We may want to drop this loop, instead provide the specific simulation as an input and output only the split cells for that seed. That way we can parallelize the whole stage (which doesn't lead to a huge overhead if we make sure MCMs are only calculated if they don't already exitst)
We may want to drop this loop, instead provide the specific simulation as an input and output only the split cells for that seed. That way we can parallelize the whole stage (which doesn't lead to a huge overhead if we make sure MCMs are only calculated if they don't already exitst)