Hi Quentin
Sorry for the confusion
If x and y are Poisson distributed, with means M_x and M_y, then x|(x+y)
follows a Binomial distribution with N = x+y, and p = M_x/(M_x+M_y)
Dave.
"Quentin L. Burrell" <[log in to unmask]> on 06/02/2002 13:42:42
To: David Mcnulty/GB/QUES/ICI@ICI
cc:
Subject: RE: Ratio of two Poisson Variables
Dave
Maybe I'm not having a good day but I just don't understand the "simple"
answer. Can you expand?
Quentin
-----Original Message-----
From: A UK-based worldwide e-mail broadcast system for the statistical
community, [mailto:[log in to unmask]]On Behalf Of
[log in to unmask]
Sent: 06 February 2002 09:34
To: [log in to unmask]
Subject: Re: Ratio of two Poisson Variables
Hi Folks,
A big thankyou to everyone who has responed both publicly and privately.
Here is a quick summary of responses:
The simple answer is use the Binomial distribution for x/(x+y) where both x
and y are Poisson.
Alternatively a partial solution to the question is given in the book:
Lindley, D.V. (1965) Probability and Statistics: Part 2 Inference
In return here is a tip I came accross some time ago.
Leslie Daly (1992). "Simple SAS macros for the calculation of exact
Binomial and Poisson confidence limits." Comput. Biol. Med. Vol 22 No 5,
pp351-361
Dave.
IMPORTANT NOTICE:
This email is confidential, may be legally privileged, and is for the
intended recipient only. Access, disclosure, copying, distribution, or
reliance on any of it by anyone else is prohibited and may be a criminal
offence. Please delete if obtained in error and email confirmation to the
sender.
IMPORTANT NOTICE:
This email is confidential, may be legally privileged, and is for the
intended recipient only. Access, disclosure, copying, distribution, or
reliance on any of it by anyone else is prohibited and may be a criminal
offence. Please delete if obtained in error and email confirmation to the
sender.
|