[Rd] Possible tweak to R intro - was RE: [R] Subseting a data.frame -

Bert Gunter gunter.berton at gene.com
Mon Oct 21 16:47:54 CEST 2013

Thank you.

I stand corrected.


On Mon, Oct 21, 2013 at 1:17 AM, Prof Brian Ripley
<ripley at stats.ox.ac.uk> wrote:
> On 18/10/2013 15:12, S Ellison wrote:
>> Transferred from R-help:
>>>> From: S Ellison
>>>> Subsetting using subset() is perhaps the most natural way of
>>>> subsetting data frames; perhaps a line or two and an example could
>>>> usefully be included in the 'Working with data frames' section of the R
>>>> Intro?
>>> From: Bert Gunter [mailto:gunter.berton at gene.com]
>>> The R Intro Manual was largely or entirely the work of Bill Venables
>>> some years ago. So it is not really a part of R's maintained document
>>> system and has thus not been kept up to date with changes like the
>>> convenience function, subset(), which is basically a wrapper for "[]"
>>> .
>>> This is not to say that your suggestion is not worthwhile, only to
>>> explain why it probably won't be acted upon.
> No, this is deliberate and R-intro is kept up-to-date (although it remains
> an introduction, not a full manual).
> Some of us think convenience functions such as subset() and transform() were
> mistakes, not least as we see the problems they cause when people try to use
> them in functions and packages.  Sooner or later you will need to learn to
> use indexing, and knowing about two systems with different scoping rules is
> too confusing for quite a few R users.
>> It's trivial enough that I could offer a 3-line patch if someone has time
>> and inclination to add it...
>> S Ellison
> --
> Brian D. Ripley,                  ripley at stats.ox.ac.uk
> Professor of Applied Statistics,  http://www.stats.ox.ac.uk/~ripley/
> University of Oxford,             Tel:  +44 1865 272861 (self)
> 1 South Parks Road,                     +44 1865 272866 (PA)
> Oxford OX1 3TG, UK                Fax:  +44 1865 272595


Bert Gunter
Genentech Nonclinical Biostatistics

(650) 467-7374

More information about the R-devel mailing list