[R] Resolving conflicting function definitions {was "Error in rm.outlier"}

Marc Schwartz MSchwartz at mn.rr.com
Sat Apr 29 16:50:14 CEST 2006


On Sat, 2006-04-29 at 16:21 +0200, Martin Maechler wrote:
> >>>>> "Marc" == Marc Schwartz (via MN) <mschwartz at mn.rr.com>
> >>>>>     on Fri, 28 Apr 2006 14:38:53 -0500 writes:
> 
>     Marc> Sachin, I don't have a definitive thought, but some
>     Marc> possibilities might be a conflict somewhere in your
>     Marc> environment with a local function or with one in the
>     Marc> searchpath.
> 
>     Marc> Use ls() to review the current objects in your
>     Marc> environment to see if something looks suspicious. It
>     Marc> did not look like 'outliers' is using a namespace, so
>     Marc> a conflict of some nature is a little more possible
>     Marc> here.
> 
>     Marc> Also use searchpaths() to get a feel for where R is
>     Marc> searching for the function. See what is getting
>     Marc> searched "above" the outliers package in the search
>     Marc> order, which might provide a clue.
> 
>     Marc> Also, try to start R from the command line using 'R
>     Marc> --vanilla', which should give you a clean working
>     Marc> environment. Then use library(outliers) and your code
>     Marc> below to see if the same behavior is present. If so,
>     Marc> perhaps there was a corruption in the package
>     Marc> installation. If not, it would support some type of
>     Marc> conflict or perhaps a corruption in your default
>     Marc> working environment.
> 
> Another possibility -- the one I use first in such circumstances --
> is
> 	conflicts()
> 
> a good ol' function that seems not be as known as it maybe
> should.

Yes indeed Martin. I will acknowledge that I was not aware of it and
reviewing the *NEWS files, see that it was added by Prof. Ripley back in
R version 0.63, which was released in November of 1998.

I wonder if it would make sense to add a comment about it to the Posting
Guide under "Surprising behavior and bugs". I have cc'd Tony here for
his comment as well.

Best regards,

Marc




More information about the R-help mailing list