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]

Re: st: Separate intercept mixed model


From   rgutierrez@stata.com (Roberto G. Gutierrez, StataCorp)
To   statalist@hsphsun2.harvard.edu
Subject   Re: st: Separate intercept mixed model
Date   Tue, 08 Feb 2011 13:05:06 -0600

Michael Mitchell <Michael.Norman.Mitchell@GMAIL.COM> asks:

> I am trying to estimate a mixed (multilevel) model using a separate
> intercept approach, and am getting an error saying that the likelihood
> evaluates to missing. Here is the model I am estimating with the error...

> . webuse pig, clear
> (Longitudinal analysis of pig weights)
> . replace week = week - 1
> (432 real changes made)
> . * create 2 groups, trt and control
> . generate trt = (id > 24)
> . * make weight 10 pounds more for treatment
> . replace weight = weight + 10     if trt == 1
> (216 real changes made)
> . * Model 0:
> . * Enter trt as factor variable
> . xtmixed weight ibn.trt c.week , nocons || id:

> Performing EM optimization:
> likelihood evaluates to missing
> r(430);

Michael then explains that this should work because he generates the
indicators manually and then runs -xtmixed- with these instead.  He is
right.

Michael has discovered a bug in how -xtmixed- is mishandling the "ibn."
factor notation for including all levels of a factor variable, including
the base category.

This will be fixed in the next ado update.

--Bobby
rgutierrez@stata.com
*
*   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