pre-specified shock histories with frame architecture?

[sbenthall]

Considering Frames as an architecture #798 ...

We seem to have converged on a way to organize the definitions and execution of model variable updates.

However, I don't think our 'frame' architecture plan is yet sensitive to the question of whether exogenous shocks can and should be precomputed.

Recall that HARK uses IID exogenous shocks, whereas Dolo uses more broadly specified exogenous processes.

@mnwhite I wonder whether and to what extent the existing Shock precomputing functionality needs to be preserved in the alternative, frame-based architecture.