My model uses several bounded uniform distributions, which I’m representing with bounded normal distributions with large, uninformative standard deviations. When I was debugging my model, I tried to verify that I was implementing these pseudo-uniform variables correctly by checking the logp of them…

Bounded variable logp

junpenglao August 12, 2018, 1:26pm 4

I dont quite get what do you mean by “falls off”.

Topic		Replies	Views
X_lowerbound__ or X_interval__ in the Trace Questions	1	701	November 21, 2017
pm.find_MAP won't follow the bounds of custom distribution Questions	1	442	December 9, 2019
Bounded variable transformation when not using HMC? version agnostic	8	628	February 14, 2023
[quick conceptual question] shouldn't the lognormal distribution be used as the likelihood more often? Questions	5	1209	August 6, 2019
Bound variables misbehaving Questions	1	504	December 10, 2020