Interface for counterfactual prediction

May be out of scope for this project (or just for now), but @SamuelBrand1 points us to epidemia's interface for counterfactuals, which I think is cool. This is somewhat straightforward to implement, as we can build off of the built-in numpyro predictive functionality, but we would need to think about the interface.

I like the epidemia interface for prediction/counterfactual: https://imperialcollegelondon.github.io/epidemia/reference/posterior_predict.epimodel.html .
The tl; dr explanation is they fit an object (which contains posterior draws for the params) and then passing newdata can create forecasts (by being longer) and/or counterfactuals (by having different covariates like an earlier lockdown). Its demoed here, https://imperialcollegelondon.github.io/epidemia/articles/europe-covid.html .

Originally posted by @SamuelBrand1 in https://github.com/CDCgov/multisignal-epi-inference/issues/169#issuecomment-2191161181

Start by reviewing the built-in NumPyro predictive functionality and the linked epidemia prediction interface and example. Clarify whether PyRenew should support longer forecasts and changed covariates through newdata, then define the interface and its completion criteria with the project maintainers.