I have a physical model where a discrete sample, Y_{sample}, is taken (poisson process), but then that sample is measured via a separate process as Z_{observed} which is dependent on the log of the poisson RV Y_{sample} with some gaussian noise. In the latter process, Y_{sampled}==0 from the poisson…

CdfPosPoisson does not have support on R, so my guess is that if you have lambda very close to 0 in the initial condition it breaks the sampler: https://www.google.com/search?q=log(1-exp(x))&oq=log(1-exp(x)) Also you might consider implementing the cdf in a more numerical stable way (see Eq. 7 in h…

Observing the log of Poisson RV

Questions

junpenglao December 13, 2020, 10:47am 3

It’s pretty close to mixture distribution likelihood like zero inflated poisson etc

Topic		Replies	Views
Custom discrete distribution for zero truncated Poisson Questions	10	2608	December 19, 2017
Zero Inflated Poisson Log-Lik Questions	3	1767	August 2, 2019
Poisson realization of log-normal distribution Questions	4	873	January 28, 2021
Issues with Poisson Regression using GLM method Questions	2	1057	April 5, 2018
The zero-inflated exponential Questions	5	1549	January 18, 2019

Observing the log of Poisson RV

Related topics