arma structure correlogram
Posted: Mon Nov 20, 2017 10:31 am
It looks like the autocorrelations in the arma structure correlogram are being drawn off by one lag. I estimated an MA(1) and asked for the arma structure correlogram and got
Note that the "Actual" spike begins at lag zero. It should begin at lag 1, I realize this is in part to avoid having the actual and theoretical on top of one another, but it would be much better if they touched or something.
For the record, here is the correlogram done straight.
- bug 1.png (20.72 KiB) Viewed 2657 times
For the record, here is the correlogram done straight.
- Capture2.PNG (9.49 KiB) Viewed 2657 times