qrunch.quantum.algorithms.second_quantization.quantum_selected_configuration_interaction.subspace_state_generator

Protocols for the states defining a QSCI diagonalization subspace, and their generators.

Classes

SubspaceState	Protocol for a state whose samples contribute determinants to a QSCI diagonalization subspace.
SubspaceStateGenerator	Protocol for strategies generating the states whose samples define a diagonalization subspace.

class SubspaceState

Bases: Protocol

Protocol for a state whose samples contribute determinants to a QSCI diagonalization subspace.

A subspace state describes which state to prepare without fixing the Hamiltonian it is prepared under or the reference state it builds upon. The caller realizes it into a circuit, so successive states of one run may be realized under different Hamiltonians and on top of different references – as adaptive strategies require, where both are refined from the determinants found so far.

__init__(*args, **kwargs)

create_circuit(hamiltonian: SecondQuantizedHamiltonian, reference_circuit: Circuit) → Circuit

Build the circuit preparing this state under the given Hamiltonian and reference.

Implementations may be randomized, in which case repeated calls return different circuits.

Parameters:

• hamiltonian (SecondQuantizedHamiltonian) – The second-quantized Hamiltonian to prepare the state under.

• reference_circuit (Circuit) – Circuit preparing the reference state this state builds upon.

Return type:

Circuit

class SubspaceStateGenerator

Bases: Protocol

Protocol for strategies generating the states whose samples define a diagonalization subspace.

In quantum-selected configuration interaction (QSCI), the Hamiltonian is diagonalized in the subspace spanned by Slater determinants sampled from a set of quantum circuits. A SubspaceStateGenerator decides which states are sampled; it does not compile them, so it never sees the Hamiltonian nor the reference state they build upon.

__init__(*args, **kwargs)

generate() → Iterator[SubspaceState]

Yield the states whose samples define the diagonalization subspace.

Return type:

Iterator[SubspaceState]