ExponentialSmoothingHyperparams

chronax.models.ExponentialSmoothingHyperparams

A Pydantic model defining the configuration parameters for the Exponential Smoothing model.

__init__(self, **data)

Initializes the hyperparameters, converting string "null" values to None.

Parameter	Type	Default	Description
trend	Optional[Literal["add", "mul", "null"]]	-	Trend component: 'add', 'mul', or None (no trend)
damped_trend	Optional[bool]	False	Whether to use damped trend
seasonal	Optional[Literal["add", "mul", "null"]]	None	Seasonal component: 'add', 'mul', or None (no seasonality)
seasonal_periods	Optional[Union[int, Literal["null"]]]	None	Number of seasonal periods (None if no seasonality)

ExponentialSmoothingModel

chronax.models.ExponentialSmoothingModel · inherits BaseModel

Exponential Smoothing model implementation.

__init__(self, params: Dict[str, Any], settings: Dict[str, Any])

Initializes the Exponential Smoothing Model wrapper.

Parameter	Type	Default	Description
params	Dict[str, Any]	-	(undocumented)
settings	Dict[str, Any]	-	(undocumented)

train(self, y_context: np.ndarray, y_target: np.ndarray, timestamps_context: np.ndarray, timestamps_target: np.ndarray, x_context: Optional[np.ndarray] = None, x_target: Optional[np.ndarray] = None, **kwargs: dict) -> "ExponentialSmoothingModel"

Trains a separate Exponential Smoothing model for each variate in a (potentially multivariate) time series.

Parameters:

Parameter	Type	Default	Description
y_context	np.ndarray	-	Historical target values for model fitting. Expects 2D arrays: (num_steps, num_targets).
y_target	np.ndarray	-	Future target values (unused; included for compatibility). Expects 2D arrays: (num_steps, num_targets).
timestamps_context	np.ndarray	-	Timestamps corresponding to y_context.
timestamps_target	np.ndarray	-	Timestamps corresponding to y_target.
x_context	Optional[np.ndarray]	None	Optional covariate data for context.
x_target	Optional[np.ndarray]	None	Optional covariate data for prediction horizon.
**kwargs	dict	-	Additional keyword arguments.

Returns: ExponentialSmoothingModel (the fitted model instance).

predict(self, y_context: np.ndarray, timestamps_context: np.ndarray, timestamps_target: np.ndarray, x_context: Optional[np.ndarray] = None, x_target: Optional[np.ndarray] = None, **kwargs: dict) -> np.ndarray

Predicts future values for each variate and concatenates the results.

Parameters:

| Parameter | Type | Default | Description | |-----------|---------------|-------------| | y_context | np.ndarray | - | Context values, shape (num_steps, num_variates). | | timestamps_context | np.ndarray | - | Timestamps for context data. | | timestamps_target | np.ndarray | - | Timestamps for target/future data. | | x_context | Optional[np.ndarray] | None | Optional covariate data for context. | | x_target | Optional[np.ndarray] | None | Optional covariate data for prediction horizon. | | **kwargs | dict | - | Additional keyword arguments. |

Returns: np.ndarray (Predictions with shape (forecast_horizon, num_variates)). Raises: ValueError (If the model has not been fitted.)