Skip to contents

Train a predictive model using (rROC-preprocessed) data

train_rROC_h2o.Rd

A convenience function to train a model using the data x_prepared to predict the outcome y. Per default, the function is a wrapper around the h2o.randomForest function.

Usage

train_rROC_h2o(
  x_prepared,
  y,
  init_h2o = TRUE,
  h2o_trainfun = function(df, col_y, cols_x, ...) {
    
    h2o::h2o.randomForest(training_frame = df, y = col_y, x = cols_x, ntrees = 1000,
    max_depth = 20, min_rows = 1, nbins = 20, seed = 4242, ...)
 },
  ...
)

Arguments

x_prepared

A data frame containing the preprocessed data to use in the model. The data frame should not contain the outcome variable. Usually the result of a call to prepare_modeldata.

y

The outcome variable. Usually a factor.

init_h2o

Whether to initialize h2o. Defaults to TRUE.

h2o_trainfun

The function to use to train the model. Defaults to h2o.randomForest with some default parameters. Needs to have df, col_y and cols_x as arguments where df is the data frame to use, col_y is the column index of the outcome variable and cols_x is a vector of column indices of the features to use to predict the outcome.

...

Arguments passed to h2o_trainfun.