R devel - Jul 2009 - Incorrect comment about ISNA(x) in Arith.h (PR#13826)

R-2.9.0/include/R_ext/Arith.h has:

int R_IsNA(double);		/* True for R's NA only */
int R_IsNaN(double);		/* True for special NaN, *not* for NA */
int R_finite(double);		/* True if none of NA, NaN, +/-Inf */

#define ISNA(x)	       R_IsNA(x)
/* True for *both* NA and NaN.

The first and last lines are contradictory - if R_IsNA is true only
for NA, not NaN, then ISNA should be the same.

Venables & Ripley S Programming p 137 indicates that
"ISNA does not detect NaN s"

Here's a little .c file and R transcript that shows that V&R are
correct,
   ISNA(x) is not True for NaN
So the last comment should read
/* True for NA, not NaN



#include "R.h"
void mysum(double *x)
{
  double sum = 0.0;
  sum = x[0] + x[1] + x[2];
  if (ISNA(sum))
    PROBLEM "sum is NA" WARNING(NULL_ENTRY);
  if (ISNAN(sum))
    PROBLEM "sum is NAN" WARNING(NULL_ENTRY);
  if (!(R_FINITE(sum)))
    PROBLEM "sum is not finite" WARNING(NULL_ENTRY);
  return;
}

> dyn.load("~/R/work/bug.so")
> .C("mysum", c(NaN, 0.0, 0.0), NAOK = TRUE, DUP = FALSE)
[[1]]
[1] NaN   0   0

Warning messages:
1: sum is NAN 
2: sum is not finite 



--please do not edit the information below--

Version:
 platform = x86_64-unknown-linux-gnu
 arch = x86_64
 os = linux-gnu
 system = x86_64, linux-gnu
 status = 
 major = 2
 minor = 9.0
 year = 2009
 month = 04
 day = 17
 svn rev = 48333
 language = R
 version.string = R version 2.9.0 (2009-04-17)

Locale:
LC_CTYPE=en_US;LC_NUMERIC=C;LC_TIME=C;LC_COLLATE=C;LC_MONETARY=C;LC_MESSAGES=en_US;LC_PAPER=en_US;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_US;LC_IDENTIFICATION=C

Search Path:
 .GlobalEnv, package:Rfixes, package:Rcode, package:aggregate,
package:splus2Rgoogle, package:stats, package:graphics, package:grDevices,
package:utils, package:datasets, package:methods, Autoloads, package:base

>>>>> Tim Hesterberg <rocket at google.com>
>>>>>     on Tue, 14 Jul 2009 06:15:17 +0200 (CEST) writes:
    > R-2.9.0/include/R_ext/Arith.h has:
    > int R_IsNA(double);		/* True for R's NA only */
    > int R_IsNaN(double);		/* True for special NaN, *not* for NA */
    > int R_finite(double);		/* True if none of NA, NaN, +/-Inf */

    > #define ISNA(x)	       R_IsNA(x)
    > /* True for *both* NA and NaN.

Hi Tim, thanks for that, but you have cited things out of
context. The last line in the above is the beginning of a longer
comment which introduces  '#define ISNAN(x)'
and is all about ISNAN(.) and hence is correct.

So really, there's no bug, not even in comments of the sources
{{which are *not* part of the API, so very pedantically cannot
  have a bug... but of course, we'd still be glad for reports
  like yours particularly if they are about an error ..}}

In order to ensure that such a confusion is less likely to
happen again, I've very slightly changed that part now
[svn rev 48943].

Regards,
Martin Maechler, ETH Zurich




    > The first and last lines are contradictory - if R_IsNA is true only
    > for NA, not NaN, then ISNA should be the same.

    > Venables & Ripley S Programming p 137 indicates that
    > "ISNA does not detect NaN s"

    > Here's a little .c file and R transcript that shows that V&R
are correct,
    > ISNA(x) is not True for NaN
    > So the last comment should read
    > /* True for NA, not NaN



    > #include "R.h"
    > void mysum(double *x)
    > {
    > double sum = 0.0;
    > sum = x[0] + x[1] + x[2];
    > if (ISNA(sum))
    > PROBLEM "sum is NA" WARNING(NULL_ENTRY);
    > if (ISNAN(sum))
    > PROBLEM "sum is NAN" WARNING(NULL_ENTRY);
    > if (!(R_FINITE(sum)))
    > PROBLEM "sum is not finite" WARNING(NULL_ENTRY);
    > return;
    > }


    >> dyn.load("~/R/work/bug.so")
    >> .C("mysum", c(NaN, 0.0, 0.0), NAOK = TRUE, DUP = FALSE)
    > [[1]]
    > [1] NaN   0   0

    > Warning messages:
    > 1: sum is NAN 
    > 2: sum is not finite 



    > --please do not edit the information below--

    > Version:
    > platform = x86_64-unknown-linux-gnu
    > arch = x86_64
    > os = linux-gnu
    > system = x86_64, linux-gnu
    > status = 
    > major = 2
    > minor = 9.0
    > year = 2009
    > month = 04
    > day = 17
    > svn rev = 48333
    > language = R
    > version.string = R version 2.9.0 (2009-04-17)

    > Locale:
    >
LC_CTYPE=en_US;LC_NUMERIC=C;LC_TIME=C;LC_COLLATE=C;LC_MONETARY=C;LC_MESSAGES=en_US;LC_PAPER=en_US;LC_NAME=C;LC_ADDRESS=C;LC_TELEPHONE=C;LC_MEASUREMENT=en_US;LC_IDENTIFICATION=C

    > Search Path:
    > .GlobalEnv, package:Rfixes, package:Rcode, package:aggregate,
package:splus2Rgoogle, package:stats, package:graphics, package:grDevices,
package:utils, package:datasets, package:methods, Autoloads, package:base

    > ______________________________________________
    > R-devel at r-project.org mailing list
    > https://stat.ethz.ch/mailman/listinfo/r-devel