Re: st: xtlsdvc, initial(my)

["V. Sarafidis" <vs242@cam.ac.uk>]

I think the answer is that you don't need the estimate of the error 
variance. Only your initial vector of estimates (your estimate of the LDV 
coefficient needs to be firstfirst).

The reason it says 
"the number of columns of my must be equal to the number of
right-hand variables plus one"

is because when you run xtlsdvc you do not put the lagged dependent 
variable in the model (so you type xtlsdvc y x , i(my) ). Therefore, you 
have 5 - say - regressors in your file plus the lagged dependent variable 
which the command xtlsdvc will create by itself.

I hope this helps

Vasilis

On Sep 7 2005, Miguel Portela wrote:


> Dear all,
>
> I am estimating a dynamic model using XTLSDVC. When I use the option
> 'initial(my)' I get
>
>    "the number of columns of my must be equal to the number of
> right-hand variables plus one"
>
> 'my' is a [1x(k+1)] matrix, which uses the results from a previous
> regression and an estimate of the error variance
>
>       mat my=e(b),evariance
>
> I have checked with matrices of different dimensions, but the answer
> is always the same. Could you please let me know if there is a
> possible solution?
>
> thanks in advance for your help,
>
> cheers,
>
> miguel
--
If I could get my membership fee back, I'd resign from the human race.
Fred Allen (1894 - 1956)


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/