From: William Dunlap <wdunlap_at_tibco.com>

Date: Thu, 05 Feb 2009 13:38:01 -0800

R-devel_at_r-project.org mailing list

https://stat.ethz.ch/mailman/listinfo/r-devel Received on Thu 05 Feb 2009 - 21:40:05 GMT

Date: Thu, 05 Feb 2009 13:38:01 -0800

> -----Original Message-----

*> From: r-devel-bounces_at_r-project.org
**> [mailto:r-devel-bounces_at_r-project.org] On Behalf Of Greg Snow
**> Sent: Thursday, February 05, 2009 1:15 PM
**> To: marc_schwartz_at_comcast.net; ted.harding_at_manchester.ac.uk
**> Cc: R-Devel
**> Subject: Re: [Rd] "open-ended" plot limits?
**>
**> I use range( 0, y ) rather than c(0, max(y)), that way if
**> there are any y values less than 0, the limits still include
**> them (and it is slightly shorter :-).
*

To mimic what plot does by default you must ignore the NA's and Inf's in y with something like

range(0,y[is.finite(y)])

It might be nice to have an na.rm-like argument for ignoring
the Inf's - it gets tedious to write

range(0, y1[!is.finite(y1)], y2[!is.finite(y2)], ...) Also, when you get into really long vectors the explicit subscripting can run you out of memory.

Bill Dunlap

TIBCO Software Inc - Spotfire Division

wdunlap tibco.com

*>
*

> This also extends to cases where you may know that you will

*> be adding additional data using points or lines, so you can
**> do ylim=range(0, y1, y2, y3) and it will give enough room to
**> add the other y variables in latter.
**>
**> Hope this helps,
**>
**> --
**> Gregory (Greg) L. Snow Ph.D.
**> Statistical Data Center
**> Intermountain Healthcare
**> greg.snow_at_imail.org
**> 801.408.8111
**>
**>
**> > -----Original Message-----
**> > From: r-devel-bounces_at_r-project.org [mailto:r-devel-bounces_at_r-
**> > project.org] On Behalf Of Marc Schwartz
**> > Sent: Thursday, February 05, 2009 1:57 PM
**> > To: ted.harding_at_manchester.ac.uk
**> > Cc: R-Devel
**> > Subject: Re: [Rd] "open-ended" plot limits?
**> >
**> > on 02/05/2009 02:48 PM (Ted Harding) wrote:
**> > > Hi Folks,
**> > > Maybe I've missed it already being available somehow,
**> > > but if the following isn't available I'd like to suggest it.
**> > >
**> > > If you're happy to let plot() choose its own limits,
**> > > then of course plot(x,y) will do it.
**> > >
**> > > If you know what limits you want, then
**> > > plot(x,y,xlim=c(x0,x1),ylim(y0,y1)
**> > > will do it.
**> > >
**> > > But sometimes one would like to
**> > > a) make sure that (e.g.) the y-axis has a lower limit (say) 0
**> > > b) let plot() choose the upper limit.
**> > >
**> > > In that case, something like
**> > >
**> > > plot(x,y,ylim=c(0,NA))
**> > >
**> > > would be a natural way of specifying it. But of course that
**> > > does not work.
**> > >
**> > > I would like to suggest that this possibility should be available.
**> > > What do people think?
**> > >
**> > > Best wishes,
**> > > Ted.
**> >
**> > Ted,
**> >
**> > Unless I am mistaken in what you are looking for:
**> >
**> > plot(x, y, ylim = c(0, max(y)))
**> >
**> > would seem do what you want. If otherwise unspecified, plot() uses
**> > range(y) to define 'ylim'.
**> >
**> > HTH,
**> >
**> > Marc Schwartz
**> >
**> > ______________________________________________
**> > R-devel_at_r-project.org mailing list
**> > https://stat.ethz.ch/mailman/listinfo/r-devel
**>
**> ______________________________________________
**> R-devel_at_r-project.org mailing list
**> https://stat.ethz.ch/mailman/listinfo/r-devel
**>
*

R-devel_at_r-project.org mailing list

https://stat.ethz.ch/mailman/listinfo/r-devel Received on Thu 05 Feb 2009 - 21:40:05 GMT

Archive maintained by Robert King, hosted by
the discipline of
statistics at the
University of Newcastle,
Australia.

Archive generated by hypermail 2.2.0, at Fri 06 Feb 2009 - 00:30:20 GMT.

*
Mailing list information is available at https://stat.ethz.ch/mailman/listinfo/r-devel.
Please read the posting
guide before posting to the list.
*