[Rsiena-help] problems with little variation in behavioral dependent variable

[Ruth M. Ripley]

Dear Tobias,

Is it possible that you have no change in this variable between two of 
your waves? It seems that the error occurs when trying to find a suitable 
starting value for the rate parameter.

Regards,

Ruth
--
Ruth M. Ripley,                         Email:ruth at stats.ox.ac.uk
Dept. of Statistics,                    http://www.stats.ox.ac.uk/~ruth/
University of Oxford,                   Tel:   01865 282851
1 South Parks Road, Oxford OX1 3TG, UK  Fax:   01865 272595


On Wed, 15 Sep 2010, Tobias Stark wrote:

> Dear all,
>
> I try to run a three-waves RSiena analysis with a behavioral dependent variable. After specifying all networks, covariates and behavior files, I try to create the list with effects of the dependent variables with the command "getEffects()". However, this is the error I get:
>> paras<-getEffects(project1)
> Error in if (val > 0.9) val <- 0.5 :
>  missing value where TRUE/FALSE needed
>
> I noticed that there is very little variation over time in the behavior variable. When I add a little more variation in the behavior variable, SIENA does read the file without an error.
> I expected SIENA to have problems in the analysis given little variation in the behavior. However, I did not expect that it already has problems reading the files in. Does anybody know what to do?
>
> Thanks in advance,
> Tobias
> _______________________________________________
> Rsiena-help mailing list
> Rsiena-help at lists.r-forge.r-project.org
> https://lists.r-forge.r-project.org/cgi-bin/mailman/listinfo/rsiena-help
>