Closed grwhumphries closed 2 years ago
I think parallelization would be more efficient if applied at the scenario level, i.e. send scenario-specific stoch_crm() calls to parallel workers via e.g. furrr::future_imap(). Do you think that would work?
Actually that's a good point, especially with the implementation of the new option 3 calculations. We should add that into the examples for the vignette (how to parallelize).
Going to close this issue because I think the way the package has been put together now, parallelization can fall to the user, or you can play with the example you've put together here.
This should be able to be parallelized I would imagine. (probably on your to-do list as well)