QuarterModelPicker: Function to choose optimal forecast model for quarterly data

In robertwwalker/AutoForecaster: Auto Magic Forecaster Optimizes by aicc (Auto Forecaster)

Description

Function to choose optimal forecast model for quarterly data

Usage

QuarterModelPicker(data, Outcome, DateVar, H.Horizon = 12)

Arguments

data	A quarterly tsibble.
Outcome	A valid variable name for the Outcome to be modelled in in 'data'.
DateVar	A valid variable name for the time index in 'data'.
H.Horizon	An integer for the forecast horizon/test subset of 'data'.

Value

A list containing:

• Accuracy.Table: accuracy for the forecast horizon against the test sample.

• test: the test set

• train: the training set

• Model.Fits: the model fits

• Model.Forecasts: the forecasts

• Min.Model: the minimum model by MAE

• Min.Report: the minimum model report

• Min.Res.Plot: the gg_tsdisplay for the minimum model

• Min.Forecast.Plot: a plot of the minimum forecast by MAE

Examples

data(NWSQ)
QuarterModelPicker(NWSQ, Outcome=log(TX), DateVar=Quarter, H.Horizon=12)

robertwwalker/AutoForecaster documentation built on Dec. 22, 2021, 5:13 p.m.