'ATAforecasting' Modelling Interface for 'fable' Framework

Allows ATA (Automatic Time series analysis using the Ata method) models from the 'ATAforecasting' package to be used in a tidy workflow with the modeling interface of 'fabletools'. This extends 'ATAforecasting' to provide enhanced model specification and management, performance evaluation methods, and model combination tools. The Ata method (Yapar et al. (2019) ), an alternative to exponential smoothing (described in Yapar (2016) , Yapar et al. (2017) ), is a new univariate time series forecasting method which provides innovative solutions to issues faced during the initialization and optimization stages of existing forecasting methods. Forecasting performance of the Ata method is superior to existing methods both in terms of easy implementation and accurate forecasting. It can be applied to non-seasonal or seasonal time series which can be decomposed into four components (remainder, level, trend and seasonal).

fable.ata

'ATAforecasting' Modelling Interface for 'fable'

This package provides a tidy R interface to the Ata method procedure using fable. This package makes use of the ATAforecasting package for R.

Installation

You can install the stable version from CRAN:

install.packages("fable.ata")

You can install the development version from Github with:

# install.packages("devtools")
devtools::install_github("alsabtay/fable.ata")

Example

USAccDeaths: Accidental Deaths in the US 1973--1978

library(fable.ata)
as_tsibble(USAccDeaths) %>% model(ata = AutoATA(value ~ trend("A") + season("A", method = "stl"))) %>% forecast(h=24)

Links

Github page

Github.io page

Github - ATAforecasting

Github.io - ATAforecasting

Project team website

Github - Intermittent Ata Method Package

Github.io Intermittent Ata Method Package

License

This package is free and open source software, licensed under GPL-3.