This package allows easy simulation of the expectation value of a Hamiltonian operator (objective value) that is produced via a sequence of controlled unitary operations on an initial state. Using numerically efficient matrix-free representations to compute the action of Hermitian operators and their generated unitaries allows for rapid evaluation of the objective function. Adjoint methods and sensitivity equations further provide access to objective gradients and Hessians or Hessian-vector products for use in gradient-based numerical optimization schemes.
- Python 3
- NumPy
- Numba
- NetworkX
- SciPy
Under the terms of Contract DE-NA0003525 with NTESS, the U.S. Government retains certain rights in this software.