prediction_interval_outliers

Functions

_select_segments_subset(ts, segments)	Create TSDataset with certain segments.
create_ts_by_column(ts, column)	Create TSDataset based on original ts with selecting only column in each segment and setting it to target.
get_anomalies_prediction_interval(ts, model)	Get point outliers in time series using prediction intervals (estimation model-based method).

create_ts_by_column(ts: etna.datasets.tsdataset.TSDataset, column: str) → etna.datasets.tsdataset.TSDataset

Create TSDataset based on original ts with selecting only column in each segment and setting it to target.

Parameters

• ts (etna.datasets.tsdataset.TSDataset) – dataset with timeseries data

• column (str) – column to select in each.

Returns

result – dataset with selected column.

Return type

TSDataset

get_anomalies_prediction_interval(ts: etna.datasets.tsdataset.TSDataset, model: Union[Type[ProphetModel], Type[SARIMAXModel]], interval_width: float = 0.95, in_column: str = 'target', **model_params) → Dict[str, List[pandas._libs.tslibs.timestamps.Timestamp]]

Get point outliers in time series using prediction intervals (estimation model-based method).

Outliers are all points out of the prediction interval predicted with the model.

Parameters

• ts (etna.datasets.tsdataset.TSDataset) – dataset with timeseries data(should contains all the necessary features).

• model (Union[Type[ProphetModel], Type[SARIMAXModel]]) – model for prediction interval estimation.

• interval_width (float) – the significance level for the prediction interval. By default a 95% prediction interval is taken.

• in_column (str) –

  column to analyze

  • If it is set to “target”, then all data will be used for prediction.

  • Otherwise, only column data will be used.

Returns

dict of outliers in format {segment: [outliers_timestamps]}.

Return type

Dict[str, List[pandas._libs.tslibs.timestamps.Timestamp]]

Notes

For not “target” column only column data will be used for learning.