lkauto.ensemble package

Submodules

lkauto.ensemble.ensemble_builder module

lkauto.ensemble.ensemble_builder.build_ensemble(train: pandas.core.frame.DataFrame, top_n_runs: pandas.core.frame.DataFrame, filer: lkauto.utils.filer.Filer, ensemble_size: int, lenskit_metric, maximize_metric: bool)

lkauto.ensemble.greedy_ensemble_selection module

class lkauto.ensemble.greedy_ensemble_selection.EnsembleSelection(ensemble_size: int, lenskit_metric, maximize_metric: bool = False)

Bases: object

An ensemble of selected algorithms

Fitting an EnsembleSelection generates an ensemble from the the models generated during the search process. Can be further used for prediction.

Parameters

• lenskit_metric (metric from lenskit) – The metric used to evaluate the models

• maximize_metric (bool = False) – If the metric is to be optimized or not.

fit(data: pandas.core.frame.DataFrame)

Fit base models (we assume the ensemble part, ensemble_fit, was already fitted here or is fitted later)

Parameters

data (DataFrame) – Dataframe with columns “user”, “item”, “rating”

predict(X_data: pandas.core.frame.DataFrame)

“user”, “item” Dataframe

ensemble_fit()

ensemble_predict()

Module contents