[R] Newbie question about by() -- update

[david.beede@mail.doc.gov]

Prof. Ripley --
Thank you for your reply.

I'm not sure how to confirm your suggestion.  I found the line that you
cited in ppr.formula in the ascii file containing the modreg R code -- can
I simply change that line and rerun my example?  Or is there more to it
than that?




Prof Brian D Ripley <ripley at stats.ox.ac.uk>@auk.stats> on 03/20/2001
03:08:48 PM

Sent by:  <ripley at auk.stats>


To:   <david.beede at mail.doc.gov>
cc:   <r-help at stat.math.ethz.ch>

Subject:  Re: [R] Newbie question about by() -- update


On Tue, 20 Mar 2001 david.beede at mail.doc.gov wrote:

> Sorry about the lack of detail.  I am running R v.1.2.2.  I can recast my
> question
> (which I think I have partially answered) more succinctly as follows:
>
> 1.  This seems to work (note that group takes values 1,2,3,4, or 5):
>  my.newfun <- function(x) myfile <- lm(award ~ ilogemp + ilogage, x)
>  test.by <- by(wintemp, as.factor(wintemp$group), my.newfun)
>
> 2.  This does not work (leaving aside whether I am using ppr correctly or
not!):
> > my.pprfun <- function(x) mypprfile <- ppr(award ~ ilogemp + ilogage,
data = x, nterms = 5,
> +                max.terms = 10, optlevel = 3)
> >  test.by <- by(wintemp, as.factor(wintemp$group), my.pprfun)
> Error in model.frame.default(formula = award ~ ilogemp + ilogage, data =
x) :
>      Object "x" not found
>
> 3.  However, this does seem to work, but I don't know enough about R to
explain why:
> > my.pprfun <- function(x) mypprfile <- ppr(award ~ ilogemp + ilogage,
> +   data = eval.parent(substitute(x),3), nterms = 5, max.terms = 10,
optlevel = 3)
> >  test.by <- by(wintemp, as.factor(wintemp$group), my.pprfun)
>
> If I understand this (and p. 69 of S Programming by VR) correctly, it
seems as though
> I have to evaluate x three environments up the call sequence from ppr,
but I don't have
> to do so for lm.  I don't yet fully understand how the environments work,
and I'm afraid
> that only a little knowledge is dangerous!  Can anyone explain the
difference between
> lm and ppr in this context?

There's a bug in ppr.formula:

    m <- eval(m, sys.parent())

should be (in R)

    m <- eval(m, parent.frame())

I think that will fix it, but please confirm.

--
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 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595




-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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 at stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._