Create configuration object for running AutoML

This function creates the configuration object required to execute get_automl_model() and run_automl_pipeline().

Usage

get_automl_config(
  include_algos = "DRF",
  max_runtime_secs = NULL,
  max_models = 1,
  max_runtime_secs_per_model = 60 * 5,
  balance_classes = FALSE
)

Arguments

include_algos

(character) vector of included algorithms (see h2o::h2o.automl())

max_runtime_secs

maximum time in seconds for H2O's AutoML to run (see h2o::h2o.automl())

max_models

maximum number of models

max_runtime_secs_per_model

maximum time in seconds for each model

balance_classes

(logical) whether to balance classes for a categorical target (see h2o::h2o.automl())

Value

Configuration object (list) to pass to get_automl_model() or run_automl_pipeline()

See also

get_automl_model(), run_automl_pipeline(), h2o::h2o.automl()

Examples

if (FALSE) {
config <- get_automl_config()
pip <- df |> run_automl_pipeline(target="y1", config)
}