Prophetverse

Univariate prophetverse forecaster - prophet model implemented in numpyro.

Quickstart

from sktime.forecasting.prophetverse import Prophetverse

estimator = Prophetverse(trend='linear', exogenous_effects=None, default_effect=None, feature_transformer=None, noise_scale=None, likelihood='normal', scale=None, rng_key=None, inference_engine=None, broadcast_mode='estimator')

Parameters(9)

trendUnion[str, BaseEffect], optional

Type of trend to use. Either “linear” (default) or “logistic”, or a custom effect object.

exogenous_effectsOptional[List[BaseEffect]], optional

List of effect objects defining the exogenous effects.

default_effectOptional[BaseEffect], optional

The default effect for variables without a specified effect.

feature_transformersktime transformer, optional

Transformer object to generate additional features (e.g., Fourier terms).

noise_scalefloat, optional

Scale parameter for the observation noise. Must be greater than 0. (default: 0.05)

likelihoodstr, optional

The likelihood model to use. One of “normal”, “gamma”, or “negbinomial”. (default: “normal”)

scaleoptional

Scaling value inferred from the data.

rng_keyoptional

A jax.random.PRNGKey instance, or None.

inference_engineoptional

An inference engine for running the model.

Examples

>>> from sktime.datasets import load_airline
>>> from sktime.forecasting.prophetverse import Prophetverse
>>> from prophetverse.effects.fourier import LinearFourierSeasonality
>>> from prophetverse.utils.regex import no_input_columns
>>> y = load_airline ()
>>> model = Prophetverse (
... exogenous_effects = [
... (
... "seasonality",
... LinearFourierSeasonality (
... sp_list = [12 ],
... fourier_terms_list = [3 ],
... freq = "M",
... effect_mode = "multiplicative",
... ),
... no_input_columns,
... )
... ],
... )
>>> model. fit (y)
>>> model. predict (fh = [1, 2, 3 ])