Front page | perl.perl5.porters |
Postings from October 2003
On Tue, Oct 07, 2003 at 07:17:12PM -0400, Ed Allen Smith <easmith@beatrice.rutgers.edu> wrote:
> In message <20031007225502.GA732@efn.org> (on 7 October 2003 15:55:03
> -0700), sthoenna@efn.org (Yitzchak Scott-Thoennes) wrote:
> >On Tue, Oct 07, 2003 at 06:33:33PM -0400, Ed Allen Smith
> ><easmith@beatrice.rutgers.edu> wrote:
> >> if ($a == $b) {
> >> $x->[0] = $a;
> >> } elsif (abs(($a**$n->[0]) - $x->[0]) > abs(($b**$n->[0]) - $x->[0])) {
> >> $x->[0] = $b;
> >> } else {
> >> $x->[0] = $a;
> >> }
> >> return $x;
> >>
> >> In other words, if adding 0.5 (equivalently, subtracting, if the result
> >>is negative) will give a more arithmetically correct result after
> >>truncating, do it; otherwise, don't. (It's likely that the same sort of
> >>thing would be helpful for the _sqrt function.)
> >
> >That's basically just a form of rounding, only not based on just
> >rounding the result.
>
> Correct. It's choosing whether to truncate, or to round .5 and up upward.
>
> >It looks more to me as if the goal here is to
> >come up with the greatest result $a such that $a**$n->[0] <= $x->[0]
>
> I can see that argument; it depends on whether one wishes to stick with
> Calc.pm always truncating to int in _root or not.
It should be consistent, and the rest of it seems to want to truncate.
(This should be documented.)
> >(not sure if that "greatest" or "<=" is meant to apply to absolute
> >values or not).
> >
> >But your general method certainly works for that.
How about this:
--- lib/Math/BigInt/Calc.pm.orig 2003-10-08 15:31:26.583782400 -0700
+++ lib/Math/BigInt/Calc.pm 2003-10-08 15:43:00.251227200 -0700
@@ -1423,7 +1423,10 @@
else
{
# fit's into one Perl scalar, so result can be computed directly
- $x->[0] = int( $x->[0] ** (1 / $n->[0]) );
+ my $result = int( $x->[0] ** (1 / $n->[0]) + .5);
+ # see if we rounded up inappropriately
+ --$result if $result ** $n->[0] > $x->[0];
+ $x->[0] = $result;
}
return $x;
}
End of Patch.
By the way, just a bit lower down, I see a /0b1(0+)/ that looks as
if it should be /0b1(0+)\z/, but I'm not sure exactly what it is doing.