Path: csiph.com!x330-a1.tempe.blueboxinc.net!usenet.pasdenom.info!news.albasani.net!newsfeed.freenet.ag!news2.euro.net!newsgate.cistron.nl!newsgate.news.xs4all.nl!post.news.xs4all.nl!not-for-mail
To: python-list@python.org
From: Terry Reedy <tjreedy@udel.edu>
Subject: Re: difference between random module in python 2.6 and 3.2?
Date: Mon, 06 Feb 2012 00:07:04 -0500
References: <jgna4q$1t08$1@ns.felk.cvut.cz> <4f2f5081$0$29992$c3e8da3$5496439d@news.astraweb.com>
Mime-Version: 1.0
Content-Type: text/plain; charset=UTF-8; format=flowed
Content-Transfer-Encoding: 7bit
User-Agent: Mozilla/5.0 (Windows NT 6.1; WOW64; rv:8.0) Gecko/20111105 Thunderbird/8.0
In-Reply-To: <4f2f5081$0$29992$c3e8da3$5496439d@news.astraweb.com>
Precedence: list
Newsgroups: comp.lang.python
Message-ID: <mailman.5464.1328504850.27778.python-list@python.org>
Lines: 47
NNTP-Posting-Host: 2001:888:2000:d::a6
Xref: x330-a1.tempe.blueboxinc.net comp.lang.python:19887

On 2/5/2012 11:01 PM, Steven D'Aprano wrote:

> Reading the docs, I would expect that when using an int as seed, you
> should get identical results.

That is similar to expecting hash to be consistent from version to version.

> There is no mention that the PRNG has changed between 2.6 and 3.2;

There is at best an informal policy. This was discussed in
http://bugs.python.org/issue9025
Antoine argued that if there were a written policy, it should be limited 
to bug-fix releases within a version. I agree.

> It appears to be a bug in 3.2, because 3.1 gives the same results as 2.6:

This change is a side effect of fixing the bug of non-uniformity 
discussed in that issue. In any case, in 2.7 and probably 3.1:

     def choice(self, seq):
         """Choose a random element from a non-empty sequence."""
         return seq[int(self.random() * len(seq))]  # raises IndexError

whereas in 3.2:

     def choice(self, seq):
         """Choose a random element from a non-empty sequence."""
         try:
             i = self._randbelow(len(seq))
         except ValueError:
             raise IndexError('Cannot choose from an empty sequence')
         return seq[i]

The change was announced in What's New in 3.2

random
The integer methods in the random module now do a better job of 
producing uniform distributions. Previously, they computed selections 
with int(n*random()) which had a slight bias whenever n was not a power 
of two. Now, multiple selections are made from a range up to the next 
power of two and a selection is kept only when it falls within the range 
0 <= x < n. The functions and methods affected are randrange(), 
randint(), choice(), shuffle() and sample().

-- 
Terry Jan Reedy