ADAPT-VQE: building iteratively efficient ansatze

class qat.plugins.AdaptVQEPlugin(operator_pool: List[Observable], n_iterations: int = 300, early_stopper: float = 1e-06)

Plugin implementation of the ADAPT-VQE algorithm, which builds ansatze by selecting operators from a given pool of operators. The method is based on Grimsley et al. article.

  • operator_pool (List[Observable]) – List of operators to choose from. The pool of commutators is internally constructed from this list.

  • n_iterations (int, optional) – Maximum number of iteration to perform. Defaults to 300.

  • commutators (List[Union[Observable, SpinHamiltonian]) – List of commutators to use when computing the energy gradients.

  • early_stopper (float, optional) – Loss value for which the run is stopped. Defaults to 1e-6.

This plugin used in the following notebooks: