0
$\begingroup$

I need a Poisson distribution in an application, but in that application the only available functions for generating distributions are Beta, Uniform, and Normal functions. Now, I am thinking is there a way I could draw a beta distribution, but change the parameters in a way so as to approximate a Poisson distribution? Thanks.

$\endgroup$
1
  • 3
    $\begingroup$ Since a Beta(1,1) is a Uniform(0,1), you can code any generic Poisson random generator using that Beta generator. Or you can use the connection between Poisson and Exponential, plus the fact that a $\chi^2_2$ is an Exponential $\cal{E}(1/2)$. $\endgroup$ – Xi'an Apr 4 '18 at 7:37
1
$\begingroup$

Answer from comments: Since a beta(1,1) distribution is uniform(0,1), you can just use that to code a standard poisson variable generator. Some algorithms are given here: https://en.wikipedia.org/wiki/Poisson_distribution#Generating_Poisson-distributed_random_variables.

Or you can use the connection between Poisson and Exponential, plus the fact that a $\chi^2_2$ is an Exponential $\cal{E}(1/2)$.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.