VotingEnsemble

class VotingEnsemble(pipelines: List[etna.pipeline.base.BasePipeline], weights: Optional[Union[List[float], Literal['auto']]] = None, regressor: Optional[Union[sklearn.tree._classes.DecisionTreeRegressor, sklearn.tree._classes.ExtraTreeRegressor, sklearn.ensemble._forest.RandomForestRegressor, sklearn.ensemble._forest.ExtraTreesRegressor, sklearn.ensemble._gb.GradientBoostingRegressor, catboost.core.CatBoostRegressor]] = None, n_folds: int = 3, n_jobs: int = 1, joblib_params: Optional[Dict[str, Any]] = None)

Bases: etna.ensembles.mixins.EnsembleMixin, etna.ensembles.mixins.SaveEnsembleMixin, etna.pipeline.base.BasePipeline

VotingEnsemble is a pipeline that forecast future values with weighted averaging of it’s pipelines forecasts.

Examples

>>> from etna.datasets import generate_ar_df
>>> from etna.datasets import TSDataset
>>> from etna.ensembles import VotingEnsemble
>>> from etna.models import NaiveModel
>>> from etna.models import ProphetModel
>>> from etna.pipeline import Pipeline
>>> df = generate_ar_df(periods=30, start_time="2021-06-01", ar_coef=[1.2], n_segments=3)
>>> df_ts_format = TSDataset.to_dataset(df)
>>> ts = TSDataset(df_ts_format, "D")
>>> prophet_pipeline = Pipeline(model=ProphetModel(), transforms=[], horizon=7)
>>> naive_pipeline = Pipeline(model=NaiveModel(lag=10), transforms=[], horizon=7)
>>> ensemble = VotingEnsemble(
...     pipelines=[prophet_pipeline, naive_pipeline],
...     weights=[0.7, 0.3]
... )
>>> _ = ensemble.fit(ts=ts)
>>> forecast = ensemble.forecast()
>>> forecast
segment         segment_0        segment_1       segment_2
feature            target           target          target
timestamp
2021-07-01          -8.84          -186.67          130.99
2021-07-02          -8.96          -198.16          138.81
2021-07-03          -9.57          -212.48          148.48
2021-07-04         -10.48          -229.16          160.13
2021-07-05         -11.20          -248.93          174.39
2021-07-06         -12.47          -281.90          197.82
2021-07-07         -13.51          -307.02          215.73

Init VotingEnsemble.

Parameters

• pipelines (List[etna.pipeline.base.BasePipeline]) – List of pipelines that should be used in ensemble

• weights (Optional[Union[List[float], Literal['auto']]]) –

  List of pipelines’ weights.

  • If None, use uniform weights

  • If List[float], use this weights for the base estimators, weights will be normalized automatically

  • If “auto”, use importances of the base estimators forecasts as weights of base estimators

• regressor (Optional[Union[sklearn.tree._classes.DecisionTreeRegressor, sklearn.tree._classes.ExtraTreeRegressor, sklearn.ensemble._forest.RandomForestRegressor, sklearn.ensemble._forest.ExtraTreesRegressor, sklearn.ensemble._gb.GradientBoostingRegressor, catboost.core.CatBoostRegressor]]) – Regression model with fit/predict interface which will be used to evaluate weights of the base estimators. It should have feature_importances_ property (e.g. all tree-based regressors in sklearn)

• n_folds (int) – Number of folds to use in the backtest. Backtest is used to obtain the forecasts from the base estimators; forecasts will be used to evaluate the estimator’s weights.

• n_jobs (int) – Number of jobs to run in parallel

• joblib_params (Optional[Dict[str, Any]]) – Additional parameters for joblib.Parallel

Raises

ValueError: – If the number of the pipelines is less than 2 or pipelines have different horizons.

Inherited-members

Methods

backtest(ts, metrics[, n_folds, mode, ...])	Run backtest with the pipeline.
fit(ts)	Fit pipelines in ensemble.
forecast([ts, prediction_interval, ...])	Make a forecast of the next points of a dataset.
load(path[, ts])	Load an object.
params_to_tune()	Get hyperparameter grid to tune.
predict(ts[, start_timestamp, ...])	Make in-sample predictions on dataset in a given range.
save(path)	Save the object.
set_params(**params)	Return new object instance with modified parameters.
to_dict()	Collect all information about etna object in dict.

Attributes

fit(ts: etna.datasets.tsdataset.TSDataset) → etna.ensembles.voting_ensemble.VotingEnsemble

Fit pipelines in ensemble.

Parameters

ts (etna.datasets.tsdataset.TSDataset) – TSDataset to fit ensemble

Returns

Fitted ensemble

Return type

self

params_to_tune() → Dict[str, etna.distributions.distributions.BaseDistribution]

Get hyperparameter grid to tune.

Not implemented for this class.

Returns

Grid with hyperparameters.

Return type

Dict[str, etna.distributions.distributions.BaseDistribution]