I have been working on the Statistical Rethinking online lessons using CmdStanPy and plotnine. The link to the notebooks is at the end of the post. At the moment, there are eight weeks finished. The two last ones, covariance and missing values imputation, are really hard for my actual Stan level. I may be able to finish it in the future. Meanwhile, I hope that this work can help others to learn and enjoy Stan.
GitHub: GitHub - asuagar/statrethink-course-cmdstanpy-2019: Statistical Rethinking: A Bayesian Course Using CmdStanPy and Plotnine
this is awesome! plotnine has been on my radar for a long time.
does the current CmdStanPy interface give you enough hooks to do everything you want?
also, I see that you also made a repo Statistical Rethinking with NumPyro. what are your thoughts on the differences / similarity / advantages / etc of NumPyro vs. Stan?