initModel: Initialize a model from the Rlgt family

In Rlgt: Bayesian Exponential Smoothing Models with Trend Modifications

Initialize a model from the Rlgt family

Description

This is an internal function that usually won't be called by users directly. It validates the model type and generates the corresponding list of parameters for the model.

Usage

initModel(
  model.type = NULL,
  use.regression = FALSE,
  seasonalityMethodId = 0,
  levelMethodId = 0,
  useSmoothingMethodForError = FALSE
)

Arguments

model.type	type of the forecasting model selected, a character object
use.regression	binary parameter indicating whether additional regressors will be used for forecasting in multivariate settings.
seasonalityMethodId	Seasonality method Id (0- HW, 1- generalized).
levelMethodId	Level method Id.
useSmoothingMethodForError	if the non-standard function for error size should be used, one based on smoothed innovations or surprises

Value

an Rlgt skeleton model

Rlgt documentation built on Sept. 16, 2023, 1:08 a.m.