Take a closer look: the differences in output are not just in trailing
digits.
We solve the latter in R itself mainly by use of e.g. options(digits=5)
in the relevant \examples{} sections.
However, we also try to remove the numerical instability in the
algorithms that leads to this. Perhaps
R gurus:
I'm trying to get another round of rconifers out and I need some advice/help
crushing differences in the examples test.
I'm trying to make sure the max sdi values are being respected.
I've added a tests/rconifers-Ex.Rout.save (from windows i386-pc-mingw32) and
when I ran R CMD check (
