[Rd] Re: Finding code (was: [R] A request)

[E.L. Willighagen]

On Tuesday 02 April 2002 12:25, ripley@stats.ox.ac.uk wrote:
> On Tue, 2 Apr 2002, E.L. Willighagen wrote:
> > On Tuesday 02 April 2002 10:05, Uwe Ligges wrote:
> > > "E.L. Willighagen" wrote:
> > > R provides the tools for easily documenting and packaging functions.
> >
> > Interesting point, because I am working on a library with the same
> > issues: how to make stuff easy to find... R offers keywords, aliases and
> > full-text search. But that is all, right?
> >
> > And these only work for installed packages (i.e. those within the search
> > path...) And (just tested), full-text search does not include searching
> > references or descriptions, right (i.e. \description{} and \reference{})?
> >
> > For example,
> >
> > > help.search("Dongarra")
> >
> > No help files found with alias or title matching `Dongarra'
> >
> > Dongarra is author from article in svd().
> >
> > Did I miss some crucial functionality of help() and help.search()? Why
> > can only keywords and aliasses be searched?
> >
> > To return to Indrajit's question: there must be several other packages
> > with which one can do statistical quality control, or at least with some
> > minor extensions... But he seems not be able to find them. Ofcourse, that
> > could be
>
> Must there?  Why do you assert that?  As far as I am aware, there are
> no such packages, and I have tested/ported all those on CRAN.

Aren't there methods to calculated means and variances? Did I misunderstood 
that those can be used in Shewart charts? Ofcourse, things might need 
extensions, like a chart plotter...
I have not tested all CRAN modules, but I would have a too positive 
impression about R... 

> > imcompetence, but even then... should R be only used by experts, and not
> > by novice? In other words, is there a reason why R does not have more
> > extensive search methods that can be used by novice to get to know R in
> > such depth that the do not have to bug the R-help@ mailling list?
>
> Because R is a volunteer project, and no one has (AFAIK) contributed such
> methods. It seems that for them to be really useful approximate matching
> is needed.
>
> I do think you have a fundamental misconception of a project such as R.
> The developers wrote R for their own purposes, which include a vehicle for
> teaching statistics and to promulgate their favourite statistical methods.
> They made what they did freely available.  What we do not do is provide
> facilities on demand, from you or anyone else.

That is why I cross-posted this message to r-devel. Is there interest in such 
functionality? At least I am interested, even in coding this extra 
functionality, but before doing so, I thought it would be nice to discuss 
this on the developers list...

> Please do not reply to this message, publicly or otherwise.

Not that I would like to have the last word, or want to offend you, but I 
already wrote my answer before reading this line...

Egon
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel 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-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._