[EMAIL PROTECTED] (rishi) wrote in message news:<[EMAIL PROTECTED]>... > Let me rephrase: > > Suppose I give a frequency and intensity for a poisson variable, such > as temperature. For days over 100 degrees, lambda = 1 (i.e., we get on > average one such day per time period). For days over 120 degrees, > lambda = .25 (we get on average one such day per four time periods). > > In other words, > > lambda for temp > 100 is 1.0 > lambda for temp > 120 is .25 > > Does this mean that lambda for temps between 100 and 120 is .75? That > is, on average we get .75 days falling in the 100 to 120 temperature > range?
While the average would be right, it is not necessarily the case that the difference between two Poissons is Poisson. It requires exactly the right kind of dependence. If the difference were Poisson then the supposition about the parameter would be correct. Glen . . ================================================================= Instructions for joining and leaving this list, remarks about the problem of INAPPROPRIATE MESSAGES, and archives are available at: . http://jse.stat.ncsu.edu/ . =================================================================
