Article View: rec.puzzles
Article #2573Re: Unfair coin (CORRECTION)
From: Dave Dodson
Date: Thu, 07 Jan 1993 18:54
Date: Thu, 07 Jan 1993 18:54
69 lines
4295 bytes
4295 bytes
In article <C0G2Mv.4yq@hplntx.hpl.hp.com> potamian@hpl.hp.com (Spyros Potamianos) writes: >Oooooooooops! I made a mistake in my previous posting about approximating >an arbitrary probability with an unfair coin. The number of tosses you >need is always unbounded (not finite). >You see, what happened is that in my mind I broke the problem in two >subproblems: > a) simulate a fair coin using an unfair coin > b) generate P (or an approximation to it) from a fair coin > >For some values of P (b) only needs a finite number of tosses (e.g. >P=0.25). However when combining both (a) and (b) to form the problem I >posted then the number of tosses is obviously unbounded.... >However the average number of tosses you need is still finite. It is not necessary to simulate a fair coin if you know the probability of heads and tails for the unfair coin. And since the original problem description (which has expired on our machine) contained a tolerance, you can quit after a finite number of steps. For example, if the unfair coin comes up tails 10% of the time and heads 90% of the time, and you want to approximate a probability of 0.2 to within a tolerance of one part in a million, use the following rules in order, one for each flip of the coin. If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, keep rolling; if heads, report failure If tails, report success; if heads, keep rolling If tails, keep rolling; if heads, report failure If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, keep rolling; if heads, report failure If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, keep rolling; if heads, report failure If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, keep rolling If tails, report success; if heads, report failure Reporting success or failure means that you stop the process. As another example with the same coin, to approximate 0.19 If tails, report success; if heads, keep rolling If tails, report success; if heads, report failure ---------------------------------------------------------------------- Dave Dodson dodson@convex.COM Convex Computer Corporation Richardson, Texas (214) 497-4234
Technical Details
Message-ID:
<1993Jan7.185427.22133@news.eng.convex.com>Path:
arc1-ca.pugleaf.net!sparky!uunet!news.gtech.com!noc.near.net!news.Brown.EDU!qt.cs.utexas.edu!yale.edu!jvnc.net!newsserver.jvnc.net!darwin.sura.net!zaphod.mps.ohio-state.edu!cs.utexas.edu!hermes.chpc.utexas.edu!news.utdallas.edu!convex!dodsonReferences:
<C0G2Mv.4yq@hplntx.hpl.hp.com>Show Full Headers
Complete Article Headers
Newsgroups: rec.puzzles
Path: arc1-ca.pugleaf.net!sparky!uunet!news.gtech.com!noc.near.net!news.Brown.EDU!qt.cs.utexas.edu!yale.edu!jvnc.net!newsserver.jvnc.net!darwin.sura.net!zaphod.mps.ohio-state.edu!cs.utexas.edu!hermes.chpc.utexas.edu!news.utdallas.edu!convex!dodson
From: Dave Dodson <dodson@convex.COM>
Subject: Re: Unfair coin (CORRECTION)
Originator: dodson@bach.convex.com
Sender: usenet@news.eng.convex.com (news access account)
Message-ID: <1993Jan7.185427.22133@news.eng.convex.com>
Date: Thu, 7 Jan 1993 18:54:27 GMT
Reply-To: dodson@convex.COM (Dave Dodson)
References: <C0G2Mv.4yq@hplntx.hpl.hp.com>
Nntp-Posting-Host: bach.convex.com
Organization: Engineering, CONVEX Computer Corp., Richardson, Tx., USA
X-Disclaimer: This message was written by a user at CONVEX Computer
Corp. The opinions expressed are those of the user and
not necessarily those of CONVEX.
Lines: 69