Hi there!
Thanks for the quick reply - I am glad to hear these plots seem correct!
With equal parts being new to the Bayesian world and not being the biggest fan of logistic regression, I may just need more foundational knowledge to accomplish my current goals. With these plots, I anticipated some S-shape, but not necessarily the missing y-axis and scale of the x-axis. Given your explanation, I follow the y-axis roughly displaying proportions or counts, etc. However, why is the x-axis scaled from zero to two? Is it a function of binning, where less than one represents zero and greater or equal to one represents one?
That being said, it was mainly the prior predictive check that left me slightly confused. Below is my final version of the graph with 1,000 samples. How exactly would one describe this visual? I am assuming it would be ideal for the prior predictive mean to be equal across zeros and ones… Would you mind helping me finish that thought?
For the posterior predictive check, given the overlap, one would state that the model is satisfactory at retrodicting the data. Does that sound correct?
I am sure a lot of this information is PyMC 101, so I greatly appreciate your guidance!