File talk:ArTimeSeries.svg
Appearance
It would be nice is a description could be added as to what each line is. I see the value on the right, but there isn't any indication what that value is (\phi?)
(120418, everettr2) I agree, the context for the figure is not clear. I think we have no data at the top, with order=1, but with φ_1 = 0, so we are only seeing the noise term (with Standard Deviation =1.) The next row is with AR(1) where the single term φ_1 = 0.3, and we are getting some data in and also have the error or noise term added. The third row is with φ_1 = 0.9. The fourth and fifth rows are for order=2 (AR(2)), so we have two values to specify φ_1 and φ_2. Questions that probably need to be asked:
- Are these plots based on some data series and is there a clue as to what that series might be, or - Are these plots based entirely on the action of the AR model, as defined by the given φ, on some randomly generated noise term? - If this is so, then have we set the random number generator to produce the same sequence for the noise for each of these plots? This would be nice if we are trying to compare the different plots in the figure. - Are there any features in these plots that should be pointed out or explained? - What is the principle value of the figure? Without some more context, we just seem to be looking at some structure + noise, or maybe noise processed to have some sort of structure on it. - Is there some text we should add to the article to tie in with this figure? I don't see any such text in the history for the article, that might have been deleted, leaving the figure orphaned. — Preceding unsigned comment added by Everettr2 (talk • contribs) 20:33, 18 April 2012 (UTC)