R/reg_rf.R

In daltoolbox: Leveraging Experiment Lines to Data Analytics

#'@title Random Forest for regression
#'@description Creates a regression object that
#' uses the Random Forest method.
#' It wraps the randomForest library.
#'@param attribute attribute target to model building
#'@param nodesize node size
#'@param ntree number of trees
#'@param mtry number of attributes to build tree
#'@return obj
#'@examples
#'data(Boston)
#'model <- reg_rf("medv", ntree=10)
#'
#'# preparing dataset for random sampling
#'sr <- sample_random()
#'sr <- train_test(sr, Boston)
#'train <- sr$train
#'test <- sr$test
#'
#'model <- fit(model, train)
#'
#'test_prediction <- predict(model, test)
#'test_predictand <- test[,"medv"]
#'test_eval <- evaluate(model, test_predictand, test_prediction)
#'test_eval$metrics
#'@export
reg_rf <- function(attribute, nodesize = 1, ntree = 10, mtry = NULL) {
  obj <- regression(attribute)

  obj$nodesize <- nodesize
  obj$ntree <- ntree
  obj$mtry <- mtry

  class(obj) <- append("reg_rf", class(obj))
  return(obj)
}

#'@importFrom randomForest randomForest
#'@export
fit.reg_rf <- function(obj, data, ...) {
  data <- adjust_data.frame(data)
  obj <- fit.predictor(obj, data)

  if (is.null(obj$mtry))
    obj$mtry <- ceiling(ncol(data)/3)

  x <- data[,obj$x]
  y <- data[,obj$attribute]

  obj$model <- randomForest::randomForest(x = x, y = y, nodesize = obj$nodesize, mtry=obj$mtry, ntree=obj$ntree)

  return(obj)
}

#'@export
predict.reg_rf  <- function(object, x, ...) {
  x <- adjust_data.frame(x)
  x <- x[,object$x]
  prediction <- predict(object$model, x)
  return(prediction)
}