Timeline for Outliers spotting in time series analysis, should I pre-process data or not?
Current License: CC BY-SA 3.0
16 events
| when toggle format | what | by | license | comment | |
|---|---|---|---|---|---|
| Apr 13, 2017 at 12:44 | history | edited | CommunityBot |
replaced http://stats.stackexchange.com/ with https://stats.stackexchange.com/
|
|
| Mar 7, 2012 at 9:17 | vote | accept | Bruder | ||
| Feb 17, 2012 at 11:53 | answer | added | Rob Hyndman | timeline score: 6 | |
| Feb 17, 2012 at 11:26 | comment | added | Rob Hyndman | You're quite right -- I should have used robust=TRUE in my post on using stl to find outliers. Setting s.window to some number allows the seasonality to change slowly over time -- the smaller the number the more rapid the change. Try s.window=5 for example. | |
| Feb 17, 2012 at 9:17 | history | edited | mpiktas | CC BY-SA 3.0 |
deleted 1 characters in body
|
| Feb 17, 2012 at 9:07 | history | edited | Bruder | CC BY-SA 3.0 |
deleted 2 characters in body
|
| Feb 17, 2012 at 8:58 | answer | added | Bruder | timeline score: 4 | |
| Feb 17, 2012 at 2:13 | answer | added | IrishStat | timeline score: 3 | |
| Feb 16, 2012 at 22:02 | history | edited | Bruder | CC BY-SA 3.0 |
added 1 characters in body
|
| Feb 16, 2012 at 21:31 | comment | added | Michelle | I'm at work at the moment, I'll favourite this question and look at the data when I'm home in around 8 hours, if no one else responds in that time. | |
| Feb 16, 2012 at 21:18 | comment | added | Bruder | Above time serie goes from Jan 2005 to Dec 2011 | |
| Feb 16, 2012 at 21:14 | comment | added | Bruder | There is no underlying seasonality effect i can think of. We sell circuit breakers so... There is a valley in August and December because the factory shuts down for summer/winter holidays. I thought calendar adjustment would get rid of that effect, but the adjusted data still exhibit a some lower than average values. I guess people tend to buy less in those months because they know the company will close and so they'll have to wait longer for their product. They stock up in July and November and refill in September/January. But that's only a wild guess. | |
| Feb 16, 2012 at 19:41 | comment | added | Michelle |
A structural break can affect seasonality as well, but it depends on the underlying data (e.g. military changing their intake numbers and time of year!). For seasonality, is there any theory behind those effects, e.g. you're selling winter clothing, or people tend to buy the products on the weekend, etc? The last time I used it, stl kept the same seasonality trend across all data - this may be an incorrect assumption.
|
|
| Feb 16, 2012 at 19:37 | comment | added | Bruder | Oooh, interesting, so it could be more than a simple level shift. I never thought of this. So I try STL decomposition before and after 2009 to see what changes? And what about seasonality? I can't understand if calendar effects are supposed to be part of seasonal component or not... Would it help if I posted the values of my time serie? | |
| Feb 16, 2012 at 19:31 | comment | added | Michelle | If you fit one time-series to the pre-shift data in 2009 and a separate series to the post-shift data, what do your two times series look like? You could have a structural break which means that there is a different model for the newer data. | |
| Feb 16, 2012 at 10:23 | history | asked | Bruder | CC BY-SA 3.0 |