Alexander Condello
Alexander Condello
Ok, I took a deeper look, it appears that we currently only accept `flux_biases` as a qubit-linked list. But we do the abstraction in [VirtualGraphComposite](https://github.com/dwavesystems/dwave-system/blob/105a722e93e94bcc530f37c09283c64076f708c4/dwave/system/composites/virtual_graph.py#L149). **Proposal** Move the `flux_biases` abstraction...
Agree the documentation should be updated to make it explicit that the cutoff is around 0.
I suppose we could update the composite to allow the user to specify a range (treating a single number as a symmetric range) e.g. ``` CutOffComposite(AutoEmbeddingComposite(sampler), .75) ``` would be...
Who are we to tell people what to use it for? :smile: But yeah, need a use-case to justify the implementation time.
There is a case to be made that that's a very poor QPU problem, though I agree distinguishing that case will be very difficult.
Fair enough, and it is definitely true that there will be many heuristics worse than the QPU at that problem. In any case, there is no way (that I know...
#255 adds `EnergyScaleWarning` and `ChainStrengthWarning`
#255 has now been merged
Let's keep them in dwave-system for now
No, I just seem to really like that spelling