Because it is not the same code. In R, the set of known functions is
hard-coded. In S-PLUS, it is extensible, and was extended by MASS (now
incorporated into the standard distribution).
On Tue, 23 Mar 2004, Mervyn G Marasinghe wrote:
> Why is deriv3() functioning differently in R from that in Splus
> using library(MASS) ? For example
>
deriv3(~(t1*log(t2)+lgamma(t1)+(1-t1)*log(y)+y/t2),c("t1","t2"),function(y,t1,t2)NULL)
> complains of lgamma.
--
Brian D. Ripley, ripley at 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