R script file including a set of forecasting methods. This script acts as a wrapper of some state of the art forecasting methods, implemented in packages like the well-known forecast. Apart from that, it also implements some other useful aspects, like wrapping the elastic package to allow easy metrics, factors, and strategic indicators gathering from an elasticsearch DB. The script also implements saving and reusing the fitted forecasting models, in order to save time and computation resources.
The script is intended to be used along with the q-rapids-forecast library, but is prepared to be used alone as well.
The available methods and the corresponding R package implementing them are as follow:
| Method Name | R Package |
|---|---|
| Arima | forecast |
| Arima (force seasonality) | forecast |
| Theta | forecast, forecastHybrid |
| ETS | forecast |
| ETS (force damped) | forecast |
| Bagged ETS | forecast |
| STL | forecast |
| Neural Network (nnetar) | forecast |
| Hybrid | forecastHybrid |
| Prophet | prohpet |
The script depends on the following packages:
- forecast
- devtools
- elastic
- forecastHybrid
- prophet
These required packages should be installed beforehand using the following commands:
install.packages("devtools")
library(devtools)
install_version("forecast", version="8.7")
install_version("prophet", version="0.4")
install_version("Rserve", version="1.7-3.1")
install_version("elastic", version="0.8.4")
install_version("forecastHybrid", version="4.2.17")
Software licensed under GNU Lesser General Public License v3.0. Permissions of this copyleft license are conditioned on making available complete source code of licensed works and modifications under the same license or the GNU GPLv3. Copyright and license notices must be preserved. Contributors provide an express grant of patent rights. However, a larger work using the licensed work through interfaces provided by the licensed work may be distributed under different terms and without source code for the larger work.