Based on the PyMC port of the book “Rethinking Statistics - e2”, I have been trying to reproduce the normal approximation to a binomial. I get the correct result for the mean, but the SD is off by a factor of 4 - and I cannot find the problem. This is also relevant for the https://github.com/pymc-d…

Problems with finding normal approximation to binomial

opherdonchin December 22, 2022, 8:37pm 4

The issue is also discussed extensively in this thread.

1 Like

Topic		Replies	Views
Beta-Binomial conjugate prior -- pm.Binomial buggy results...? Questions	9	726	November 3, 2021
The standard deviation of posterior distribution remains the same compared to prior v5 modeling	16	401	January 21, 2024
Sampler issues on Beta prior Binomial likelihood v5 bug	5	331	December 8, 2022
Any appetite for a quadratic approximation PR? Development	4	1115	July 8, 2021
Estimating standard deviation of very coarsely binned data v5	6	326	January 25, 2023