Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!lll-crg!rutgers!sri-spam!sri-unix!hplabs!sdcrdcf!ism780c!tim
From: tim@ism780c.UUCP (Tim Smith)
Newsgroups: net.sources
Subject: Re: mz.c
Message-ID: <4319@ism780c.UUCP>
Date: Fri, 7-Nov-86 19:32:57 EST
Article-I.D.: ism780c.4319
Posted: Fri Nov  7 19:32:57 1986
Date-Received: Sun, 9-Nov-86 03:16:28 EST
References: <4247@ism780c.UUCP> <823@zen.BERKELEY.EDU> <1125@cit-vax.Caltech.Edu>
Reply-To: tim@ism780c.UUCP (Tim Smith)
Followup-To: net.sources.d
Organization: Interactive Systems Corp., Santa Monica, CA
Lines: 19

Note: followups are being redirected to net.sources.d

In article <1125@cit-vax.Caltech.Edu> trent@cit-vax.UUCP (Ray Trent) writes:
>
>I suppose this works...but isn't:
>
>     #define oneof(n)    (rand() % (n) + 1)
>
>more intuitive and efficient?

Yes, but it also doesn't work!  If n is not a divisor of 32768, the
distribution will be off.  For example, if n is 24000, then numbers
from 1 to 8000 will be given twice as often as numbers from 9000 to 24000.

-- 
emordnilapanalpanama

Tim Smith       USENET: sdcrdcf!ism780c!tim   Compuserve: 72257,3706
                Delphi or GEnie: mnementh