[Rd] symbols: xlim and ylim cannot be specified (PR#639)

[maechler@stat.math.ethz.ch]

>>>>> "Uwe" == Uwe Ligges <ligges@statistik.uni-dortmund.de> writes:

    Uwe> stahel@stat.math.ethz.ch wrote:
    >> 
    >> symbols(iris[,1],iris[,2],rectangles=as.matrix(iris[,3:4]),inches=0.3,
    >>         ylim=c(1,5))
    >> Error in plot.default(NA, NA, type = "n", ylim = ylim, xlim = xlim, xlab = xlab,  :
    >> formal argument "ylim" matched by multiple actual arguments
    >> 
    >> Since plot is called inside symbols, with specified xlim and ylim,
    >> any specification through  ,...  hurts this call.
yes,
and Uwe's fix cures this problem.
    [Uwe, you should have CC'ed  R-bugs, not (only) R-devel .. that's why I
     copy it again now].

and I have committed it for the release patches (!! i.e. "1.1.1 patched")

    >> I am pretty sure that you won't like to hear the suggestion that
    >> this error message
    >>    formal argument "anything" matched by multiple actual arguments
    >>    should only be a warning.

    >> I would find this more convenient since foreseeing these collisions
    >> and programming such arguments as explicit arguments of the "top"
    >> function is a nuisance when writing functions just for one project.

Maybe this should be discussed a bit:
The error can only be made into a warning when it makes sense to assume
that it is always okay, to take the *LAST* actual argument as the 
desired one.   This is the case probably whenever "..." comes at the end of the
formal argument list -- which it is in about 95% of cases...

Hmm, what do others think?

    >> Ceterum censeo: Thanks for your efforts for developing such a useful
    >> environment.
Merci!

    Uwe> Suggestion for a fix:

      function (x, y, circles, squares, rectangles, stars, thermometers, 
	  boxplots, inches = TRUE, add = FALSE, fg = 1, bg = NA, xlab = "", 
	  ylab = "", xlim = NULL, ylim = NULL, ...) 
		     ^^^^^^^^^^^^^^^^^^^^^^^^^
      {
      ## [...] snip ##
	if (!add) { 
      ### check, wheter xlim and ylim are already specified:
	      if(is.null(xlim)) {
		  xlim <- range(x, na.rm = TRUE)
		  xlim <- xlim + c(-1, 1) * (0.1 * diff(xlim))
	      }
	      if(is.null(ylim)) {
		  ylim <- range(y, na.rm = TRUE)
		  ylim <- ylim + c(-1, 1) * (0.1 * diff(ylim))
	      }
	      plot(NA, NA, type = "n", ylim = ylim, xlim = xlim, xlab = xlab, 
		  ylab = ylab, ...)
	  }
	  .Internal(symbols(x, y, type, data, inches, bg, fg, ...))
      }

    Uwe> Regards,
    Uwe> Uwe Ligges

Martin Maechler

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._