Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at statalist.org.


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

st: Convergence not achieved in negative binomial regression modeling


From   Vineet Saini <vsaini@ucalgary.ca>
To   "statalist@hsphsun2.harvard.edu" <statalist@hsphsun2.harvard.edu>
Subject   st: Convergence not achieved in negative binomial regression modeling
Date   Sat, 09 Jul 2011 06:30:07 -0600

Dear Statalist users,

I am running stepwise ­nbreg to model antimicrobial resistance in bacteria
and I get the message that "convergence not achieved" as following:


. stepwise, pr (0.95) pe(0.90) :nbreg penr_tot pnbdctuse cs1dctuse lnlctuse
cs3lctuse pnclctuse pnsysuse cs3sysuse tmpsysuse tetsysuse phnsysuse
herddurd_cn zone barntype [pw=weight], exp(farmcowday_tot)  irr
convergence not achieved
convergence not achieved
r(430);



Later, I inserted -iterate (500) in the command, and then got the following
output:

. stepwise, pr (0.95) pe(0.90) :nbreg penr_tot pnbdctuse cs1dctuse lnlctuse
cs3lctuse pnclctuse pnsysuse cs3sysuse tmpsysuse tetsysuse phnsysuse
herddurd_cn zone barntype [pw=weight], exp(farmcowday_tot) iterate(50) irr
                      begin with full model
p < 0.9500            for all terms in model

Negative binomial regression                      Number of obs   =
79
Dispersion           = mean                       Wald chi2(13)   =
226.58
Log pseudolikelihood = -50010.674                 Prob > chi2     =
0.0000

----------------------------------------------------------------------------
--
             |               Robust
    penr_tot |        IRR   Std. Err.      z    P>|z|     [95% Conf.
Interval]
-------------+--------------------------------------------------------------
--
   pnbdctuse |   1.652111   .4883559     1.70   0.089     .9256107
2.948833
   cs1dctuse |   .2436561   .1924028    -1.79   0.074     .0518354
1.145322
    lnlctuse |   1.088507   .1223067     0.75   0.450     .8733506
1.356669
   cs3lctuse |   18.87842   11.36569     4.88   0.000     5.800991
61.43685
   pnclctuse |   .7654628   .0987489    -2.07   0.038     .5944488
.985675
    pnsysuse |   .5867255   .1123691    -2.78   0.005     .4031008
.8539967
   cs3sysuse |   .8746823   .0965915    -1.21   0.225     .7044527
1.086047
   tmpsysuse |   1.464673     .27624     2.02   0.043     1.012052
2.119722
   tetsysuse |   1.657977   .2532772     3.31   0.001     1.228985
2.236714
   phnsysuse |   14.31775   5.312295     7.17   0.000     6.919181
29.62749
 herddurd_cn |   .0261432   .0185816    -5.13   0.000     .0064917
.1052835
        zone |   1.986719   .3173496     4.30   0.000     1.452678
2.717087
    barntype |   .1920082   .1312393    -2.41   0.016      .050294
.7330328
farmcowday_tot| (exposure)
-------------+--------------------------------------------------------------
--
    /lnalpha |  -17.92744          .                             .
.
-------------+--------------------------------------------------------------
--
       alpha |   1.64e-08          .                             .
.

1. Has the model converged and how do I know that?
2. What does the blanks (dots) for robust SE and 95% CI for /lnalpha and
alpha indicate?  That the model has still not converged??

Thanks a lot,
Vineet


-- 
Vineet Saini
 PhD Candidate
 Faculty of Veterinary Medicine
 The University of Calgary
 HS 3330, Hospital Drive, NW
 Calgary, AB, Canada T2N 4N1
 Email: vsaini@ucalgary.ca
 Tel: 403.975.7919 (cell)
        403.210.6813 (work)
 
"Until we see what we are, we cannot take steps to become what we should be"
~ Charlotte P. Gilman ~




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


© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index