Probabilistic-Programming-and-Bayesian-Methods-for-Hackers icon indicating copy to clipboard operation
Probabilistic-Programming-and-Bayesian-Methods-for-Hackers copied to clipboard

Published 20 hours ago • verified publisher icon CamDavidsonPilon

Chap2, Poisson from data rate issue

Open AliMoallemi opened this issue 5 years ago • 0 comments

Build graph

data = tf.constant([10., 5.], dtype=tf.float32) rv_poisson = tfd.Poisson(rate=tf.reduce_mean(data))## instead of -->>rv_poisson = tfd.Poisson(rate=1./tf.reduce_mean(data)) in Poisson rate is mean and in Exponential it should be 1/mean poisson = rv_poisson.sample()

AliMoallemi avatar Feb 09 '20 16:02 AliMoallemi