AdaptiveConfig

class AdaptiveConfig(tolerance_error=1e-08, layout_method='trivial', optimizer='bfgs2', global_search='basin_hopping', n_trials_warm=1, n_trials_cold=1, warm_start_from_level=1, refine_best_trial=False, parallel_mode='openmp', use_float=False, round_single_qubit_rotations=None, seed=None, verbosity='silent', barrier=False, compress=None, level_max=25, level_min=0, layout_objective='gate_count')

Configuration for the adaptive decomposition strategy.

The adaptive strategy grows the ansatz level by level, optimizing at each step and stopping once tolerance_error is met, then runs the tolerance-respecting compression pass to drop redundant layers. One level here is one full sweep over the topology’s edges, so n_qubits - 1 layers on a star.

The fields this strategy defines come first below. The rest are inherited from SynthesisConfig, which documents them in full.

Variables:

  • compress – Whether to run the backend circuit-compression pass after the initial decomposition. Tolerance-respecting: the backend removes a layer only if, after re-optimizing the reduced circuit, the residual error is still below tolerance_error; a removal that would cross the tolerance is rejected and the previous circuit kept. None (the default) leaves it to the gate set; the gate sets of this package compress.

  • level_max – Maximum number of ansatz levels the backend may use before giving up.

  • level_min – Minimum number of ansatz levels the search may use. 0 (the default) starts from the circuit without entanglers, which a target such as the identity needs. Raise this only to skip levels a target is known to outgrow. Read by the adaptive strategy alone; the others take no floor.

  • tolerance_error – Target compilation error; small enough makes the compilation numerically exact. (more details)

  • layout_method – Boundary qubit relabeling, or the strategy that searches for one. (more details)

  • layout_objective – What a layout_method search minimizes over the candidates. "gate_count" (the default) counts the native entanglers. "error" ranks a candidate by its overall error: its compilation error and the errors the gate set gives its gates, combined. Every candidate meets tolerance_error, so "error" puts the mapping whose gates cost least first. It needs the error of every native gate that is not a frame change. "duration" ranks a candidate by how long its circuit runs: each gate starts once the qubits the gate set says it reserves are free. It needs the duration of every native gate that is not a frame change, and the compiled circuit’s duration is recorded on metadata["duration"], in the unit the gate set’s durations are given in.

  • optimizer – Backend numerical optimizer that fits the ansatz parameters. (more details)

  • global_search – Which global search wraps BFGS2’s local optimizer. Other optimizers ignore it. (more details)

  • n_trials_warm – Optimizations per candidate gate structure seeded from the previous level’s result. (more details)

  • n_trials_cold – Optimizations per candidate gate structure from a fully random point. (more details)

  • warm_start_from_level – First level whose warm trials seed from the previous one. (more details)

  • refine_best_trial – Whether a further optimization refines the best trial. (more details)

  • parallel_mode – Backend parallelization mode. (more details)

  • use_float – Whether to run the backend in single precision. (more details)

  • seed – Seed for the backend’s random number generator. (more details)

  • verbosity – How much the compiler prints. (more details)

  • round_single_qubit_rotations – The angles to round single-qubit rotation angles onto where the fit allows. (more details)

  • barrier – Whether to insert a barrier after every native entangler, for readability when drawing. (more details)

Raises:

ConfigError – If level_min is negative or above level_max.