On Thu, 15 Jan 2004, Prof Brian Ripley wrote:
> I would like to deprecate La.eigen. It is used in a few packages
> (ade4, fpc, gss, mvtnorm and smoothSurv on CRAN), but only in usages where
> replacing `La.eigen' by `eigen' would call exactly the same code.
>
> The reason for wanting to deprecate it is that little-used interfaces tend
> to get overlooked, e.g. PR#5406, a report on eigen, needed to be applied
> to La.eigen as well. We have also fixed NaNs issues in eigen but not
> La.eigen.
>
> Does anyone have code that uses the extra features of eigen?
Oops: of La.eigen, that is the `method' argument.
^^^
--
Brian D. Ripley, ripley@stats.ox.ac.uk
Professor of Applied Statistics, http://www.stats.ox.ac.uk/~ripley/
University of Oxford, Tel: +44 1865 272861 (self)
1 South Parks Road, +44 1865 272866 (PA)
Oxford OX1 3TG, UK Fax: +44 1865 272595