Closed WrathfulSpatula closed 2 years ago
These two simulation approaches aren't really as natural or perfect of a match as I supposed in the first issue comment. It's worth keeping the idea in mind, though. Moving to backlog.
I don't see a viable approach to doing this, for now. Closing.
Learning from
QBdt
, we see thatQEngineShard
could attach twoQInterface
instances "under" each of its two probability amplitudes, as by multiplication of eachQEngineShard
probability amplitude on their respectiveQInterface
branch. Further, even with differentQEngineShard
probability amplitudes, theQInterface
instances could be represented by the same pointer value, to oneQInterface
instance.The issue is to implement a
QEngineShard
as aQBdtNode
to at least a single level of depth. This might be accomplished with inheritance ofQEngineShard
fromQBdtNode
, in part.Long-term,
QEngineShard
might implement a multi-levelQBdtNode
hierarchy between otherQEngineShards
, but I haven't fully developed the requisite theory, and perhaps that won't turn out to be the best approach.