NetNews Offline 2/NetNews Offline Volume 2.iso/news/comp/std/c/406

Text File | 1996-08-06 | 1.8 KB | 51 lines

Path: sun001.spd.dsccc.com!spd!jmccarty
From: [e-mail] (Mike McCarty)
Newsgroups: comp.lang.c,comp.std.c,comp.lang.c++
Subject: Re: Floating Point arithmetic problem
Date: 15 Feb 1996 19:11:17 GMT
Organization: DSC Communications Corporation, Plano, Texas USA
Message-ID: <4g00gl$[e-mail]>
References: <[e-mail]>
NNTP-Posting-Host: aplo139.spd.dsccc.com

In article <[e-mail]>,
Kuntal Shah <[e-mail]> wrote:
)
)I am having a wierd problem with floating point arithmetic. Gurus on
)the net, please bail me out. I am working on the SUN 4.1.x platform.
)
)I have a "double" variable say d to which  I need to add certain float
)numbers  of  moderate magnitude (say   less than 10000). This addition
)occurs in a loop in my program which  get executed more than a million
)times depending on my testcase.

[stuff cut]

)This addition  usually results in  a  few insignificant digits getting
)accumulated in my double variable. For eg.
)
)  f = 12.99  is represented as 12.9900000000000002131628207
)  f = 11.111 is represented as 11.1110000000000006536993169
)
)and so on. I am interested in comparisons with 4 digits precision.
)
)Now coming  to   the  problem. The  insignificant  digits   due to the
)floating  point representation keep  accruing and  there comes a stage
)when the accrued value exceeds 0.0001 which  results in failure of the
)if condition in the  above block of code,  when ideally no such  thing
)should have occurred.

[stuff cut]

Every 1000 or 10000 or whatever, do a roundoff. Easiest way is:

    #define FACTOR    1e4

    d = floor(d*FACTOR+0.5)/FACTOR;

Mike
----
char *p="char *p=%c%s%c;main(){printf(p,34,p,34);}";main(){printf(p,34,p,34);}

I don't speak for DSC.         <- They make me say that.