Dear Peter and Martin,
On Thu, 17 Mar 2011 18:08:18 +0100
peter dalgaard <pda...@gmail.com> wrote:
>
> On Mar 17, 2011, at 16:14 , Martin Maechler wrote:
>
> >>>>>> peter dalgaard <pda...@gmail.com>
> >>>>>> on Thu, 17 Mar 2011 15:45:01 +0100 writes:
> >>
> >
> >> Back to the original question:
> >
> >> The current rstandard() code reads
> >
> > ## FIXME ! -- make sure we are following "the literature":
> > rstandard.glm <- function(model, infl = lm.influence(model, do.coef=FALSE),
> > ...)
> > {
> > res <- infl$wt.res # = "dev.res" really
> > res <- res / sqrt(summary(model)$dispersion * (1 - infl$hat))
> > res[is.infinite(res)] <- NaN
> > res
> > }
> >
> >> which is "svn blame" to ripley but that is due to the 2003
> >> code reorganization (except for the infinity check from
> >> 2005). So apparently, we have had that FIXME since
> >> forever... and finding its author appears to be awkward
> >> (Maechler, perhaps?).
> >
> > yes, almost surely
> >
> >> I did try Bretts code in lieu of the above (with a mod to
> >> handle $dispersion) and even switched the default to use
> >> the Pearson residuals. Make check-devel sailed straight
> >> through apart from the obvious code/doc mismatch, so we
> >> don't have any checks in place nor any examples using
> >> rstandard(). I rather strongly suspect that there aren't
> >> many user codes using it either.
> >
> >> It is quite tempting simply to commit the change (after
> >> updating the docs). One thing holding me back though: I
> >> don't know what "the literature" refers to.
> >
> > well, "the relevant publications on the topic" ...
> > and now define that (e.g. using the three 'References' on the
> > help page).
>
> I count 5 actually... IIRC, the first two do not deal with glm diagnostics.
> The last two are by Fox, and, presumably, he is around to chime in if he
> wants. The middle one, by Williams, does define both standardized Pearson and
> standardized deviance residuals.
Though I don't have it in front of me, I recall that my Applied Regression text
follows Williams and defines both standardized deviance and standardized
Pearson residuals. As well, there are new editions of both these sources:
Fox, J. (2008) Applied Regression Analysis and Generalized Linear Models,
Second Edition (Sage)
Fox, J. and S. Weisberg (2011) An R Companion to Applied Regression, Second
Edition (Sage)
I'd take Williams as the definitive reference. I'll send a follow-up message if
my memory proves faulty.
Best,
John
>
> Or did you mean the three on ?glm.summaries? I would assume Davison and Snell
> to be the operative one, but I don't have it to hand.
>
> Anyways, given that de default for residuals.glm is deviance residuals, I
> suppose that rstandard.glm should have the same default for consistency, and
> that is also the least disruptive variant. I see no reason not to make
> standardized Pearson residuals an option.
>
> > Really, that's what I think I meant when I (think I) wrote that FIXME.
> > The point then I think was that we had code "donations", and they
> > partly were clearly providing functionality that was (tested)
> > "correct" (according to e.g. McCoullagh & Nelder and probably
> > another one or two text books I would have consulted ... no
> > large Wikipedia back then),
> > but also provided things for which there was nothing in "the
> > literature", but as the author provided them with other good
> > code, we would have put it in, as well....
> > == my vague recollection from the past
> >
> > Martin
> >
> >> --
> >> Peter Dalgaard Center for Statistics, Copenhagen Business
> >> School Solbjerg Plads 3, 2000 Frederiksberg, Denmark
> >> Phone: (+45)38153501 Email: pd....@cbs.dk Priv:
> >> pda...@gmail.com
> >
> >> ______________________________________________
> >> R-devel@r-project.org mailing list
> >> https://stat.ethz.ch/mailman/listinfo/r-devel
>
> --
> Peter Dalgaard
> Center for Statistics, Copenhagen Business School
> Solbjerg Plads 3, 2000 Frederiksberg, Denmark
> Phone: (+45)38153501
> Email: pd....@cbs.dk Priv: pda...@gmail.com
>
> ______________________________________________
> R-devel@r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-devel
------------------------------------------------
John Fox
Sen. William McMaster Prof. of Social Statistics
Department of Sociology
McMaster University
Hamilton, Ontario, Canada
http://socserv.mcmaster.ca/jfox/
______________________________________________
R-devel@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-devel
