Skip Menu |

This queue is for tickets about the Math-Pari CPAN distribution.

Report information
The Basics
Id: 14891
Status: new
Priority: 0/
Queue: Math-Pari

People
Owner: Nobody in particular
Requestors: kynn [...] panix.com
Cc:
AdminCc:

Bug Information
Severity: (no value)
Broken in: (no value)
Fixed in: (no value)



Date: Mon, 3 Oct 2005 12:18:41 -0400 (EDT)
From: <kynn [...] panix.com>
To: bug-Math-Pari [...] rt.cpan.org
Subject: Incorrect conversion of unnormalized IEEE doubles
severity=high Math::Pari v2.010603 See below for other system details. Math::Pari produces wrong results when it converts unnormalized IEEE doubles to PARI reals. For example print POSIX::DBL_MIN()/2 1.1125369292536e-308 ...while print Math::Pari::PARI( POSIX::DBL_MIN()/2 ) 1.668805393880401037E-308 ...which is actually POSIX::DBL_MIN*(3/4). This represents a relative error of 50%, which is far from negligible. It appears that Math::Pari::PARI is converting numbers of the form POSIX::DBL_MIN()/2**$n (where $n is a positive integer) into numbers of the form POSIX::DBL_MIN()*((2**$n)-1)/(2**(-$n-1)) This problem occurs silently when an unnormalized IEEE double is automatically promoted to a PARI real, as happens, for example, in arithmetical operations involving both the unnormalized IEEE double and a PARI real. ----------------------------------------------------------------- --- Site configuration information for perl v5.8.6: Configured by kynn at Tue Apr 12 12:24:11 EDT 2005. Summary of my perl5 (revision 5 version 8 subversion 6) configuration: Platform: osname=linux, osvers=2.4.18-686-smp, archname=i686-linux-ld uname='linux luna 2.4.18-686-smp #1 smp sun apr 14 12:07:19 est 2002 i686 unknown ' config_args='' hint=recommended, useposix=true, d_sigaction=define usethreads=undef use5005threads=undef useithreads=undef usemultiplicity=undef useperlio=define d_sfio=undef uselargefiles=define usesocks=undef use64bitint=undef use64bitall=undef uselongdouble=define usemymalloc=n, bincompat5005=undef Compiler: cc='cc', ccflags ='-fno-strict-aliasing -pipe -I/usr/local/include -D_LARGEFILE_SOURCE -D_FILE_OFFSET_BITS=64', optimize='-O2', cppflags='-fno-strict-aliasing -pipe -I/usr/local/include' ccversion='', gccversion='2.95.4 20011002 (Debian prerelease)', gccosandvers='' intsize=4, longsize=4, ptrsize=4, doublesize=8, byteorder=1234 d_longlong=define, longlongsize=8, d_longdbl=define, longdblsize=12 ivtype='long', ivsize=4, nvtype='long double', nvsize=12, Off_t='off_t', lseeksize=8 alignbytes=4, prototype=define Linker and Libraries: ld='cc', ldflags =' -L/usr/local/lib' libpth=/usr/local/lib /lib /usr/lib libs=-lnsl -lgdbm -ldbm -ldb -ldl -lm -lcrypt -lutil -lc perllibs=-lnsl -ldl -lm -lcrypt -lutil -lc libc=/lib/libc-2.2.5.so, so=so, useshrplib=false, libperl=libperl.a gnulibc_version='2.2.5' Dynamic Linking: dlsrc=dl_dlopen.xs, dlext=so, d_dlsymun=undef, ccdlflags='-Wl,-E' cccdlflags='-fpic', lddlflags='-shared -L/usr/local/lib' Locally applied patches: --- @INC for perl v5.8.6: /home/kynn/local/lib/perl/my /home/kynn/local/lib/perl/5.8.4 /home/kynn/local/lib/perl/5.6.1 /home/kynn/local/lib/perl5/5.8.6/i686-linux-ld /home/kynn/local/lib/perl5/5.8.6 /home/kynn/local/lib/perl5/i686-linux-ld /home/kynn/local/lib/perl5 /home/kynn/local/lib/perl5/5.8.6/i686-linux-ld /home/kynn/local/lib/perl5/5.8.6 /home/kynn/local/lib/perl5/site_perl/5.8.6/i686-linux-ld /home/kynn/local/lib/perl5/site_perl/5.8.6 /home/kynn/local/lib/perl5/site_perl . --- Environment for perl v5.8.6: HOME=/home/kynn LANG=C LANGUAGE (unset) LD_LIBRARY_PATH=/home/kynn/local/lib:/opt/java/jdk/lib/i386 LOGDIR (unset) PATH=.:/home/kynn/local/bin:/home/kynn/local/usr/bin:/home/kynn/local/local/bin:/bin:/usr/bin:/usr/local/bin:/usr/X11R6/bin:/home/kynn/java/javacc/bin:/opt/java/jdk/bin:/usr/pbs/bin:/home/kynn/opt/ant/bin PERL5LIB=/home/kynn/local/lib/perl/my:/home/kynn/local/lib/perl/5.8.4:/home/kynn/local/lib/perl/5.6.1:/home/kynn/local/lib/perl5 PERL5_CPANPLUS_CONFIG=/home/kynn/.cpanplus/config PERL_BADLANG (unset) SHELL=/usr/bin/zsh
Date: Tue, 4 Oct 2005 18:57:34 -0700
From: Ilya Zakharevich <nospam-abuse [...] ilyaz.org>
To: "kynn [...] panix.com via RT" <bug-Math-Pari [...] rt.cpan.org>
Subject: Re: [cpan #14891] Incorrect conversion of unnormalized IEEE doubles
RT-Send-Cc:
On Mon, Oct 03, 2005 at 12:19:11PM -0400, kynn@panix.com via RT wrote: Show quoted text
> Math::Pari produces wrong results when it converts unnormalized IEEE > doubles to PARI reals. For example > > print POSIX::DBL_MIN()/2 > 1.1125369292536e-308 > > ...while > > print Math::Pari::PARI( POSIX::DBL_MIN()/2 ) > 1.668805393880401037E-308 > > ...which is actually POSIX::DBL_MIN*(3/4). This represents a relative > error of 50%, which is far from negligible. > > It appears that Math::Pari::PARI is converting numbers of the form > > POSIX::DBL_MIN()/2**$n > > (where $n is a positive integer) into numbers of the form > > POSIX::DBL_MIN()*((2**$n)-1)/(2**(-$n-1))
Thanks for a great bug report? Could you also produce a great bug fix? ;-) The problem is (IIRC) that Math::Pari uses a PARI API which (apparently) is very rarely used otherwise. So I would think it is a GP/PARI bug exposed by Math::Pari... Thanks, Ilya