Randomizing physical position of embedded graph in QPU
Thinking high level and knowing that there are several embedding strategies in Ocean's DWave.
If I have a problem for which the physical embedded graph on the QPU < Pegasus physical graph on the QPU, each time I sample a set of solutions from the QPU:
a) is the embedding algorithm always choosing the same subgraph with the same Qubits/Connections on the QPU?
b) is the embedding algorithm choosing a random subgraph with different Qubits/Connections on the QPU?
c) is there a parameter to switch between behaviour a) and b) ?
E.g in this call:
response = EmbeddingComposite(DWaveSampler(token=self.sapi_token, endpoint=self.url, solver='Advantage_system1.1'), ).sample_qubo(self.qubo_dict, chain_strength=self.chainstrength, num_reads=self.numruns,annealing_time=self.annealing_time, auto_scale=True, reduce_intersample_correlation=False)
where could I define the parameter mentioned in question c) ?