Intervals in SARIMA prediction are not displayed

[andypol]

This is final prediction. The interval boundaries are also written in the legend, but they simply do not appear. The weird thing is that if I try a deterministic model on the same set, they will show up in that picture. But once I estimate the stochastic model and specify that the GLS method be used, it simply doesn't work.

[EViews Gareth]

Sorry, you're not being very clear. Your image shows a group. What's that group? Where did it come from? What's the output of the forecast?

[andypol]

I am not really sure about what you mean. Here is the estimation output. In the picture there is a group consists of : forecast of number of flats, the original variable and the upper and lower limits of the confidence interval. Original variable comes from an excel file on my computer, flatsf is prediction created in eviews as well as limits of the interval by entering the command: series down=flatsf-1.96*se and series up=flatsf+1.96*se.

[andypol]

But I feel there is either problem with the file or in the program because when I tried a different computer it doesn't work as well.

[andypol]

Alright, after several hours of desperate searching for reason, I probably found why is that happening.when I don't leave the item coef. uncertainty selected, so the intervals are finally displayed. But my friend has the same version of the program and does not have to do this. Could it be because of some update?

*different file but also a stochastic model

[startz]

andypol,

Part of the problem is that you know exactly what you've done, but no one on this end does. It probably feels obvious to you, but it isn't.

For example, the picture does not appear to be what happens when you push Ok in the forecast box. Where did the variables DOWN and Up come from?

Explain like the forum readers are idiots...