EViews.com

Hi there,

I'm not so familiar with eviews and I want to analyze my data with the structural break unit root model of Perron (1989). I do not have to test for unit root. And the data of the breaks are known.

I have data of 3 brands in 40 stores from the same supermarket concern and consists of 117 weeks. So, in total I've got 6880 observations.
The first question is how I make a workfile of this weeks. I know how to make it in years or months, but in weeks that's a different story I know.

The second question is how I regress the model. Can someone help me with the process? I never worked with Eviews before, only with SPSS.

I hope my question is clear!

Thank you in advance.

There is an add-in you can download for the related Zivot-Andrews unit root test. Go to Add-ins/Download Add-ins and get the ZAURoot add-in. [edit] It sounds as though you want to structure your workfile as a panel (search this forum for links), but note that there is no panel component to either the Zivot-Andrews or the Perron test. You would have to compute for each cross-section separately.