Hello,
I assume you're running the 64-bit version of EViews. When you try to create the large group of dummies, does EViews utilize all the available memory on your system (a simple check of Windows Task Manager should be sufficient)?
three-dimensional panel data
Moderators: EViews Gareth, EViews Jason, EViews Steve, EViews Moderator
-
- EViews Developer
- Posts: 563
- Joined: Thu Apr 25, 2013 7:48 pm
-
- Non-normality and collinearity are NOT problems!
- Posts: 3775
- Joined: Wed Sep 17, 2008 2:25 pm
Re: three-dimensional panel data
eviewsaoa wrote:I'll phrase the doubt differently. I'm not specifically interested in getting the values of the country-pair dummies, but I do want them to be included in my regression (i.e., I want the coefficients to be conditional on these pair-fixed effects). This is something really common when using this type of techniques, so I'm not really asking for any econometric doubt. My dataset is large enough to be able to do this in a consistent manner, and I do know another software that is capable of doing this really easily. But I want to stick to Eviews.
How many observations do you have per pair?
And what's the other software and how does it do it? Maybe that will give someone here a clue about how to do it in EViews?
Re: three-dimensional panel data
I have the 64-bit version of EViews, indeed. I have 6 observations per pair.
The software I was referring to is Stata. They have a wide range of commands where country-pair fixed effects are taken into account in the estimation. For example, xtpqml or xtpoisson
The software I was referring to is Stata. They have a wide range of commands where country-pair fixed effects are taken into account in the estimation. For example, xtpqml or xtpoisson
-
- EViews Developer
- Posts: 563
- Joined: Thu Apr 25, 2013 7:48 pm
Re: three-dimensional panel data
I still suspect that the core issue is the quantity of time/memory needed to explicitly process/store the dummy variables. While including such dummy variables is conceptually equivalent to a fixed effects model, efficiently estimating fixed effects involves selective demeaning of the data rather than additional regressors. EViews provides this feature for 2D cross sections, and Stata apparently does so for larger dimensions as well. Taking a step back from your original question regarding creating a large number of dummy variables, can you reorganize your data and estimation as a 2D panel, where the cross section id is the Cartesian product of the country ids? You can then use EViews' native (efficient) support for fixed effects rather than having to manually include country-country dummies.
-
- Non-normality and collinearity are NOT problems!
- Posts: 3775
- Joined: Wed Sep 17, 2008 2:25 pm
Re: three-dimensional panel data
Does EViews offer fixed effects for a Poisson model?
-
- EViews Developer
- Posts: 563
- Joined: Thu Apr 25, 2013 7:48 pm
Re: three-dimensional panel data
Hmm, it appears not.
Who is online
Users browsing this forum: No registered users and 30 guests