BaseForecaster

base_forecaster.py · inherits ABC

BaseForecaster defines the shared interface and common infrastructure for all models in Chronax.

This class implements the core forecaster contract.

Instance Attributes: * alias: model name, declared in model's __init__. * conformal_params: a conformal_intervals object, used for prediction intervals. * model_: stores fitted model post-training.

Class Attributes: * uses_exog: bool (Default: False). Boolean representing model's exogenous variable handling.

new(self) -> Self

Returns a shallow copy of the object, used internally to clone a model without mutating state.

__repr__(self) -> str

Returns the model's alias for easy identification.

fit(self, y, X=None) -> Self

Fit the model to univariate time series y. Must set self.model_ and return self.

Parameter	Type	Default	Description
y	jnp.ndarray	-	Univariate time series.
X	jnp.ndarray | None	None	Optional exogenous input.

Returns: Self (the fitted forecaster; sets self.model_).

predict(self, h, X=None, level=None) -> dict

Generate h-step-ahead forecasts.

Parameter	Type	Default	Description
h	int	-	(undocumented)
X	jnp.ndarray | None	None	(undocumented)
level	list[int | float] | None	None	(undocumented)

Returns: dict (Returns a dict with at least {"mean": jnp.ndarray}).

forecast(self, y, h, X=None, X_future=None, level=None, fitted=False) -> dict

Stateless fit+predict on y, forecasting h steps ahead. Must return a dict with at least {"mean": jnp.ndarray}. Subclasses may extend the signature with model-specific optional parameters.

Parameter	Type	Default	Description
y	jnp.ndarray	-	(undocumented)
h	int	-	(undocumented)
X	jnp.ndarray | None	None	(undocumented)
X_future	jnp.ndarray | None	None	(undocumented)
level	list[int | float] | None	None	(undocumented)
fitted	bool	False	(undocumented)

Returns: dict (Must return a dict with at least {"mean": jnp.ndarray}).

forward(self, y, h, X=None, X_future=None, level=None, fitted=False) -> dict

Update the model on new data y and forecast h steps ahead. Default delegates to forecast(). Subclasses with warm-start behavior (e.g. Holt, HoltWinters, ETS) override this.

Parameter	Type	Default	Description
y	jnp.ndarray	-	(undocumented)
h	int	-	(undocumented)
X	jnp.ndarray | None	None	(undocumented)
X_future	jnp.ndarray | None	None	(undocumented)
level	list[int | float] | None	None	(undocumented)
fitted	bool	False	(undocumented)

Returns: dict

conformity_scores(self, y, X=None) -> jnp.ndarray

Computes the model's conformity score on y as a 2D JAX array. A model's conformity score is the absolute difference between forecasted and actual values across h positions and n_windows. Uses vmap for parallelization over windows.

Parameter	Type	Default	Description
y	jnp.ndarray	-	(undocumented)
X	jnp.ndarray | None	None	(undocumented)

Returns: jnp.ndarray Raises: ValueError

add_confidence_intervals(fcst, cs, level, method) -> dict

Calculates confidence intervals at level(s) for forceasted values based on conformity_score.

Parameter	Type	Default	Description
fcst	dict	-	(undocumented)
cs	jnp.ndarray	-	(undocumented)
level	list[int | float]	-	(undocumented)
method	str	-	(undocumented)

Returns: dict Raises: ValueError