Path: news.mathworks.com!not-for-mail
From: <HIDDEN>
Newsgroups: comp.soft-sys.matlab
Subject: Re: Question on Poisson distrbution
Date: Thu, 2 Apr 2009 15:57:01 +0000 (UTC)
Organization: The MathWorks, Inc.
Lines: 18
Message-ID: <gr2n8d$on6$1@fred.mathworks.com>
References: <gr2lcg$kg7$1@fred.mathworks.com>
Reply-To: <HIDDEN>
NNTP-Posting-Host: webapp-02-blr.mathworks.com
Content-Type: text/plain; charset="ISO-8859-1"
Content-Transfer-Encoding: 8bit
X-Trace: fred.mathworks.com 1238687821 25318 172.30.248.37 (2 Apr 2009 15:57:01 GMT)
X-Complaints-To: news@mathworks.com
NNTP-Posting-Date: Thu, 2 Apr 2009 15:57:01 +0000 (UTC)
X-Newsreader: MATLAB Central Newsreader 1187260
Xref: news.mathworks.com comp.soft-sys.matlab:529776

"Ag" <ag@gmail.com> wrote in message <gr2lcg$kg7$1@fred.mathworks.com>...
> .....
> I am working with random variables and I want to generate random values with a Poisson distribution whose ranges are between 0.8 and 1 and with a mean distribution of 5.
> ......
> a=0.8 + (1-0.8) * poissrnd(5,1,10)
> ......

  I believe you are confusing the notions of matlab's 'rand' function with those of 'poissrnd'.  The output of 'rand' is essentially a continuous, uniform distribution on the interval [0,1], and when you write

 x = a + (b-a)*rand

you get a uniform distribution over the interval [a,b].

  No such result can be obtained from 'poissrnd' using such a computation.  In the first place the poisson distribution is discrete with integer values, and in the second place it is unbounded.  Like normal distributions, there is no limit to its possible size.  The 'a' in your computation could very easily extend far beyond the interval [.8,1].

  You need to rethink what it is you wish to achieve.

Roger Stafford