Python API — LightGBM 4.6.0.99 documentation (original) (raw)
Data Structure API
| Dataset(data[, label, reference, weight, ...]) | Dataset in LightGBM. |
|---|---|
| Booster([params, train_set, model_file, ...]) | Booster in LightGBM. |
| CVBooster([model_file]) | CVBooster in LightGBM. |
| Sequence() | Generic data access interface. |
Training API
| train(params, train_set[, num_boost_round, ...]) | Perform the training with given parameters. |
|---|---|
| cv(params, train_set[, num_boost_round, ...]) | Perform the cross-validation with given parameters. |
Scikit-learn API
| LGBMModel(*[, boosting_type, num_leaves, ...]) | Implementation of the scikit-learn API for LightGBM. |
|---|---|
| LGBMClassifier(*[, boosting_type, ...]) | LightGBM classifier. |
| LGBMRegressor(*[, boosting_type, ...]) | LightGBM regressor. |
| LGBMRanker(*[, boosting_type, num_leaves, ...]) | LightGBM ranker. |
Dask API
Added in version 3.2.0.
| DaskLGBMClassifier(*[, boosting_type, ...]) | Distributed version of lightgbm.LGBMClassifier. |
|---|---|
| DaskLGBMRegressor(*[, boosting_type, ...]) | Distributed version of lightgbm.LGBMRegressor. |
| DaskLGBMRanker(*[, boosting_type, ...]) | Distributed version of lightgbm.LGBMRanker. |
Callbacks
| early_stopping(stopping_rounds[, ...]) | Create a callback that activates early stopping. |
|---|---|
| log_evaluation([period, show_stdv]) | Create a callback that logs the evaluation results. |
| record_evaluation(eval_result) | Create a callback that records the evaluation history into eval_result. |
| reset_parameter(**kwargs) | Create a callback that resets the parameter after the first iteration. |
Plotting
| plot_importance(booster[, ax, height, xlim, ...]) | Plot model's feature importances. |
|---|---|
| plot_split_value_histogram(booster, feature) | Plot split value histogram for the specified feature of the model. |
| plot_metric(booster[, metric, ...]) | Plot one metric during training. |
| plot_tree(booster[, ax, tree_index, ...]) | Plot specified tree. |
| create_tree_digraph(booster[, tree_index, ...]) | Create a digraph representation of specified tree. |