[R] lm.influence under R2.6.1

[Marc Schwartz]

On Thu, 2007-12-06 at 14:00 -0500, Gerard Tromp wrote:
> Greetings!
> 
> Recently when I tried to use lm.influence I get the following error:
> Error in .Fortran("lminfl", model$qr$qr, n, n, k, as.integer(do.coef),  :
>   Fortran symbol name "lminfl" not in DLL for package "base"
> 
> This occurs on both Linux and Windows platforms (details below).

I presume that the Linux message does not refer to DLLs? :-)

> Searching the mail lists and other sources indicates that the fortran code
> for lminfluence (lminfl.f) was moved in the source code tree suggesting it
> should be found in package stats. Not clear to me why the "hardcoded" path
> is package base.
> The offending line in the R code for lm.influence is:
> res <- .Fortran("lminfl", model$qr$qr, n, n, k, as.integer(do.coef),
> 
> There were no other messages from people experiencing errors with
> lm.influence on the mailing list.
> 
> Can someone else confirm that they are experiencing the same problem?
> 
> Also, as a workaround, is there a way of forcing R to look in stats for the
> fortran symbol (similar to using the double and triple colon for addressing
> specific namespaces)?

<snip>

I can run example(lm.influence) and example(influence.measures) without
issue on:

  R version 2.6.1 Patched (2007-11-26 r43541)

on an F8 Linux system.

I don't see anything immediately obvious in the NEWS file for 2.6.1
patched to suggest that something has changed here.

Can you provide a reproducible example here and/or try to upgrade to
patched?

I would also try to run:

  R --vanilla

from the command line to provide a clean environment to be sure that
none of the other packages that you are loading are causing conflicts.

HTH,

Marc Schwartz