SampleSet energy

Hey everyone, how does dwave calculated sampleSet energy? I know that when we sample using (say) sample_ising(), we get a sampleSet object with: samples, energy of each sample and number of occurrences. My question is how does dwave find the energy associated with a particular sample?

I couldn't find many references on this online. Thanks!


Helpful link:-
(https://docs.ocean.dwavesys.com/en/latest/docs_cloud/reference/generated/dwave.cloud.solver.StructuredSolver.sample_ising.html#dwave.cloud.solver.StructuredSolver.sample_ising)

0

Comments

1 comment
  • Hi Sehmimul,

    The energy landscape is defined by the linear and quadratic biases of a problem formulation.

    For example, consider an example where the objective function is -2a+b-4ab and a,b are binary variables. The solution space for this example is:

    Submitting this problem to the QPU:

    >>> from dwave.system import DWaveSampler, EmbeddingComposite
    >>> sampler = EmbeddingComposite(DWaveSampler())
    >>> Q = {('a', 'a'): -2, ('b', 'b'): 1, ('a','b'): -4}
    >>> sampleset = sampler.sample_qubo(Q, num_reads = 10)
    >>> print(sampleset)
       a  b energy num_oc. chain_.
    0  1  1   -5.0      10     0.0
    ['BINARY', 1 rows, 10 samples, 2 variables]

    As expected, the example above returns the lowest energy solution with a=b=1 and energy -5.0.

    You might also find our getting started guide helpful: https://docs.dwavesys.com/docs/latest/doc_getting_started.html

    0
    Comment actions Permalink

Please sign in to leave a comment.

Didn't find what you were looking for?

New post