[R] Some more help needed...

Patrick Connolly p.connolly at hortresearch.co.nz
Mon Jan 19 20:56:29 CET 2004

On Mon, 19-Jan-2004 at 06:36PM +0100, Peter Dalgaard wrote:

|> "v.demart at libero.it" <v.demart at libero.it> writes:
|> > As an absolute beginner, still reading "Modern Applied Statistics
|> > with S" and exercising with its examples, I'm frequently stopped
|> > by what it looks to be R poor help system (or is it my gigantic
|> > ignorance?). I mean that using help many arguments of a command
|> > seems to be given for granted like for instance:
|> > ...............................
|> > ?lines
|> >  lines(x, ...)
|> > 
|> >      ## Default S3 method:
|> >      lines(x, y = NULL, type = "l", col = par("col"),
|> >            lty = par("lty"), ...)
|> > 
|> > Arguments:
|> > 
|> >     x, y: coordinate vectors of points to join.
|> > 
|> >     type: character indicating the type of plotting; actually any of
|> >           the 'type's as in 'plot'.
|> > 
|> >      col: color to use. This can be vector of length greater than one,
|> >           but only the first value will be used.
|> > 
|> >      lty: line type to use.
|> > 
|> >      ...: Further graphical parameters (see 'par') may also be supplied
|> >           as arguments, particularly, line type, 'lty' and line width,
|> >           'lwd'.
|> > ................................
|> > 
|> > How could I quickly know during an R-session what values should be "col" set to have red, how could I set "lty" etc.?
|> Well, you might take a hint and look at ?par, in which this is in fact 
|> explained. The above text is not saying that very explicitly, I agree. 

... that hint being given a little further down the help for lines,
where it says:

See Also:

     'points', 'plot', and the underlying "primitive" 'plot.xy'.

     'par' for how to specify colors.

I can't think of a way that makes it easier to get to than that.

|> However, it could be a good idea if we found a nice way of
|> integrating this sort of tabular material in the help system. The
|> case that really annoys me is that to get at the vaæues for 'pch',
|> you need to run example(points), which is both nonobvious and
|> disruptive if you are in the middle of constructing a complex plot
|> command.

Even in that case, it's not a big deal to run a separate R session in
another workspace (or even better, viewport in the pre-Gnome2 days)
where such ancillary tasks can run.  There are OS limitations, of

Patrick Connolly
Mt Albert
New Zealand 
Ph: +64-9 815 4200 x 7188
I have the world`s largest collection of seashells. I keep it on all
the beaches of the world ... Perhaps you`ve seen it.  ---Steven Wright 

More information about the R-help mailing list