[Rd] Legend frame error (PR#9236)

Peter Dalgaard p.dalgaard at biostat.ku.dk
Mon Sep 18 22:43:40 CEST 2006

tgibson at augustcouncil.com writes:

> Hello,
> The frame drawn around the legend appears to miscalculate when the
> x-axis decreases rather than increases.  Pasted below is a simple
> testcase.  When the x-axis decreases, the width of the legend frame
> appears to be calculated based on the length of the first item, rather
> than the maximum length of all items.

More likely, it is calculating the maximum of the endpoints, which
will of course be the one furthest to the left if the x axis is
reversed. (As reordering the legends would have shown you.)

The construction

    if (is.null(text.width))
        text.width <- max(strwidth(legend, units = "user", cex = cex))
    else if (!is.numeric(text.width) || text.width < 0)
        stop("'text.width' must be numeric, >= 0")

seems to be the culprit. max() should be min() in the reversed case.

And if I'm not mistaken, there's an additional issue: If the x axis is
reversed, an explicit text.width setting should in fact be negative.

   O__  ---- Peter Dalgaard             Øster Farimagsgade 5, Entr.B
  c/ /'_ --- Dept. of Biostatistics     PO Box 2099, 1014 Cph. K
 (*) \(*) -- University of Copenhagen   Denmark          Ph:  (+45) 35327918
~~~~~~~~~~ - (p.dalgaard at biostat.ku.dk)                  FAX: (+45) 35327907

More information about the R-devel mailing list