On Tue, 31 Aug 2004 14:13:49 +0100, "Simon Fear" <[EMAIL PROTECTED]> wrote :
>I've searched on CRAN for axes, axis, and other terms >I've already forgotten, without (re)discovering the >reason for S using "non-joining" axes by default, instead >of box("l"). > >MASS points me towards Cleveland (1993) but I don't >have ready access to this any more. Could someone >give me a one-liner to justify this choice to a sceptic? > >It's something to do with not mis-interpreting the axes >intersection as (0,0), isn't it? I'm not sure I understand your question: the default in S is a full box unless you ask for no box (bty="n"), isn't it? The justification for this is "The four scale lines also provide a clearly defined region where our eyes can search for data. With just two, data can be camouflaged by virtue of where they lie" (p. 35). There's also advice to keep the data away from the axes and to put the ticks outside the box to avoid hiding extreme points. I can't spot a discussion in Cleveland of the reason the axes don't join when bty="n" is specified. Duncan Murdoch ______________________________________________ [EMAIL PROTECTED] mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html