Stata The Stata listserver
[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: gllamm_"insufficient information"?


From   Stas Kolenikov <[email protected]>
To   [email protected]
Subject   Re: st: gllamm_"insufficient information"?
Date   Fri, 30 Jul 2004 15:00:59 -0400 (EDT)

> When I run the model: Gllamm depvar indepvar, i(idfirm, idanalyst,
> idyear) trace
>
> a message saying that there are "insufficient observations". Does
> this mean that -gllamm- requires at least 2 observations per group
> (in my case per analyst in the same year for a specific firm) to be
> able to run or my data are inappropriately arranged? I could not find
> anything about this on FAQ or in the book.

-gllamm-'s identifiers that you specify in -i()- option are taken to
indicate hierarchy of the observations. Thus Stata believes that all firms
are nested into analysts, and analysts in turn are nested into years. This
is not how your data is organized, and thus -gllamm- is unhappy about it.

You would have to use Stata's outdated and no longer documented -eq-
command; see examples in the end of the help file. The GLLAMM book sold
through Stata webstore should have it all that :).

Here's my example:

clear
set obs 20
g id = _n
g eid = invnorm(uniform())
expand 10
bysort id : gen year = _n
bysort year (id): gen eyear = 0.3*invnorm(uniform())
* created the panel structure
gen who = int(uniform()*5)
bysort who (id year) : g ewho = 0.5*invnorm(uniform())
* added 5 experts
g x = uniform()
g y = x + eid + ewho + eyear + 0.2*invnorm(uniform())
* variances:
* firm effect : 1^2   = 1
* year effect : 0.3^2 = 0.09
* expert effect : 0.5^2 = 0.25
* error term : 0.2^2 = 0.4

g one = 1
eq year : one
eq who : one
gllamm y x , i(id) eqs(year who) nrf(2)

It has not converged in five minutes on my laptop, so with your 1000 of
firms... you'd be best off leaving it running on your departmental server
overnight.

 ---                                    Stas Kolenikov
 --       Ph.D. student in Statistics at UNC-Chapel Hill
 - http://www.komkon.org/~tacik/  -- [email protected]

* This e-mail and all attachments to it are not intended to provide any
* reasonable point of view and was transmitted to you in error. It
* should be immediately deleted by all recipients unless they really
* enjoy communicating with the author :). Other restrictions apply.

*
*   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/



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