[prev in list] [next in list] [prev in thread] [next in thread] 

List:       sas-l
Subject:    Re: Help on proc mixed error message
From:       Edzard van Santen <evsanten () ACESAG ! AUBURN ! EDU>
Date:       2006-07-31 17:25:29
[Download RAW message or body]

if it is a repeated measures analysis you may have specified the wrong subject stratum. Suggest
you run proc tabulate and create a table with the number of non missing obs for factor
combinations. Is the output what you expected/ Strange things can happen when data are
transferred.



On Mon, 31 Jul 2006 17:46:36 +0100, Joakim Englund <joakim.englund@GMAIL.COM> wrote:

>Hi!
>
>When I ran proc mixed recently I got the following error message:
>
>NOTE: An infinite likelihood is assumed in iteration 0 because of a
>nonpositive residual variance estimate.
>
>WARNING: Stopped because of infinite likelihood.
>
>
>The program is used at a routinely basis and has never before produced
>this type of log message. There is nothing unusual about the input
>data.
>
>What can potentially cause this problem, i.e. how can I get negative
>residual variance estimates?
>What can I do about it?
>
>Thanks forehand,
>Joakim Englund

[prev in list] [next in list] [prev in thread] [next in thread]