Re: R-beta: Teach material for R

Thomas Lumley (thomas@biostat.washington.edu)
Thu, 26 Mar 1998 09:12:49 -0800 (PST)


Date: Thu, 26 Mar 1998 09:12:49 -0800 (PST)
From: Thomas Lumley <thomas@biostat.washington.edu>
To: r-help@stat.math.ethz.ch
Subject: Re: R-beta: Teach material for R
In-Reply-To: <199803261543.QAA06470@sophie.ethz.ch>

On Thu, 26 Mar 1998, Martin Maechler wrote:

> >>>>> "Kenneth" == Kenneth Nordstrom <nordstro@noppa.helsinki.fi> writes:
> 
>     Kenneth> Howdy!  I'm a newbie in R, but have nevertheless gone out on a
>     Kenneth> limb and used R this spring for teaching our stats students
>     Kenneth> "modern data-analysis".  I've been using RNotes, but to get
>     Kenneth> some more (statistical) meat on the bones, I've complemented
>     Kenneth> it with material from Venables-Ripley. Most of the V-R stuff
>     Kenneth> (that I've tried) seems to work out of the box.
> 
>     Kenneth> Q-1: Is this generally the case (except for obvious
>     Kenneth> differences with S+ like 3-dim graphics)? If yes, why is it
>     Kenneth> necessary to port MASS if you can source the files on their
>     Kenneth> diskette? Anyone tried this, or used other material for
>     Kenneth> teaching data-analysis with R a la V-R?
> 
> You must have been that you didn't have problems with "MASS".
> Part of it certainly is because V&R use a very good S programming style.

Some other things involved in porting even when the S code is ok:
* C, and even more so FORTRAN, tends to need minor changes since R only
allows double precision (FORTRAN programmers often use REAL instead of
DOUBLE PRECISION, which is quite a bit faster on some computers (though 
not x86s))
* The help pages need to be converted to R help format (we have a Perl
script for this)
* The directory structure for an R package is different.

These are all routine (unlike the occasional changes to the interpreted
code), but still need to be done to make a package (as distinct from a
code fragment) useful.

> 
>     Kenneth> Q-2: How does R compare with XploRe? The latter seems to have
>     Kenneth> the advantage that more macros/functions are already written,
>     Kenneth> it has a nice graphical user interface (a consideration when
>     Kenneth> teaching) as well as 3-dim graphics (also dynamical). Has
>     Kenneth> anyone compared these in terms of capabilities as programming
>     Kenneth> languages? (Note that you can get XploRe for LINUX for free
>     Kenneth> and without expiry date!)
> 

One advantage of R for teaching purposes is its compatibility with
S(-PLUS), which is becoming very popular out there in the 'Real World',
particularly in biostatistics. 


Thomas Lumley
-----------------------------------------------------+------
Biostatistics		: "Never attribute to malice what  :
Uni of Washington	:  can be adequately explained by  :
Box 357232		:  incompetence" - Hanlon's Razor  :
Seattle WA 98195-7232	:				   :
------------------------------------------------------------

-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-help 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-help-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._