Add notebook on prior selection and prior-predictive simulation

[jessegrabowski]

This will also be able how to use PyMC. Interesting topics include:

1. Using preliz to choose priors (including using pz.maxent and .plot_interactive
2. Using pymc machinery to get sample statistics as symbolic functions, for example stationary autocovariance matrix and IRF statistics
3. Use sample_prior_predictive to show the effect of varying priors on different quantities of interest

Some of this can come from the 2nd half of the introduction to gEconpy notebook, which is way too bloated.