Skip to content

UnivariateModel

UnivariateModel(
   order: int, family: str, bias: bool
)

Base class for univariate models.

Args

  • order (int) : Order of the model.
  • family (str) : Family of the model. Defaults to "AR".

Attributes

  • order : Order of the model.
  • family : Family of the model.

Methods:

.forecast

.forecast(
   initial_conditions: Union[pd.DataFrame, np.ndarray, list] = None, horizon: int = 1
)

Simulate the model forward in time.

Args

  • initial_conditions (ArrayLike) : Initial conditions for the model.
  • horizon (int) : Number of steps to forecast. Defaults to 1.

Returns

  • forecast (ndarray) : Forecasted values.

AutoRegressive

AutoRegressive(
   p: int = 1, bias: bool = True
)

Class for purely autoregressive models of order p.

Args

  • p (int) : Order of the model. Defaults to 1.
  • bias (bool) : Whether to include a bias term in the model. Defaults to True.

Attributes

  • p : Order of the model.
  • coef : Coefficients of the model.

Methods:

.fit

.fit(
   data: Union[pd.DataFrame, np.ndarray]
)

Fit the model to the data.

MovingAverage

MovingAverage(
   q: int = 1, bias: bool = True
)

Class for purely moving average models of order q.

Args

  • q (int) : Order of the model. Defaults to 1.
  • bias (bool) : Whether to include a bias term in the model. Defaults to True.

Attributes

  • q : Order of the model.
  • coef : Coefficients of the model.

Methods:

.fit

.fit(
   data: Union[pd.DataFrame, np.ndarray], n_iterations: int = 100
)

Fit the model to the data.