On 6/18/07, Julia Proudnikova <[EMAIL PROTECTED]> wrote: > Hello, > > We have a problem with function lmer. This is our code: > > Get_values<-function(ff_count, fixed_factors, rf_count, random_factors, > y_values) > { > SA<-matrix(as.array(c(fixed_factors, random_factors)), ncol=3) > data<-as.data.frame(SA) > y<-as.array(y_values) > > dd<-data.frame(SA) > for(i in 1:(ff_count+rf_count)){ > dd[,i]<-as.factor(data[,i]) > } > > fit_full=lmer(y~dd[,1]+dd[,2]+(1|dd[,3]),method="ML") > fit_full > } > > A<-c(0,0,0,0,1,1,1,1,0,0,0,0,1,1,1,1,0,0,0,0,1,1,1,1,0,0,0,0,1,1,1,1) > B<-c(0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1,0,1) > C<-c(0,0,1,1,0,0,1,1,0,0,1,1,0,0,1,1,0,0,1,1,0,0,1,1,0,0,1,1,0,0,1,1) > Y<-c(5,3,4,1,1,2,6,1,5,3,7,1,2,3,1,1,5,3,4,1,1,2,6,1,5,3,7,1,2,3,1,1) > r<-Get_values(2, c(A,B),1,c(C),Y) > r > > R output: > Error in inherits(x, "factor") : object "dd" not found > > Can this function work with random array? Because this code is > working:
The full explanation of why lmer fails to find dd has to do with the way names are resolved in a call to model.frame. However, there may be a way to solve your problem by redesigning your function so you don't need to worry about what model.frame does. Why not pass the data as a data frame and pass the names of the fixed factors, random factors and response variable as character strings? Your current design of creating a matrix, then converting it to a data frame then converting numeric variables back to factors is a bit convoluted. If you knew that you were only going to have one random factor you could generate the formula as substitute(y ~ ff + (1|rf), list(y = as.name(y_name), ff = parse(paste(ff_names, collapse = "+")), rf = as.name(rf_name)) It gets a bit trickier with multiple random factors. Having said all this, it does appear that the call to model.frame inside lmer is getting the wrong environment from the formula and I will correct that. If you need more detail about the redesign I am suggesting, feel free to contact me off-list. ______________________________________________ R-help@stat.math.ethz.ch 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.