xgb_bin: Automatically models categorical targets with random search...
In prescient/modelpipe: Fast and opinionated model development pipelines.

Automatically models categorical targets with random search hyperparameter optimziation. XGBoost model trains with lossguide and histogram tree method to accelerate tuning.

xgb_bin(
  df_train,
  df_test,
  tune_rounds = 25L,
  verbose = T,
  max_rounds = 10^6,
  cv_folds = 5L,
  early_stopping_rounds = 10L,
  folds = NULL
)

`df_train`	Training data.frame with column called "target" for training. All columns should be numeric and prepared with a package like vtreat.
`df_test`	Testing data.frame with column called "target" for evaluation. All columns should be numeric and prepared with a package like vtreat.
`tune_rounds`	Integer (e.g. 25L) indicating the number of hyperoptimization tuning rounds.
`verbose`	Print model iterations (T/F).
`max_rounds`	Maximum number of rounds to use in model fitting.
`cv_folds`	Integer (e.g. 5L) that sets the number of cross validation folds to use in model tuning.
`early_stopping_rounds`	Integer (e.g. 10L) that sets xgb.cv early_stopping_rounds parameter.
`folds`	Allows users to specify their own folds (e.g. stratified folds).