During iterative solving, it can sometimes be useful to "mask" certain scores. For example, if X cannot be a parent of Y due to some external constraints, then we would "mask" all parent sets of Y which include X. For sparse parent lists, this can be implemented simply by adding a bitset which indicates which parent sets are masked.
During iterative solving, it can sometimes be useful to "mask" certain scores. For example, if X cannot be a parent of Y due to some external constraints, then we would "mask" all parent sets of Y which include X. For sparse parent lists, this can be implemented simply by adding a bitset which indicates which parent sets are masked.