Releases
2.4.0
Highlights
Add module etna.experimental.prediction_intervals
with new model-agnostic methods to generate prediction intervals
Add notebook about prediction intervals
Add DeepARNativeModel
Optimize TSDataset.to_hierarchical_dataset
, BasePipeline._validate_backtest_dataset
, datasets.utils.duplicate_data
Full changelog
Added
Add params_to_tune
for DeepStateModel
(#115 )
Handle new functionality for prediction intervals in the plot_forecast
(#130 )
Add get_historical_forecasts
to pipelines for forecast estimation at each fold on the historical dataset (#143 )
ConformalPredictionIntervals
method for prediction intervals estimation (#152 )
Add DeepARNativeModel
(#114 )
EmpiricalPredictionIntervals
method for prediction intervals estimation (#173 )
Prediction intervals tutorial notebook (#189 )
Changed
Change warning condition on loading object saved under different library version (#31 )
Fixed
Speed up segment column creation in TSDataset.to_hierarchical_dataset
(#194 )
Speed up BasePipeline._validate_backtest_dataset
(#194 )
Speed up datasets.utils.duplicate_data
(#194 )
You can’t perform that action at this time.