Skip to content

termplot

2 messages · Robert Michael Inman, John Fox

#
I have used glm and stepAIC to choose a best model.  I can use termplot to
assess the contribution of each explanatory variable in the glm.  However
the final model after running stepAIC includes interaction terms, and when I
do termplot I get "Error in `[.data.frame`(mf, , i) : undefined columns
selected".  I also see the termplot detail saying "Nothing sensible happens
for interaction terms."   

Is it possilbe to determine explanatory power of final set of variables when
interactions are included?  
 
Thanks

- Bob
#
Dear Bob,

Take a look at the effects package, described in
<http://www.jstatsoft.org/v08/i15/paper>.

I hope this helps,
 John

------------------------------
John Fox, Professor
Department of Sociology
McMaster University
Hamilton, Ontario, Canada
web: socserv.mcmaster.ca/jfox
On
I
happens
when
http://www.R-project.org/posting-guide.html