I’m having difficulty finding resources that explain exactly what happens under the hood in an ordered transformation.
with pm.Model() as m12_4:
cutpoints = pm.Normal(
"cutpoints",
0.0,
1.5,
transform=pm.distributions.transforms.ordered,
shape=6,
testval=np.arange(6) - 2.5,
)
R = pm.OrderedLogistic("R", 0.0, cutpoints, observed=trolley_df.response.values - 1)