[R] rgamma question

[Thomas Lumley]

On Fri, 6 Feb 2004, Icabalceta, Jorge L. wrote:

> Sorry to bother you. I am sort of confused with the random number generation
> from a gamma distribution. For some reason, if a variable X~gamma(shape,
> scale)I have been told that the mean mu can be eithe mu=shape/scale or
> mu=shape*scale. Why is that so?

It isn't.  The mean is always shape*scale.  However, the notation
Gamma(alpha,beta) is ambiguous.  Some people use beta=scale (in which
case mean=alpha*beta)  and others use beta=1/scale (in which case
mean=alpha/beta).  A few people even use use alpha=mean, beta=variance,
but they usually know who they are.


	-thomas

Thomas Lumley			Assoc. Professor, Biostatistics
tlumley at u.washington.edu	University of Washington, Seattle