Bookmark and Share

Notice: On March 31, it was announced that Statalist is moving from an email list to a forum. The old list will shut down at the end of May, and its replacement, statalist.org is already up and running.


[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: anyone using xsmle-pls help


From   Federico Belotti <f.belotti@gmail.com>
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: anyone using xsmle-pls help
Date   Wed, 15 May 2013 22:50:50 +0200

The current version of -xsmle- does not directly allow any of these tests. 
Nonetheless, you can obtain information criteria statistics or perform LR, Wald or specification (Hausman) testing using conventional Stata commands. Some examples may be found at the following URL

http://www.econometrics.it/?p=312#comments

To my knowledge, some of the tests you have mentioned can be performed using Shehata's -spregrext- command.

HTH
Federico
 
On May 15, 2013, at 7:51 PM, ken first wrote:

> Dear Federico,
> 
> Thank you very much for your help.
> 
> I am wondering as xsmle does provide other tests--- such as
> Breusch-Godfrey Test:,  AIC, GLOBAL Moran MI Test or Spatial lag
> likelihood ratio tests. What would you suggest to do for testing these
> for panel data?
> 
> Also if you know how to conduct lagrange multiplier tests (LM-ERR and
> LM-LAG) and Breusch-Pagan tests for panel dataset?
> 
> Many thanks--really appreciate your help
> 
> Ken
> 
> 
> 
> On 15 May 2013 14:51, Federico Belotti <f.belotti@gmail.com> wrote:
>> Dear Ken
>> 
>> Your conformability error is due to the size of the spatial weights matrix.
>> -xsmle- requires a (n x n) matrix of spatial weights, where n is the number of units (in your case n=30). This matrix can be a Stata matrix  or a -spmat- object and it may follow any spatial weighting scheme, though it is usual to normalize spatial weights so that either the row or column sums are equal to 1. Furthermore, -xsmle- does not allow the use of multiple W. This means that the spatial weights matrix is forced to be the same for each cross-section and the command will automatically replicate it for all time periods.
>> 
>> The reason why -xsmle- allows the use of -spmat- objects is to overcome Stata's matrix limits (which depends on Stata's flavor: 40x40 (Small) 800x800 (IC) and 11,000x11,000 (SE or MP)). Thus, bigger matrices can be specified as -spmat- objects.
>> 
>> HTH,
>> Federico
>> 
>> On May 15, 2013, at 3:19 PM, ken first wrote:
>> 
>>> Dear Stata users,
>>> 
>>> could someone please help with this matrix problem? I got this message
>>> when using xsmle.
>>> 
>>> 
>>> I have a panel data for 10 years and 30 regions.
>>> 
>>> 
>>> 
>>> First, I have created a panel matrix using spawmat. The matrix is
>>> similar to the sample file you provided.
>>> 
>>> 
>>> 
>>> =93spatwmat, name(abc) xcoord( longitude ) ycoord( latitude ) band(0 5)
>>> standardize
>>> 
>>> 
>>> 
>>> 1. Inverse distance weights matrix abc (row-standardized)
>>> 
>>>  Dimension: 300x300
>>> 
>>>  Distance band: 0 < d <=3D 5
>>> 
>>>  Friction parameter: 1
>>> 
>>>  Minimum distance: 0.0
>>> 
>>>  1st quartile distance: 7.8
>>> 
>>>  Median distance: 12.2
>>> 
>>>  3rd quartile distance: 17.6
>>> 
>>>  Maximum distance: 47.1
>>> 
>>>  Largest minimum distance: 14.14
>>> 
>>>  Smallest maximum distance: 24.60
>>> 
>>> 
>>> 
>>> Beware! 30 locations have no neighbors
>>> 
>>>  You are advised to extend the distance band =93
>>> 
>>> 
>>> 
>>> 
>>> 
>>> 2. I have then use xsmle and received the following message
>>> 
>>> 
>>> 
>>> xsmle variable a b c d, wmat(abc)
>>> 
>>> 
>>> 
>>>                      *:  3200  conformability error
>>> 
>>>           _xsmle_est():     -  function returned error
>>> 
>>>                <istmt>:     -  function returned error
>>> 
>>> 
>>> 
>>> I have also converted the stata matrix into a spmat file and still get the
>>> same message.
>>> 
>>> 
>>> 
>>> I have also tried to use spmat to create matrix as the help indicates but
>>> spmat seems to be for creating matrices for cross sectional data.
>>> 
>>> 
>>> 
>>> 
>>> 
>>> It would be highly appreciated if you could help with this. thank you in
>>> advance .
>>> 
>>> Ken
>>> *
>>> *   For searches and help try:
>>> *   http://www.stata.com/help.cgi?search
>>> *   http://www.stata.com/support/faqs/resources/statalist-faq/
>>> *   http://www.ats.ucla.edu/stat/stata/
>> 
>> --
>> Federico Belotti, PhD
>> Research Fellow
>> Centre for Economics and International Studies
>> University of Rome Tor Vergata
>> tel/fax: +39 06 7259 5627
>> e-mail: federico.belotti@uniroma2.it
>> web: http://www.econometrics.it
>> 
>> 
>> *
>> *   For searches and help try:
>> *   http://www.stata.com/help.cgi?search
>> *   http://www.stata.com/support/faqs/resources/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/faqs/resources/statalist-faq/
> *   http://www.ats.ucla.edu/stat/stata/

-- 
Federico Belotti, PhD
Research Fellow
Centre for Economics and International Studies
University of Rome Tor Vergata
tel/fax: +39 06 7259 5627
e-mail: federico.belotti@uniroma2.it
web: http://www.econometrics.it


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


© Copyright 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   Site index