Package index

Functions

Parsnip bindings

This package includes parsnip bindings and functions which add lightgbm to the parsnip boost_tree() model. These bindings load when the package loads.

add_boost_tree_lightgbm()

Wrapper to add lightgbm engine to the parsnip boost_tree model specification. Gets called when the package loads

train_lightgbm()

Boosted trees via LightGBM

pred_lgb_reg_num()

predict_lightgbm_regression_numeric

multi_predict(<_lgb.Booster>)

Model predictions across many sub-models

LightGBM hyperparameters

LightGBM dials:: paramter functions that can be used with tune_ functions. Pass them to set_engine to use.

max_bin() max_depth() add_to_linked_depth() learning_rate() feature_fraction() feature_fraction_bynode() min_data_in_leaf() min_gain_to_split() num_leaves() bagging_fraction() bagging_freq() min_data_per_group() max_cat_threshold() lambda_l1() lambda_l2() cat_smooth() cat_l2()

LightGBM engine-specific hyperparameters

Tidymodels helpers

Utility functions for common Tidymodels tasks, such as predicting on new data, cleaning outputs, and saving/loading fit objects.

axe_recipe()

Strip environment data from recipe objects to shrink their size

axe_tune_data()

Strip attribute data from CV objects to shrink their size

lgbm_load()

Load a LightGBM model from disk

lgbm_save()

Save a LightGBM model to disk