[R] fixed variance of lmer object

Martin Maechler maechler at stat.math.ethz.ch
Sat Mar 4 10:38:42 CET 2006

>>>>> "BDR" == Prof Brian Ripley <ripley at stats.ox.ac.uk>
>>>>>     on Sat, 4 Mar 2006 08:28:06 +0000 (GMT) writes:

    BDR> On Sat, 4 Mar 2006, Johannes Hüsing wrote:
    >> Martin Maechler <maechler at stat.math.ethz.ch> [Fri, Mar
    >> 03, 2006 at 05:38:52PM CET]:
    >>>>>>>> "Johannes" == Johannes Hüsing <hannes at ruhrau.de>
    >>>>>>>> on Wed, 1 Mar 2006 21:01:39 +0100 writes:
    Johannes> Dear expRts, could anybody nudge me toward some
    Johannes> documentation about how to extract the
    Johannes> variance-covariance matrix of the fixed parameters
    Johannes> from an lmer object?
    >>>  vcov( . )
    >>  thank you! I just found it out myself by looking at
    >> https://svn.r-project.org/R-packages/trunk/Matrix/R/lmer.R.
    >>>  as ``for all good objects '' :-)
    >>  Which brings me back to the core of my question: Where
    >> was I supposed to find this piece of information
    >> myself. The logical place, in my opinion, would be
    >> Section 11.3 of "An Introduction to R". Am I missing
    >> something here? If desired, I can write a patch for
    >> R-intro.texi.

    BDR> In Venables & Ripley (1994-2002), whose invention this
    BDR> was.


But since you've offered it, yes, I'd be glad for a patch to
R-intro.texi (that file and none else!!) which talks about
vcov() --- and other patches too which would talk about features
in R (and S), e.g.,  logLik(), that were not present (or not
widely known) when the first version of the introduction was
written, and hence have been overlooked there.

Martin Maechler, ETH Zurich

More information about the R-help mailing list