AutoForecaster

AutoForecaster is an R package to automate the production of a collection of time series models applied to tsibble data structures defined by index and key.

The pkgdown site and the main github development page.

Installation

# Install development version from GitHub
devtools::install_github("robertwwalker/AutoForecaster")

Usage

library(AutoForecaster)

The package contains four (current) forecasting functions.

DayWModelPicker(data, Outcome, index, H.Horizon=14)
MonthModelPicker(data, Outcome, index, H.Horizon=12)
DaysModelPicker(data, Outcome, index, H.Horizon=14)
QuarterModelPicker(data, Outcome, index, H.Horizon=12)

supported by

DayWModelFitter(data, Outcome)
MonthModelFitter(data, Outcome)
DaysModelFitter(data, Outcome)
QuarterModelFitter(data, Outcome)

that fits the models.