以下是引用reconomics在2005-8-16 11:10:17的发言： 这个问题我也有疑问，希望高手能给予指点！[em24][em24][em24]

也很想知道，期待高手解答

Please, Does anybody know how to implement cointegration tests with panel?

Roberto Zanola Dipartimento di Economia Universit?degli Studi di Torino Via Po 53 10124 Torino [Italy]

-------------------------------------------------------------------------------- If you could give us more details of your panel data, it would help. I assume that you are testing to see if the linear combination of two or more non-stationary time series is stationary . If so, I suggest that you look into Eviews, by Quantitative Micro Software. SAS has unit root tests, but I am not sure they have any co-integration tests. You may be able to use SAS's State-space methods along with co-integration tests to test for co-integration. It involves some coding. In Eviews it is the co-integration tests and examples are there. I suggest you check their website. Their phone number is 714-856-3368. Please try http://www.estima.com may also contain some info about co-integration.

Hope this helps.

Kattamuri Sarma

[此贴子已经被作者于2005-8-21 22:10:18编辑过]

我找到这个检测的方法并且可以用的程序了。是RATS。不过，总是输入不正确数据。谁知道，怎样把panel data从excel转到RATS吗？想知道这个程序code的，请跟贴。

好像没有人说哦