Statalist The Stata Listserver


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

st: error in ML computation GLLAMM


From   Inna Khousnoullina <inna.khousnoullina@uni-konstanz.de>
To   statalist@hsphsun2.harvard.edu
Subject   st: error in ML computation GLLAMM
Date   Fri, 08 Sep 2006 15:19:50 +0200

Dear all,

I can't find en error in my computation of a gllamm-model. The same
model provided useful estimation of coefficients and variances for
other dependent variables of my dataset. I don't understand why the
same model does not function with another dependent variable.


lnf38_4 is my depvar, I control for modus (modus==1), intnr is 2-level
and point 3rd-level.

hier my output:

. gllamm lnf38_4 if modus==1, i(intnr point) adapt
(error occurred in ML computation)
(use trace option and check correctness of initial model)

. gllamm lnf38_4 if modus==1, i(intnr point) adapt trace
 
General model information
-----------------------------------------------------------------------------
dependent variable:         lnf38_4
family:                     gauss
link:                       ident
equation for fixed effects  _cons
  
Random effects information for 3 level model
-----------------------------------------------------------------------------
***level 1 equation:
 
   log standard deviation
   lns1: _cons
 
 
***level 2 (intnr) equation(s):
 
   standard deviation of random effect
   intn: _cons
 
***level 3 (point) equation(s):
 
   standard deviation of random effect
   poin: _cons
 
number of level 1 units = 2137
number of level 2 units = 416
number of level 3 units = 160
 
Initial values for fixed effects
 

      Source |       SS       df       MS              Number of obs =  
 2137
-------------+------------------------------           F(  0,  2136) =  
 0.00
       Model |           0     0           .           Prob > F      =  
    .
    Residual |  4384.96984  2136   2.0528885           R-squared     = 
0.0000
-------------+------------------------------           Adj R-squared = 
0.0000
       Total |  4384.96984  2136   2.0528885           Root MSE      = 
1.4328

------------------------------------------------------------------------------
    __00000C |      Coef.   Std. Err.      t    P>|t|     [95% Conf.
Interval]
-------------+----------------------------------------------------------------
       _cons |   1.626379   .0309942    52.47   0.000     1.565597   
1.687161
------------------------------------------------------------------------------
  
start running on  8 Sep 2006 at 15:05:07
 
Non-adaptive log-likelihood: -3638.4494
 
First iteration of adaptive quadrature:
 
Updating posterior means and variances
log-likelihood: 
-3638.4556 -3638.4556 
------------------------------------------------------------------------------
Iteration 0:
Coefficient vector:
     lnf38_4:     lns1:     intn:     poin:
       _cons     _cons     _cons     _cons
r1  1.626379  .3596239        .5        .5

overflow at level 1 ( 1 missing values)
Hessian has become unstable or asymmetric (D)
(error occurred in ML computation)
(use trace option and check correctness of initial model)
finish running on  8 Sep 2006 at 15:06:53
  
I would be thankful for any advice.

Inna Khousnoullina
. 


*
*   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 1996–2014 StataCorp LP   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index