st: xtpmg stopps with conformability error

[Jan Ditzen <jan@ditzen.net>]

Dear Statalist Users,
I am running a Monte Carlo Simulation on a panel with a different number 
of countries. Among others, I would like to use the pooled mean group 
estimator. Thus I am using the user-written command "xtpmg". If I 
increase the number of countries (N) from 85 to 86 the command crashes 
with the error: conformability error r(503). Obviously there is a matrix 
calculation which is not possible. I enabled the trace option to see 
where it crashes and the last few lines are:

    - matrix `param'=nullmat(`param') , e(b)
    = matrix __00000D=nullmat(__00000D) , e(b)
    - quie matrix accum `cpsr'=`ec' $SRx if `touse' & `ivar'==`i', 
`constant'
    = quie matrix accum __00000J=__ec D.slsk_d D.slsh_d D.slngd_d if 
__000000 & id==86, noconstant
    - matrix `G'[(`j'-1)*(`ks')+1,(`j'-1)*(`ks')+1]=`cpsr'/(`sig')
    = matrix 
__00000N[(86-1)*(__000008)+1,(86-1)*(__000008)+1]=__00000J/(__00000G)
conformability error
      local j=`j'+1
      }
----------------------------------------------------------------------------------------------------------------- 
end xtpmg.Estimate ---
---------------------------------------------------------------------------------------------------------------------------- 
end xtpmg ---
r(503);


Has anyone encountered this problem as well and has a solution (besides 
reducing the number of observations) at hand?

Thanks a lot in advance
Jan Ditzen

--
Jan Ditzen
Wackerhagen 8
20537 Hamburg
- Germany -

Email:jan@ditzen.net
Skype: janditzen
Tel: +49176 23894413

*
*   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/