Generate statistical description from model

[null-a]

I wonder if it would be useful to be able to generate a description of a brmp model in the form of a series of statistical formulas. Something like:

model = defm('hits | player', df, family=Binomial(num_trials=45))

# Produces:
#
# hits_i ~ Binomial(p_i, num_trials=45)
# logit(p_i) ~ a_player
# a_player ~ Normal(0,1)

We might be able to implement this as a backend (degenerate -- no inference methods), thereby making it available to users through something like:

model.generate(backend=math)

Presumably this could generate LaTeX for pretty rendering in notebooks.