Skip to content

solver

eigenkit.core.solver

The base class every quantum solver inherits.

It lives in core rather than in algorithms because the four solvers that inherit it sit in three different packages, and optimization and chemistry depend on core alone. In algorithms, inheriting it would make those two packages depend on algorithms, which they otherwise never touch.

BaseQuantumSolver

Bases: ABC

Base class for quantum solvers that compute a result in a single call.

Unlike BaseQuantumEstimator, a solver has no training phase and no fitted state, and it does not follow the scikit-learn estimator interface. It receives a mathematical problem directly in solve() and returns a structured result object.

Four classes inherit from it: HHL, IQAE, QAOA and VQE. PortfolioRiskEstimator does not, by design. It is a scikit-learn-style estimator with fit(), estimate() and var_at(), and it has no solve() method. As a result, isinstance(x, BaseQuantumSolver) tells you whether an object can be given a problem directly, and it returns False for the estimator.

Each subclass sets its own provider default in its signature and passes it to this class. All of them use "numpy". Two of the modes of IQAE, a callable oracle and shot_mode="provider", need provider="pennylane".

Parameters:

Name Type Description Default
provider ProviderSpec

The backend to use: a registered name ("numpy", "pennylane", "qiskit" and so on), a config instance, or a backend object. The default is "numpy", the built-in simulator that the core installation includes.

 'numpy'
Example 
from eigenkit.core import BaseQuantumSolver
from eigenkit.optimization import QAOA

isinstance(QAOA(), BaseQuantumSolver)   # True

solve(*args, **kwargs) abstractmethod

Compute and return the solution.

Each subclass declares the signature that its problem needs. HHL takes a matrix and a right-hand side, IQAE takes an oracle, QAOA takes a problem, and VQE takes a Hamiltonian and the number of qubits.