Bug #108643 for Math-BigInt: A rational power and a negative base produces a "uninitialized" warning

Tue Nov 10 02:02:25 2015 TRIZEN [...] cpan.org - Ticket created

Subject:

A rational power and a negative base produces a "uninitialized" warning

Considering the following code: === BEGIN-CODE === use Math::BigFloat; my $mone = Math::BigFloat->new(-1); my $three = Math::BigFloat->new(3); print $mone**($mone / $three); === END-CODE === Math::BigFloat produces the following warning: Use of uninitialized value in numeric le (<=) at /usr/share/perl5/core_perl/Math/BigFloat.pm line 2555.

Wed Nov 11 06:27:34 2015 peter.john.acklam [...] gmail.com - Correspondence added

Thanks for the report. I have checked this case against old versions of Math::BigFloat, and it turns out that this is a very old bug. Math::BigFloat does not support complex numbers, so raising a negative number to a non-integer power should result in a NaN.

Wed Nov 11 06:27:34 2015 The RT System itself - Status changed from 'new' to 'open'

Fri Dec 18 23:28:05 2015 TRIZEN [...] cpan.org - Correspondence added

On Wed Nov 11 06:27:34 2015, pjacklam wrote: Show quoted text

> Thanks for the report. I have checked this case against old versions > of Math::BigFloat, and it turns out that this is a very old bug. > > Math::BigFloat does not support complex numbers, so raising a negative > number to a non-integer power should result in a NaN.

A related issue, which triggers a warning on the same line as the code above, is the following: === BEGIN-CODE === use Math::BigFloat; Math::BigFloat->precision(-10); # commenting this, the warning is not triggered my $x = Math::BigFloat->new(1.24); my $y = Math::BigFloat->new(9.14); CORE::say $x**$y; === END-CODE=== Outputs: Use of uninitialized value in numeric le (<=) at /usr/share/perl5/site_perl/Math/BigFloat.pm line 2555. 1.0000000000