The dsread output is little-endian, as that's the native format for 
floats on the Wintel platform.  The byte order should stay the same if 
converting directly to a float, using a data structure like (C/C++):

union {
     char bytes[8];
     double value;
}

If reading the values with a SAS HEX informat, the bytes will need to be 
reversed.  It's obviously trivial for me to add an endian-ness option, 
I'll do that later....

Chris.

On 02/03/2010 02:06, Roger DeAngelis(xlr82sas) wrote:
> Hi,
>
>    It looks like we may need to swap bytes(little endian to big endian). I
> will look into it tonight.
>
>    As a side note, SAS reserves 28 floats for missing values. It should be
> easy to convert these to NaN on input to R.
>
> You can test this in SAS by converting the 16 char floats to ieee8. in SAS
> and doing a put. The result will be A, B...Z, . and _.
>
> SAS code that produced the listing is below.
>
> Here are the floats that map to the 28 missing values in SAS
>
> A  FFFFFD0000000000
> B  FFFFFC0000000000
> C  FFFFFB0000000000
> D  FFFFFA0000000000
> E  FFFFF90000000000
> F  FFFFF80000000000
> G  FFFFF70000000000
> H  FFFFF60000000000
> I  FFFFF50000000000
> J  FFFFF40000000000
> K  FFFFF30000000000
> L  FFFFF20000000000
> M  FFFFF10000000000
> N  FFFFF00000000000
> O  FFFFEF0000000000
> P  FFFFEE0000000000
> Q  FFFFED0000000000
> R  FFFFEC0000000000
> S  FFFFEB0000000000
> T  FFFFEA0000000000
> U  FFFFE90000000000
> V  FFFFE80000000000
> W  FFFFE70000000000
> X  FFFFE60000000000
> Y  FFFFE50000000000
> Z  FFFFE40000000000
> _  FFFFFF0000000000
> .  FFFFFE0000000000
>
> data mis;
> retain A .A B .B C .C D .D E .E F .F G .G H .H I .I J .J K .K L .L M .M
>         N .N O .O P .P Q .Q R .R S .S T .T U .U V .V W .W X .X Y .Y Z .Z
>        _ ._ DOT .;
> array mis[28] A B C D E F G H I J K L M N O P Q R S T U V W X Y Z _ DOT;
>    do idx=1 to 28;
>       hex=put(mis[idx],ieee8.);
>       xeh=put(hex,hex16.);
>       put @1 mis[idx] @6 xeh;
>    end;
>
>
>    

        [[alternative HTML version deleted]]

______________________________________________
R-help@r-project.org mailing list
https://stat.ethz.ch/mailman/listinfo/r-help
PLEASE do read the posting guide http://www.R-project.org/posting-guide.html
and provide commented, minimal, self-contained, reproducible code.

Reply via email to