Dear Peter, Thanks for your concise answer, it works perfectly.
By the way, I fully agree that "data" or "df" are not good names for data.frames and I am/was aware of that and I usually avoid those names (not consequently though I've to admit, it is too tempting ;). However, if one uses those evil names, one cannot expect to receive meaningful error messages. Thus, I was not astonished by the peculiar error message itself (in fact I was well aware that this has to do with the bad naming and the fact that "data" is, above all, a function) and I suspect the error to be due to environment issues. I tried the workaround with passing the very same data argument explicitly to update: > update(models[[1]], . ~ ., data = dat) which worked but which left the stale impression of redundancy and even more dangerous error proneness: what happens if the name of the data frame is changed earlier? Finally, your suggestion with > update(models[[1]], . ~ ., data = model.frame(models[[1]])) solved all the issues (and I was wondering why I did not try it out myself, so obviously I was not seeing the wood for the trees). So, thanks a lot for your help. Have a nice day. KR, -Thorn ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
