Path: csiph.com!optima2.xanadu-bbs.net!xanadu-bbs.net!feeder.erje.net!1.eu.feeder.erje.net!eternal-september.org!feeder.eternal-september.org!mx02.eternal-september.org!.POSTED!not-for-mail
From: Paul Rubin <no.email@nospam.invalid>
Newsgroups: comp.lang.python
Subject: Re: Bug in floating point multiplication
Date: Thu, 02 Jul 2015 08:26:06 -0700
Organization: A noiseless patient Spider
Lines: 12
Message-ID: <87a8vepomp.fsf@nightsong.com>
References: <55955048$0$1662$c3e8da3$5496439d@news.astraweb.com>
Mime-Version: 1.0
Content-Type: text/plain
Injection-Info: mx02.eternal-september.org; posting-host="38b554b2114f997ed08ebb809b0b9924"; logging-data="22645"; mail-complaints-to="abuse@eternal-september.org";	posting-account="U2FsdGVkX188QhQ/bccFd6agG62KPh0z"
User-Agent: Gnus/5.13 (Gnus v5.13) Emacs/24.3 (gnu/linux)
Cancel-Lock: sha1:bmI5a0ThKd+XqE37mg8b0HYGHEs= sha1:5lO9MYxATRsp0buDUBekZa/w5SY=
Xref: csiph.com comp.lang.python:93431

Steven D'Aprano <steve@pearwood.info> writes:
> x = 1 - 1/2**53
> assert x == 0.9999999999999999

In Python 2.x I don't see how that assert can possibly succeed, since
x is the integer 1.  But I tested it anyway on 2.7.5 under Fedora 19
and it threw an assertion error.

I changed it to say 1 - 1/2.0**53 and then the loop runs to completion.

sys.version is:
2.7.5 (default, Nov  3 2014, 14:33:39) \n[GCC 4.8.3 20140911 (Red Hat 4.8.3-7)]