Run random forest, usually returning only diagnostic values. — make_rf_diagnostics_old • envModel

Random forest is run via caret::train.

Usage

make_rf_diagnostics_old(
  df,
  clust_col = "cluster",
  env_cols,
  trees = 999,
  folds = 3,
  reps = 5,
  tune_length = NULL,
  save_rf = NULL
)

Arguments

df: Dataframe with clusters, context and environmental columns.
clust_col: Character. Name of column with cluster membership.
env_cols: Character. Name of columns with environmental data.
trees: Numeric. How many trees in the forest?
folds: Numeric. How many folds to use in repeated cross-validation.
reps: Numeric. How many reps to use in repeated cross-validation.
tune_length: Numeric. Passed to caret::train argument tuneLength.
save_rf: Character. File name for saving the train object as (.rds).

Value

Tibble (row) of diagnostics