To use Ron Ziegler's famous phrase, that argument is "no longer operative". The code for the bxp function, which is what is called to actually do the plotting, has a section if (!is.null(width)) { if (length(width) != n | any(is.na(width)) | any(width <= 0)) stop("invalid boxplot widths") width <- 0.8 * width/max(width) } else if (varwidth) { width <- 0.8 * sqrt(z[[i]]$n/nmax) } if (n == 1) width <- 0.4 else width <- rep(0.8, n) The effect of either an explicit "width" argument or a TRUE "varwidth" argument is being overridden in the next statement. Since the default value of "width" is NULL, I think this can be repaired by replacing the last three lines shown there by if (is.null(width)) { if (n == 1) width <- 0.4 else width <- rep(0.8, n) } P.S. For those of you who don't know who Ron Ziegler was (which means you are too young to remember Watergate), he was Richard Nixon's press secretary. When it was shown that a statement he had made to the press the previous day was an outright lie he responded, "That statement is no longer operative.". =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=- r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html Send "info", "help", or "[un]subscribe" (in the "body", not the subject !) To: r-devel-request@stat.math.ethz.ch =-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-=-