Re: GammaDistribution versus PoissonDistribution
- To: mathgroup at smc.vnet.net
- Subject: [mg79000] Re: GammaDistribution versus PoissonDistribution
- From: Bill Rowe <readnewsciv at sbcglobal.net>
- Date: Sun, 15 Jul 2007 01:09:45 -0400 (EDT)
On 7/14/07 at 2:42 AM, peter_van_summeren at yahoo.co.uk (P_ter) wrote: >Hello, as far as I understand the GammaDistribution has two >parameters: alfa and beta. With beta =1 one should get the >PoissonDistribution with paramter alfa-1. I did, just to check: You are incorrect here. The Poisson distribution is a discrete distribution. The gamma distribution is a continuous distribution. There are no parameters for the gamma distribution that will result in a Poisson distribution. However, there is a relationship between the Poisson distribution and gamma distribution. Specifically, suppose the number of events (n) in some specified length of time (t) has a Poisson distribution. The the time between subsequent events will have a gamma distribution with beta = 1 (an exponential distribution) and alpha = n/t. In general the time from a given event k to event k+m, will be gamma distributed with beta = m and alpha = n/t. -- To reply via email subtract one hundred and four