I've been finding that the "loose ends" in many of these older
codes cause more trouble
than it is worth in their use. When I encounter them, I've attempted to
re-program the
algorithm in R. A lot of the Fortran code is because of the software structure
the author
used and nothing to do with the job to be done.
If you can prepare an R function for this, you'd be doing the R community a
favour. You
may also find that a judicious combination of optimize() and grid search gets
the task
done satisfactorily.
Best,
JN
> Date: Sat, 2 Jul 2011 01:43:03 +0530
> From: Mohit Dayal <ken.sensei at gmail.com>
> To: R-devel at r-project.org
> Subject: [Rd] AS Algorithms
>
> I would like to use one of the AS Algorithms that used to be published in
> the journal Applied Statistics of the Royal Statistical Society (Series C).
> FORTRAN code based on these are available on the Statlib website at
>
>
http://lib.stat.cmu.edu/modules.php?op=modload&name=PostWrap&file=index&page=apstat/
[snip]>
> *BTW I am looking at AS 133 : Finding the global maximum or minimum of a
> function of 1 variable.
> *
>