Override this to customize the tqdm bar for predicting.

Override this to customize the tqdm bar for the validation sanity run.

Override this to customize the tqdm bar for training.

Override this to customize the tqdm bar for validation.

This decorator adds support for quantile interval metrics with sanity checks, processing, and extraction of quantiles from the intervals.

Returns whether the pipeline is invertible or not. A pipeline is invertible if all transformers in the pipeline are themselves invert

Applies some input / output processing to PLForecastingModule.forward. Note that this wrapper must be added to each of PLForecastingModule's subc

Whether the series is deterministic.

Whether the series is stochastic (probabilistic).

Whether the scorer expects a probabilistic prediction as the first input.

Whether the series is stochastic (probabilistic).

Whether the scorer is trainable.

Whether the Scorer is trainable.

Whether the series is univariate.

Whether the Scorer is a univariate scorer.

Whether the given timestamp or integer is within the time interval of the series. `ts` does not need to be an element of the series' time ind

Interval Width (IW). IL gives the width / length of predicted quantile intervals. For the true series :math:`y` and predicted stochastic or

Interval Winkler Score (IWS) [1]_. IWS gives the length / width of the quantile intervals plus a penalty if the observation is outside the interv

The lagged feature names the model has been trained on. The naming convention for target, past and future covariates is: ``"{name}_{type}_lag

The lagged label name for the model's estimators. The naming convention is: ``"{name}_target_hrz{i}"``, where: - ``{name}`` the

Last value of the univariate series. Returns ------- float The last value of this univariate deterministic time s

Returns the likelihood (if any) that the model uses for probabilistic forecasts.

Loads a model from a given path or file handle. Parameters ---------- path Path or file handle from whic

Loads a model from a given path or file handle. Parameters ---------- path Path or file handle from whic

Mean Absolute Ranged Relative Error (MARRE). For the true series :math:`y` and predicted series :math:`\\hat{y}` of length :math:`T`, it is compu

Mean Absolute Scaled Error (MASE) (see [1]_ for more information on scaled forecasting errors). It is the Mean Absolute Error (MAE) scaled by the

Mean Continuous Ranked Probability Score (MCRPS). MCRPS is a proper scoring rule that generalises the Mean Absolute Error (MAE) to probabilistic

Mean Error (MERR). For the true series :math:`y` and predicted series :math:`\\hat{y}` of length :math:`T`, it is computed per component/colu

The metadata of this series. If defined, the metadata is given as a dictionary.

Reference implementation for AUTC metric.

The minimum number of samples for training the model.

The minimum required length for the training series.

Mean Interval Non-Conformity Score for Quantile Regression (MINCS_QR). MINCS_QR gives the time-aggregated INCS_QR :func:`~darts.metrics.metrics.i

Mean Interval Width (MIW). MIW gives the time-aggregated width / length of predicted quantile intervals. For the true series :math:`y` and p

Mean Interval Winkler Score (IWS) [1]_. MIWS gives the time-aggregated length / width of the quantile intervals plus a penalty if the observation

Patches plt.show() and closes all plots / figures from memory at the end of the test.

Mean Quantile Loss (MQL). Also known as Pinball Loss. QL is a metric that quantifies the accuracy of a specific quantile :math:`q` from the p

Mean Squared Error (MSE). For the true series :math:`y` and predicted series :math:`\\hat{y}` of length :math:`T`, it is computed per compone

Mean Squared Scaled Error (MSSE) (see [1]_ for more information on scaled forecasting errors). It is the Mean Squared Error (MSE) scaled by the M

This decorator further adapts the metrics that took as input two (or three for scaled metrics with `insample`) univariate/multivariate ``Time

This decorator transforms a metric function that takes as input two univariate TimeSeries instances into a function that takes two equally-si

The number of components (columns) in the series.

The number of samples contained in the series.

The number of time steps in the series.

Name of the data transformer.

Returns the number of distribution parameters for a single target value.

List of numeric static variables in model