IBM AIX problems

Ross Ihaka ihaka@stat.auckland.ac.nz
Tue, 9 Jun 1998 08:42:49 +1200 (NZST)


Hmmm.  I think all the code in R should use the FINITE macro rather than
the finite function.  On non IEEE platforms FINITE can be defined to check
for a special NA pattern.

It would be possible to rename the macro to something more obscure if
that would help.  E.g. R_FINITE.

Medium-term this whole area needs some work.  It may be that we will have
to have an R_specific implementation of functions like "finite"  and
"isnan".  This is because there are just too many platforms which seem
to screw them up.

The prize (as usual) goes to Windows, where (0.0 == NaN) is true.

	Ross
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
r-devel mailing list -- Read http://www.ci.tuwien.ac.at/~hornik/R/R-FAQ.html
Send "info", "help", or "[un]subscribe"
(in the "body", not the subject !)  To: r-devel-request@stat.math.ethz.ch
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._