_(error) |>  
    e_tooltip() |>  e_datazoom(show = F) |>  e_show_loading() |>
    e_add_nested("itemStyle", color) |> e_flip_coords() |> 
    e_y_axis(inverse = TRUE)
  
  return(res)
}