a strange logical bug (PR#162)

ripley@stats.ox.ac.uk writes:

On RH 5.2

       The isnan() function returns a non-zero value if value  is
       "not-a-number" (NaN), and 0 otherwise.

and that non-zero is not defined!

Uh-oh...

and isnan is used in Arith.c for ISNAN and that is used in coerce.c
for do_isna (why?)

is.na(0/0) # Splus compat.

So ISNAN needs to be redefined on Linux at least.  Using ISNA not ISNAN
solves this one for me.

But...(see above)

A better fix is 

#define ISNAN(x) (isnan(x)!=0)

Next problem: Messing with Arith.h and make'ing doesn't do anything
(dependency missing, obviously)

O__  ---- Peter Dalgaard             Blegdamsvej 3  
  c/ /'_ --- Dept. of Biostatistics     2200 Cph. N   
 (*) \(*) -- University of Copenhagen   Denmark      Ph: (+45) 35327918
~~~~~~~~~~ - (p.dalgaard@biostat.ku.dk)             FAX: (+45) 35327907
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._

as.numeric(is.na(c(NA,NA,1)))

[1] -1 -1  0

Seems to affect all numeric NA's:

A)))

[1] -1

This is on RedHat 5.2 (As Jim, I suspect)

On RH 5.2

I don't seem to have this problem on Debian 2.2 (with glibc2.1):

as.numeric(is.na(as.numeric(NA)))

[1] 1

as.numeric(is.na(c(NA,NA,1)))

[1] 1 1 0


Nicholas


-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
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
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._