On Monday 26 Apr 2010 14:05:16 Shlomi Fish wrote:
> Hi Rob,
>
> On Monday 26 Apr 2010 13:38:07 Rob Coops wrote:
> > Hi list,
> >
> > I am just wondering if I sumbled upon an error in perl an error in my
> > logic or somehtign else see the below perl one liners
> >
> > $ perl -e '$n = 0.945; $r = sprintf("%.2f", $n); print "$r\n";'
> > 0.94
> >
> > $ perl -e '$n = 0.9451; $r = sprintf("%.2f", $n); print "$r\n";'
> > 0.95
> >
> > $ perl -e '$n = 0.9450; $r = sprintf("%.2f", $n); print "$r\n";'
> > 0.94
> >
> > $ perl -e '$n = 0.94500; $r = sprintf("%.2f", $n); print "$r\n";'
> > 0.94
> >
> > $ perl -e '$n = 0.945001; $r = sprintf("%.2f", $n); print "$r\n";'
> > 0.95
> >
> > Now I expected perl to round the number: 0.945 to 0.95 and not to 0.94
> > like it is doing. I have no idea why it is doing this but I sure don't
> > like it. The problem is I have to do some VAT calculations on thousands
> > of invoices and invoice elements. I found that in rare cases I'm of by 1
> > cent or even 3 cents if there are lots of elements. Going over the logic
> > I ran into this problem.
> >
> > Cany anyone advise me how to deal with this as I need to find a way to
> > round correctly in all cases not just in cases where Perl decides it is a
> > good idea to do so. :-)
>
> The problem is likely caused by the fact that floating-point numbers in
> most modern computers are represented by using binary digits (and binary
> fractions) instead of decimal ones. So there may be rounding errors. As a
> result, it is not advisable to use the regular floating-point numbers for
> dealing with monetary values. Some of the options you may wish to consider
> instead are:
>
> 1. Use a rational library for computation such as Math::BigRat.
>
> 2. Use a binary-coded-decimal (BCD) system. Last time I investigated I
> couldn't find one for Perl, though , and I've misplaced the link to the
> downloadable software library at IBM for that. (Though Python has something
> like that in pure-Python)
>
> 3. Work always in the smallest unit - cents or whatever.
>
I eventually forgot to add the link to this article:
http://docs.sun.com/source/806-3568/ncg_goldberg.html
"What Every Computer Scientist Should Know About Floating-Point Arithmetic".
It's kinda long, but I mostly enjoyed it and found it enlightening.
Regards,
Shlomi Fish
--
-----------------------------------------------------------------
Shlomi Fish http://www.shlomifish.org/
My Aphorisms - http://www.shlomifish.org/humour.html
God considered inflicting XSLT as the tenth plague of Egypt, but then
decided against it because he thought it would be too evil.
Please reply to list if it's a mailing list post - http://shlom.in/reply .
--
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
http://learn.perl.org/
