qrunch.tools.minimizers.parameter_sorting_strategy.descending_abs_value_strategy

Implements a parameter sorting strategy that sorts parameters by descending absolute value.

Classes

DescendingAbsoluteValueSortingStrategy

Sort tail parameters by descending absolute value.
class DescendingAbsoluteValueSortingStrategy

Bases: ParameterSortingStrategy

Sort tail parameters by descending absolute value.

Semantics

Given an initial_guess of length L (values for the last L symbolic parameters of the circuit), the strategy:

  1. Collects those tail symbolic gates and re-bases their indices locally to 0..L-1.

  2. Orders the parameters by -abs(value) using a stable sort so ties preserve original relative order within the tail.

  3. Returns the permuted parameter values and the associated permutation of local indices.

Characteristics

  • Ignores numeric-only gates entirely.

  • Stable: equal absolute values retain original input order.

  • Deterministic (no RNG usage).

  • Returns empty arrays if L == 0.

__init__(*args, **kwargs)
sort_parameters(initial_guess: ndarray[Any, dtype[float64]], circuit: Circuit, operator: HermitianPauliSum | None) → ParameterSortingStrategyResult

Sort the selected (tail) symbolic parameters by descending absolute value.

Parameters:

  • initial_guess (ndarray[Any, dtype[float64]]) – The initial guess array of length \(L\).

  • circuit (Circuit) – The parameterized quantum circuit corresponding to \(|\psi(\theta)\rangle\).

  • operator (HermitianPauliSum | None) – The Hermitian operator \(O\) used to compute expectation values: \(\langle \psi(\theta) | O | \psi(\theta) \rangle\).

Return type:

ParameterSortingStrategyResult