Datasets

Datasets included with the R-package

agaricus.train

Training part from Mushroom Data Set

agaricus.test

Test part from Mushroom Data Set

bank

Bank Marketing Data Set

Data Input / Output

Data I/O required for LightGBM

dim(<lgb.Dataset>)

Dimensions of an lgb.Dataset

dimnames(<lgb.Dataset>) `dimnames<-`(<lgb.Dataset>)

Handling of column names of lgb.Dataset

get_field()

Get one attribute of a lgb.Dataset

set_field()

Set one attribute of a lgb.Dataset object

slice()

Slice a dataset

lgb.Dataset()

Construct lgb.Dataset object

lgb.Dataset.construct()

Construct Dataset explicitly

lgb.Dataset.create.valid()

Construct validation data

lgb.Dataset.save()

Save lgb.Dataset to a binary file

lgb.Dataset.set.categorical()

Set categorical feature of lgb.Dataset

lgb.Dataset.set.reference()

Set reference of lgb.Dataset

lgb.convert_with_rules()

Data preparator for LightGBM datasets with rules (integer)

Machine Learning

Train models with LightGBM and then use them to make predictions on new data

lightgbm()

Train a LightGBM model

lgb.train()

Main training logic for LightGBM

predict(<lgb.Booster>)

Predict method for LightGBM model

lgb.cv()

Main CV logic for LightGBM

Saving / Loading Models

Save and load LightGBM models

lgb.dump()

Dump LightGBM model to json

lgb.save()

Save LightGBM model

lgb.load()

Load LightGBM model

lgb.model.dt.tree()

Parse a LightGBM model json dump

saveRDS.lgb.Booster()

saveRDS for lgb.Booster models

readRDS.lgb.Booster()

readRDS for lgb.Booster models

Model Interpretation

Analyze your models

lgb.get.eval.result()

Get record evaluation result from booster

lgb.importance()

Compute feature importance in a model

lgb.interprete()

Compute feature contribution of prediction

lgb.plot.importance()

Plot feature importance as a bar graph

lgb.plot.interpretation()

Plot feature contribution as a bar graph

Miscellaneous

Ungroupable functions to troubleshoot LightGBM

lgb.unloader()

Remove lightgbm and its objects from an environment