EViews.com

I have an unbalanced panel which consists of: 1) hundreds of unique individual id; 2) a group id (A,B,C); 3) T time periods; 4) other variables

I want to create a panel in which the individuals are aggregated by group, so that the panel is 3xT.

I have created a new page using group id as the cross id, then performed a matched merge, contracting using the mean. I must be doing something wrong because the values of the other variables (weight, height, etc.) are the same across groups for corresponding time periods. Is there a step-by-step example of how to correctly do what seems like should be a common application?

Could you provide your workfile (or an example of similar structure)?

Attached is a workfile. I want to aggregate dq by proptype and activitydt. Then in another page I want to aggregate dq by purposetype and activitydt. BTW I am using EVIEWS 8 Enterprise 64 bit.

These are the steps I took, which seemed to work:

http://www.eviews.com/Forum_support/Pan ... llapse.htm

Thank you that does the trick!