I have been using lda for a while with success but recently re-installed my Ubuntu system and R. Now when I run R - lda I get a question about a variable [16 in the specific case] that does not exist. There are only 14 variables + one factor variable, in the data-frame! I am running the same model with the same set of variables that I ran previously [but a smaller group of 62 samples]. I tried re-installing base-R, which contains the MASS library but got the same result.
I had run a previous session with 16 variables but did rm(list=ls(all=TRUE)) to make sure I flushed the buffers and started a new R session. I even tried shutting down the system and starting from scratch.