Timeline for OLS with ARMA errors, tip or two
Current License: CC BY-SA 4.0
12 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Feb 21, 2019 at 15:10 | vote | accept | Kiril E. Proykov | ||
| Feb 21, 2019 at 15:10 | answer | added | Kiril E. Proykov | timeline score: 0 | |
| Feb 17, 2019 at 22:34 | vote | accept | Kiril E. Proykov | ||
| Feb 21, 2019 at 15:10 | |||||
| Feb 17, 2019 at 22:01 | answer | added | IrishStat | timeline score: 2 | |
| Feb 17, 2019 at 21:54 | comment | added | IrishStat | I would simply leave them in as they seem to absorb/fix the issue the 6 indicators are being measure/tested on a 1 by 1 basis AND not collectively ...whereas the ar(7) is a global/composite efffect. I wonder when a series of very productive comments become an answer .... I guess only a moderator knows for sure !.. | |
| Feb 17, 2019 at 21:47 | comment | added | Kiril E. Proykov | Yep but that's the weird part, when I add them to the original OLS (the one that needs fixing) the dummies are not significant. Then I take the error from the OLS and model an ARIMA structure to it and in the error itself I do not have weekly seasonality but in the individual series in the PACF it picks regularly on the 7th + lag. I guess its very hard to advice without seeing the data. I will ask tomorrow if I can share it. Nevertheless, you have helped me a lot :) Thank you. | |
| Feb 17, 2019 at 21:36 | comment | added | IrishStat | your findings based upon my answer suggest that you add 6 seasonal dummies (6 seasonl pulses) ARiMA doesn't fix determnistic effects | |
| Feb 17, 2019 at 21:34 | comment | added | Kiril E. Proykov | I just checked my series and I can identify very strong weekly seasonality (every 7 days). I guess that causes the funky ARMA(3,7). Weird thing is that, when I take diff(x,7) the seasonality still stays in the data. | |
| Feb 17, 2019 at 21:31 | comment | added | IrishStat | tsoutliers my be useful but I am not sure. I apologize for the deletion of my answer as someone didn't recognize that I had actually answered your question. Do you agree ? | |
| Feb 17, 2019 at 21:20 | comment | added | Kiril E. Proykov | Hi IrishStat, I have filtered the original series for deterministics (but not for outliers). How would advice me to check the residual structure for pulses, level shifts, etc.? | |
| Feb 17, 2019 at 20:54 | comment | added | IrishStat | It sounds to me that there may be deterministic structure in the residuals ( pulses,level shifts,seasonal pulses) or an error variance that is not constant. Post your data in a csv file and I will try and help. | |
| Feb 17, 2019 at 20:40 | history | asked | Kiril E. Proykov | CC BY-SA 4.0 |