variational_circuit

Variational quantum circuit module. Provides support to variational circuits on mixed states. Based on QuTiP

Three sub-modules are provided:

• measure: provides some measures of qubit systems
  • com_measure: return the probability distribution of the measurement in computational basis.
  • sep_purity: measure the purity of subsystems.
  • fid_ref: measure the fidelity between the subsystem of a state and a reference state.
  • c_entropy: measure the entropy of the output of the measurement in computational basis.
  • dst: use destructive swap test [10.1103/PhysRevA.87.052330] to obtain purity of [sub]systems.
• vcirc: provides variational circuit class
  • Create a variational circuit instance with N qubits: vc = vcirc(N)
  • Input state can be attached: vc.add_input(state)
  • Add a layer of ansatz by vc.add_ansatz(x)
    • Customized structure can be used vc.add_ansatz(x,structure=custom_struc)
    • Define structure as functions:

      def custom_struc(parameters,N):
          qc = QubitCircuit(N)
          # qc.add_gate(...) ... add quatnum gates ...
          return qc

  • Evaluation is simple: vc.apply_to(state)
  • Ansatzes can be added to subsystems: vc.add_ansatz(x,pos=[0,1])
  • The unitary given by the circuit can be obtained by vc.compress()
  • Structure of the ansatzes can be read from vc.structures, which contains a list of the tuples (name_of_ansatz,subsystem,hyper-parameters).
  • Current parameters can be read from vc.para.
• optimize: utilize scipy.optimize to optimize variational circuits
  • circ_maximize and circ_minimize:

        res = circ_minimize(x0,input_state,vcircuit,test_function,*args)