Jesse Berwald

Your first code block basically lays out what I'm doing already (with more code, of course). So yes, that seems like a nice solution.

As a side note, it seems like the TilingComposite excludes a fair amount of real estate in the example above. Am I understanding it's tiling procedure correctly in relation to...

A dict would indeed be preferable. Eg., `{qubit_name: offset}`, where `qubit_name` is the variable name in the BQM and any qubits not included in the dict are inferred to be...

The standard terminology for this functionality seems to be `feed_dict`, which I personally dislike, but has permeated the tech world via TF. What about `parameter_dict` instead of `..._default`? It's more...

I would also advocate for simple information such a `max_chain_length` and `number_max_chains` to be returned as attributes of some sort. That way a user can decide whether the embedding is...

@boothby, regarding your comments above, how would chain connectivity information recorded? Do you imagine sending more vectors to `find_embedding()` would be the way? Eg., `vector[vector[int]] chain_couplers`?