Statalist


[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

RE: st: RE: SUR with panel data


From   "Karfakis, Panagiotis (ESAF)" <[email protected]>
To   [email protected]
Subject   RE: st: RE: SUR with panel data
Date   Thu, 21 Aug 2008 17:01:04 +0200

missing observations is a different problem onto which I don't have any
particular knowledge/advice
sorry!!
good luck

-----Original Message-----
From: [email protected]
[mailto:[email protected]] On Behalf Of [email protected]
Sent: 21 August 2008 16:53
To: [email protected]
Subject: Re: st: RE: SUR with panel data


Dear Panagiotis,

many thanks for the answer. Actually, my doubt about xtsur is that (as  
I understand from the structure of the example dataset  
xtsur_example.dta) no missing observation on c can be accomodated,  
i.e. for each given x and t I should have information for ALL the  
commodity classes, isn't it? Obviously, I would not drop out all  
observation with missing c.
Kind regards,

Vittorio Marzano


Quoting "Karfakis, Panagiotis (ESAF)" <[email protected]>:

> have you tried: xtsur
> if its not installed write in command line:
>
> net search xtsur
>
> best
>
> -----Original Message-----
> From: [email protected]
> [mailto:[email protected]] On Behalf Of 
> [email protected]
> Sent: 21 August 2008 16:27
> To: [email protected]
> Subject: st: SUR with panel data
>
>
> Dear statalist users,
>
> I am going to estimate an International trade model using a database 
> with observations related to o-d pairs of countries (x) for different 
> years (t) and commodity classes (c). This is therefore a panel 
> dataset, and in addition there is potential correlation between 
> commodity classes equations for the o-d pair x at the year t. 
> Moreover, the dataset is not complete, since some c?s are missing for 
> xt pairs and some t?s are missing for xc pairs.
>
> I should therefore perform a SUR estimation within a panel data
> framework: how can I do it using STATA? I understand that sureg and 
> xtreg solve separately the two problems, but is there a way to combine 
> them? I would also be glad to know which tests can be run in order to 
> contrast this estimation with a simple panel data estimation (i.e. 
> neglecting correlations among commodity classes).
>
> Thanks in advance.
>
>
>
> ----- Fine del messaggio inoltrato -----
>
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
> *
> *   For searches and help try:
> *   http://www.stata.com/help.cgi?search
> *   http://www.stata.com/support/statalist/faq
> *   http://www.ats.ucla.edu/stat/stata/
>



*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/
*
*   For searches and help try:
*   http://www.stata.com/help.cgi?search
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/



© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index