Gives the number of active grid cells before row element j, in column i. Parameters ---------- elem (i,j) index,

Gives the number of activate grid cells in each column. Returns ------- np.ndarray of shape (n+1,) The number of

process the torch_metrics parameter.

Convert lags to the dictionary format, assuming the lags are shared across the components

Helper function that constructs the labels array `y` from the target series. This is done by by looping over each time in `feature_ti

Convert a dataframe column to a pandas Index or DatetimeIndex. Parameters ---------- df The dataframe containing the time column.

Applies Dynamic Time Warping to `actual_series` and `pred_series` before passing it into the metric. Enables comparison between series of dif

add time dimension to static context

Extract the full docstring from a Python file. Returns: Tuple of (title, body) where body is the docstring content without the title sect

Computes a moving average of this series' values and returns a new TimeSeries. The returned series has the same length and time axis

Fits the scorer on the given time series. If a sequence of series, the scorer is fitted on the concatenation of the sequence. The as

Determine empirically good embedding sizes (formula taken from fastai). Args: n (int): number of classes max_size (int, optio

Helper function that returns the times shared by all of the specified series that can be used to create features and labels. This is

Helper function that returns the times shared by all specified series that can be used to create features and labels. This is perform

Helper function called by `_get_feature_times` that extracts all of the times within `future_covariates` that can be used to create f

Get the title and docstring body for a module or package. Args: module_name: Full module name like "darts.models.forecasting" sou

Get the value of an option.

Helper function to create the X and y arrays by building them block by block (one per covariates).

Evaluate model on multiple series, for all 4 supported metric functions

simple helper that returns prediction for the individual test cases

generates multivariate test case data. Target series is a sine wave stacked with a repeating linear curve of equal seasonal length. Covariates

Helper function to test that the model raises an error when calling `predict()` or `historical_forecasts()` after `fit_from_dataset()` with mi

Test `SingleEncoder.encode_train()`

Test which includes adding custom seasonalities and future covariates. The tests compare the output of univariate and stochastic forecasting w

Generates historical forecasts by simulating predictions at various points in time throughout the history of the provided (potentially multipl

Called by dtw to initialize the window to a certain size. Parameters ---------- n The width of the windo

Return a deterministic series with the kurtosis of each component computed over the samples of the stochastic series. This works only

A wrapper around PyTorch Lightning's `Tuner.lr_find()`. Performs a range test of good initial learning rates, to reduce the amount of

Gives the mean distance between pair-wise elements in the two series after warping. Returns ------- float The mea

Computes the ratio of missing values Parameters ---------- series The time series to compute ratio on Returns -----

Plots the Autocorrelation Function (ACF) of `ts`, highlighting it at lag `m`, with corresponding significance interval. Uses :func:`statsmode

Plots the Tolerance Curve for evaluating forecast alignment. The tolerance curve shows, for each tolerance level (as a percentage of the act

Predict the ``n`` time step following the end of the training series, or of the specified ``series``. Prediction is performed with a PyTorch

Returns the distribution parameters as a single Tensor, extracted from the raw model outputs.

Helper function to process and check either of the background or foreground series input to `_ForecastingModelExplainer`. If no input was pro

Raises a DeprecationWarning. Parameters ---------- message The message of the ValueError. logger The logger inst

Reset the option to its default value.

Rotates half the hidden dims of the input.

Compute anomaly score(s) for the given series. Predicts the given target time series with the forecasting model, and applies the scorer(s)

to be set from TorchForecastingModel before calling trainer.predict() and reset at self.on_predict_end()

Plot the results of the scorer. Computes the score on the given series input. And plots the results. The plot will be composed of th

Return a deterministic series with the skew of each component computed over the samples of the stochastic series. This works only on

Returns a new Index with the same type as the input `index`, containing the values between `start` and `end` included. If start and end are n

symmetric Mean Absolute Percentage Error (sMAPE). For the true series :math:`y` and predicted series :math:`\\hat{y}` of length :math:`T`, it is

Splits all provided TimeSeries instances into train and test sets according to the provided timestamp. Parameters ---------- feature

Transforms a (sequence of) of series by calling the user-implemeneted `ts_transform` method. In case a ``Sequence[TimeSeries]`` is passed as

Warps the two time series according to the warp path returned by `DTWAlignment.path()`, which minimizes the pair-wise distance. This w

Data transformer to apply a custom function to a (sequence of) ``TimeSeries`` (similar to calling :func:`TimeSeries.map()` on each se

Darts' Progress Bar for `TorchForecastingModels`. Allows to customize for which model stages (sanity checks, training, validation, prediction

Abstract class for all inference datasets that can be used with Darts' `TorchForecastingModel`. Provides samples to compute forecast

Applies the (invertible) transform `transformed_series = scale * series + translation`. When 'fitting' this transform, the `s

Applies the transform `transformed_series = scale * series + translation`. When 'fitting' this transform, the `scale` and `tr

Parameters ---------- ffn One of Darts' Position-wise Feed-Forward Network variants from darts.models.components.

XGBoost Model Parameters ---------- lags Lagged target `series` values used to predict the next time step/s.

CatBoost Model Parameters ---------- lags Lagged target `series` values used to predict the next time step/s.

Random Forest Model Note: `RandomForest` is deprecated and will be removed in a future version. Use :class:`~darts.models.forecasting

Temporal Fusion Transformers (TFT) for Interpretable Time Series Forecasting. This is an implementation of the TFT architecture, as outlined

TiRex foundation model for zero-shot time series forecasting. This is a Darts wrapper around the TiRex model introduced in Auer et a

An implementation of the Theta method with configurable `theta` parameter. See [1]_. The training time series is de-seasonalized acc

TimesFM 2.5 Model for zero-shot forecasting. This is an implementation of Google's TimesFM 2.5 model, ported from `google-research/ti