qrunch.chemistry.ground_state_problem.calculators.vqe_creator

Module containing all the VQE (Variational Quantum Eigensolver) algorithms with chemistry-tailored defaults.

Functions

`chemistry_default_boson_estimator`()	Create an ExcitationGateEstimator configured for BEAST-VQE (recommended default).
`chemistry_default_boson_sampler`()	Create an ExcitationGateSampler configured for BEAST-VQE (recommended default).
`chemistry_default_fermion_estimator`()	Create an ExcitationGateEstimator configured for FAST-VQE (recommended default).
`chemistry_default_fermion_sampler`()	Create an ExcitationGateSampler configured for FAST-VQE (recommended default).
`fast_gate_selector_creator`()	Create a FAST Gate selector creator, with chemistry defaults.
`vqe_creator`()	Create a VQE algorithm creator.

Classes

`IterativeVqeCreator`	creator for all types of VQE algorithms.
`OrbitalOptimizedIterativeVqeCreator`	creator for all types of VQE algorithms.
`VqeCreator`	creator for all types of VQE algorithms.

class IterativeVqeCreator

Bases: object

creator for all types of VQE algorithms.

static beast() → BeastSecondQuantizationAdaptiveVqeCreator

Narrow the VQE type to BEAST-VQE (without orbital optimization) with chemistry defaults.

Return type:: BeastSecondQuantizationAdaptiveVqeCreator

static paired_electron_approximation() → SecondQuantizationAdaptiveVqeCreator

Narrow to paired electron approximation version (without orbital optimization).

Return type:: SecondQuantizationAdaptiveVqeCreator

static standard() → SecondQuantizationAdaptiveVqeCreator

Narrow the VQE type to an adaptive VQE (defaults to FAST-VQE), with chemistry default settings.

An adaptive VQE is a VQE that grows the ansatz during the optimization, by selecting gates from a pool. The difference between FAST-VQE and other adaptive schemes, such as ADAPT-VQE (https://arxiv.org/pdf/1812.11173) is the gate selection strategy. FAST-VQE uses a heuristic strategy to select gates (either a heuristic gradient or heuristic selected CI), while ADAPT-VQE uses the exact gradient.

To use ADAPT-VQE instead of FAST-VQE, use the with_gate_selector() method with a gate_selector created by following the adapt-path in the gate_selector_creator().

Return type:: SecondQuantizationAdaptiveVqeCreator

class OrbitalOptimizedIterativeVqeCreator

Bases: object

creator for all types of VQE algorithms.

static beast() → BeastSecondQuantizationAdaptiveOrbitalOptimizationVqeCreator

Narrow the VQE type to BEAST-VQE with orbital optimization with chemistry defaults.

BEAST-VQE is an iterative VQE (grows the ansatz during the optimization, by selecting gates from a pool), that is specific to bosonic encoding of paired electrons in molecular ground state energy problems.

In addition, the adaptive orbital optimization VQE also optimizes the molecular orbitals to find a better basis for the problem. This is particularly useful in the bosonic-encoding case, where the orbital optimization can significantly improve the accuracy of the results.

Return type:: BeastSecondQuantizationAdaptiveOrbitalOptimizationVqeCreator

static paired_electron_approximation() → SecondQuantizationAdaptiveOrbitalOptimizationVqeCreator

Narrow to paired electron approximation version with orbital optimization.

It grows the ansatz during the optimization, by selecting gates from a pool, that is specific to bosonic encoding of paired electrons in molecular ground state energy problems. This enables a specialized gate set (only singles, corresponding to paired-excitations, are needed), and a highly simplified Hamiltonian whose expectation value can be measured only using three distinct measurement bases.

In addition, the adaptive orbital optimization VQE also optimizes the molecular orbitals to find a better basis for the problem. This is particularly useful in the bosonic-encoding case, where the orbital optimization can significantly improve the accuracy of the results.

Return type:: SecondQuantizationAdaptiveOrbitalOptimizationVqeCreator

static standard() → SecondQuantizationAdaptiveOrbitalOptimizationVqeCreator

Narrow the VQE type to an adaptive VQE (defaults to FAST-VQE), with chemistry default settings.

An adaptive VQE is a VQE that grows the ansatz during the optimization, by selecting gates from a pool. The difference between FAST-VQE and other adaptive schemes, such as ADAPT-VQE (https://arxiv.org/pdf/1812.11173) is the gate selection strategy. FAST-VQE uses a heuristic strategy to select gates (either a heuristic gradient or heuristic selected CI), while ADAPT-VQE uses the exact gradient.

To use ADAPT-VQE instead of FAST-VQE, use the with_gate_selector() method with a gate_selector created by following the adapt-path in the gate_selector_creator().

Return type:: SecondQuantizationAdaptiveOrbitalOptimizationVqeCreator

class VqeCreator

Bases: object

creator for all types of VQE algorithms.

static fixed_ansatz() → SecondQuantizationBasicVqeCreator

Narrow the VQE type to a basic VQE, with chemistry defaults.

A basic VQE is a VQE that does not grow the ansatz, but uses a fixed ansatz provided by the user.

Return type:: SecondQuantizationBasicVqeCreator

static iterative() → IterativeVqeCreator

Narrow the VQE type to an iterative VQE (defaults to FAST-VQE), with chemistry default settings.

An adaptive VQE is a VQE that grows the ansatz during the optimization, by selecting gates from a pool. The difference between FAST-VQE and other adaptive schemes, such as ADAPT-VQE (https://arxiv.org/pdf/1812.11173) is the gate selection strategy. FAST-VQE uses a heuristic strategy to select gates (either a heuristic gradient or heuristic selected CI), while ADAPT-VQE uses the exact gradient.

To use ADAPT-VQE instead of FAST-VQE, use the with_gate_selector() method with a gate_selector created by following the adapt-path in the gate_selector_creator().

Return type:: IterativeVqeCreator

static iterative_with_orbital_optimization() → OrbitalOptimizedIterativeVqeCreator

Narrow to an iterative VQE with orbital optimization.

It defaults to OO-FAST-VQE, with chemistry default settings.

An adaptive VQE is a VQE that grows the ansatz during the optimization, by selecting gates from a pool. The difference between FAST-VQE and other adaptive schemes, such as ADAPT-VQE (https://arxiv.org/pdf/1812.11173) is the gate selection strategy. FAST-VQE uses a heuristic strategy to select gates (either a heuristic gradient or heuristic selected CI), while ADAPT-VQE uses the exact gradient.

To use ADAPT-VQE instead of FAST-VQE, use the with_gate_selector() method with a gate_selector created by following the adapt-path in the gate_selector_creator().