R devel - Aug 2001 - Reading Inf and NaN values under windows (PR#1072)

Under windows, R supports IEEE floating point arithmetic, but doesn't
allow conversion of Inf and NaN from character to numeric.

R> is.nan(NaN)
[1] TRUE
R> as.numeric(as.character(NaN))
[1] NA
Warning message: 
NAs introduced by coercion 
R> is.infinite(Inf)
[1] TRUE
R> as.numeric(as.character(Inf))
[1] NA
Warning message: 
NAs introduced by coercion 

whereas under Linux

R> as.numeric(as.character(NaN))
[1] NaN
R> as.numeric(as.character(Inf))
[1] Inf

S-PLUS 3.3 for Windows gives

S-PLUS> as.numeric(as.character(Inf))
[1] Inf
S-PLUS> as.numeric(as.character(NaN))
[1] NA
Warning messages:
  1 missing value  generated coercing from character to numeric in: 
        as.numeric(as.character(NaN))

I first came across this problem when trying to read in a dataset with
Inf values using read.table().  Under Windows, the variable with the
infinite value is converted to a factor. The scan() function also refuses
to read in Inf and NaN as numeric values under Windows.

Martyn

--please do not edit the information below--

Version:
 platform = i386-pc-mingw32
 arch = x86
 os = Win32
 system = x86, Win32
 status = 
 major = 1
 minor = 3.0
 year = 2001
 month = 06
 day = 22
 language = R

Windows NT 4.0 (build 1381) Service Pack 6

Search Path:
 .GlobalEnv, package:ctest, Autoloads, package:base

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

On Fri, 24 Aug 2001 plummer@iarc.fr wrote:
> Under windows, R supports IEEE floating point arithmetic, but doesn't
> allow conversion of Inf and NaN from character to numeric.
>
> R> is.nan(NaN)
> [1] TRUE
> R> as.numeric(as.character(NaN))
> [1] NA
> Warning message:
> NAs introduced by coercion
> R> is.infinite(Inf)
> [1] TRUE
> R> as.numeric(as.character(Inf))
> [1] NA
> Warning message:
> NAs introduced by coercion
>
> whereas under Linux
>
> R> as.numeric(as.character(NaN))
> [1] NaN
> R> as.numeric(as.character(Inf))
> [1] Inf
>
> S-PLUS 3.3 for Windows gives
>
> S-PLUS> as.numeric(as.character(Inf))
> [1] Inf
> S-PLUS> as.numeric(as.character(NaN))
> [1] NA
> Warning messages:
>   1 missing value  generated coercing from character to numeric in:
>         as.numeric(as.character(NaN))
>
> I first came across this problem when trying to read in a dataset with
> Inf values using read.table().  Under Windows, the variable with the
> infinite value is converted to a factor. The scan() function also refuses
> to read in Inf and NaN as numeric values under Windows.
It's the same cause: strtod is used.  That's not required to support
IEC 60559 (to be correct) specials, but glibc happens to.

We have R_strtod: we should make proper use of it.  I'll do that for
1.3.1.

Brian
>
> Martyn
>
> --please do not edit the information below--
>
> Version:
>  platform = i386-pc-mingw32
>  arch = x86
>  os = Win32
>  system = x86, Win32
>  status >  major = 1
>  minor = 3.0
>  year = 2001
>  month = 06
>  day = 22
>  language = R
>
> Windows NT 4.0 (build 1381) Service Pack 6
>
> Search Path:
>  .GlobalEnv, package:ctest, Autoloads, package:base
>
>
-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-.-
> 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
>
_._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._._
>
-- 
Brian D. Ripley,                  ripley@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 272860 (secr)
Oxford OX1 3TG, UK                Fax:  +44 1865 272595

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