New user here, teething problems are everywhere.
The main issue is when I run a model with NUTS, it crashes the notebook kernel I am running. Under Metropolis, the sampling function runs, but there is no output to examine. The code doesn’t seem to be faulty, nothing out of the ordinary from the tutorial notebooks I’ve seen.
Not really sure why this is happening, code below if anyone would have a look. Any help appreciated
with pm.Model() as linear:
# Priors
alpha = pm.Uniform(‘alpha’, lower=-10, upper=10)
beta = pm.Uniform(‘beta’, lower=-10, upper=10)
# Likelihood
phat = alpha + beta * distance
logitp = pm.Binomial('logitp', p=phat, n=len(distance), observed= logit)
where distance = array([ 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18,
19, 20], dtype=int64)
and logit = array([ 2.63018153, 1.59566833, 1.04939831, 0.36080434, 0.19176195,
0.12516314, -0.1502822 , -0.76310577, -0.68565651, -0.77010822,
-1.05939158, -1.15496523, -0.7985077 , -1.60226942, -1.86321843,
-1.66587922, -1.56608747, -1.84845481, -1.67397643])