SARIMA model is considered a decomposition model. However, in the literature, I only found figures about its forecasts results and none about its components. By cntrast, in the case of STL algorithm od DECOMPOSE or wavelet transform, the components are usually extracted and visualized.
So is it possible to extract the components and visualize them?
Hello,
I am not sure I understand: do you want to access the model's stored results or to visualize the series' components?
Each software (R, Stata, etc) gives you the possibility to access the procedure's stored results. (check documentation).
On the other hand, (S)ARIMA is not a "decomposition model", per se, better try another approach i.e. decomposition / unobserved components model (UCM) analysis.
If you want to isolate and explore the components (trend-cycle, seasonality, remainders), you may use whatever decomposition method you want (STL, X11/12, etc).
For R I recommend "seasonal" (http://www.seasonal.website/seasonal.html) or "feasts" package; For Stata try "ucm" https://www.stata.com/features/overview/unobserved-components-model/
Regards,
C.
- Badges
- Science topic
- Similar topics
- Phytochemistry
- Extracts