qgd_N_Qubit_Decomposition_adaptive.py

Go to the documentation of this file.

"""
Created on Tue Jun 30 15:44:26 2020
Copyright 2020 Peter Rakyta, Ph.D.

Licensed under the Apache License, Version 2.0 (the "License");
you may not use this file except in compliance with the License.
You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

Unless required by applicable law or agreed to in writing, software
distributed under the License is distributed on an "AS IS" BASIS,
WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
See the License for the specific language governing permissions and
limitations under the License.

You should have received a copy of the GNU General Public License
along with this program.  If not, see http://www.gnu.org/licenses/.

@author: Peter Rakyta, Ph.D.
"""




import numpy as np
from os import path
from squander.decomposition.qgd_N_Qubit_Decomposition_adaptive_Wrapper import qgd_N_Qubit_Decomposition_adaptive_Wrapper
from squander.gates.qgd_Circuit import qgd_Circuit




class qgd_N_Qubit_Decomposition_adaptive(qgd_N_Qubit_Decomposition_adaptive_Wrapper):
    
    

    def __init__( self, Umtx, level_limit_max=8, level_limit_min=0, topology=None, config={}, accelerator_num=0 ):

        
        self.qbit_num = int(round( np.log2( len(Umtx) ) ))

        # validate input parameters

        topology_validated = list()
        if isinstance(topology, list) or isinstance(topology, tuple):
            for item in topology:
                if isinstance(item, tuple) and len(item) == 2:
                    item_validated = (np.intc(item[0]), np.intc(item[1]))
                    topology_validated.append(item_validated)
                else:
                    print("Elements of topology should be two-component tuples (int, int)")
                    return
        elif topology == None:
            pass
        else:
            print("Input parameter topology should be a list of (int, int) describing the connected qubits in the topology")
            return
        

        # config
        if not( type(config) is dict):
            print("Input parameter config should be a dictionary describing the following hyperparameters:") #TODO
            return

        # call the constructor of the wrapper class
        super().__init__(Umtx, self.qbit_num, level_limit_max, level_limit_min, topology=topology_validated, config=config, accelerator_num=accelerator_num)



    def Start_Decomposition(self,):

     # call the C wrapper function
        super().Start_Decomposition()

    def get_Initial_Circuit(self):

     # call the C wrapper function
        super().get_Initial_Circuit()
        

    def Compress_Circuit(self):

     # call the C wrapper function
        super().Compress_Circuit()


    def Finalize_Circuit(self):

     # call the C wrapper function
        super().Finalize_Circuit()


    def Reorder_Qubits( self, qbit_list ):

     # call the C wrapper function
        super().Reorder_Qubits(qbit_list)



    def List_Gates(self):

     # call the C wrapper function
        super().List_Gates()



    def get_Circuit( self ):
        
        # call the C wrapper function
        return super().get_Circuit()



    def get_Qiskit_Circuit( self ):

        from squander import Qiskit_IO
        
        squander_circuit = self.get_Circuit()
        parameters       = self.get_Optimized_Parameters()
        
        return Qiskit_IO.get_Qiskit_Circuit( squander_circuit, parameters )





    def get_Cirq_Circuit( self ):

        import cirq


        # creating Cirq quantum circuit
        circuit = cirq.Circuit()

        # creating qubit register
        q = cirq.LineQubit.range(self.qbit_num)

        # retrive the list of decomposing gate structure
        gates = self.get_Gates()

        # constructing quantum circuit
        for idx in range(len(gates)-1, -1, -1):

            gate = gates[idx]

            if gate.get("type") == "CNOT":
                # adding CNOT gate to the quantum circuit
                circuit.append(cirq.CNOT(q[self.qbit_num-1-gate.get("control_qbit")], q[self.qbit_num-1-gate.get("target_qbit")]))

            if gate.get("type") == "CRY":
                # adding CRY gate to the quantum circuit
                print("CRY gate needs to be implemented")

            elif gate.get("type") == "CZ":
                # adding CZ gate to the quantum circuit
                circuit.append(cirq.CZ(q[self.qbit_num-1-gate.get("control_qbit")], q[self.qbit_num-1-gate.get("target_qbit")]))

            elif gate.get("type") == "CH":
                # adding CZ gate to the quantum circuit
                circuit.append(cirq.CH(q[self.qbit_num-1-gate.get("control_qbit")], q[self.qbit_num-1-gate.get("target_qbit")]))

            elif gate.get("type") == "SYC":
                # Sycamore gate
                circuit.append(cirq.google.SYC(q[self.qbit_num-1-gate.get("control_qbit")], q[self.qbit_num-1-gate.get("target_qbit")]))

            elif gate.get("type") == "U3":
                print("Unsupported gate in the Cirq export: U3 gate")
                return None;

            elif gate.get("type") == "RX":
                # RX gate
                circuit.append(cirq.rx(gate.get("Theta")).on(q[self.qbit_num-1-gate.get("target_qbit")]))

            elif gate.get("type") == "RY":
                # RY gate
                circuit.append(cirq.ry(gate.get("Theta")).on(q[self.qbit_num-1-gate.get("target_qbit")]))

            elif gate.get("type") == "RZ":
                # RZ gate
                circuit.append(cirq.rz(gate.get("Phi")).on(q[self.qbit_num-1-gate.get("target_qbit")]))

            elif gate.get("type") == "X":
                # X gate
                circuit.append(cirq.x(q[self.qbit_num-1-gate.get("target_qbit")]))

            elif gate.get("type") == "Y":
                # Y gate
                circuit.append(cirq.y(q[self.qbit_num-1-gate.get("target_qbit")]))

            elif gate.get("type") == "Z":
                # Z gate
                circuit.append(cirq.z(q[self.qbit_num-1-gate.get("target_qbit")]))


            elif gate.get("type") == "SX":
                # RZ gate
                circuit.append(cirq.sx(q[self.qbit_num-1-gate.get("target_qbit")]))


        return circuit



    def import_Qiskit_Circuit( self, qc_in ):  

        from qiskit import QuantumCircuit, transpile
        from qiskit.circuit import ParameterExpression
        from qgd_python.gates.qgd_Circuit_Wrapper import qgd_Circuit_Wrapper

        qc = transpile(qc_in, optimization_level=0, basis_gates=['cz', 'u3'], layout_method='sabre')
        #print('Depth of imported Qiskit transpiled quantum circuit:', qc.depth())
        print('Gate counts in teh imported Qiskit transpiled quantum circuit:', qc.count_ops())
        #print(qc)
        

        # get the register of qubits
        q_register = qc.qubits

        # get the size of the register
        register_size = qc.num_qubits

        # prepare dictionary for single qubit gates
        single_qubit_gates = dict()
        for idx in range(register_size):
            single_qubit_gates[idx] = []

        # prepare dictionary for two-qubit gates
        two_qubit_gates = list()

        # construct the qgd gate structure            
        Circuit_ret = qgd_Circuit_Wrapper(register_size)
        optimized_parameters = list()

        for gate in qc.data:

            name = gate.operation.name
            if name == 'u3':
                # add u3 gate 
                qubits = gate.qubits

                qubit = q_register.index( qubits[0] )   # qubits[0].index
                single_qubit_gates[qubit].append( {'params': gate.operation.params, 'type': 'u3'} )

            elif name == 'cz':
                # add cz gate 
                qubits = gate.qubits

                qubit0 = q_register.index( qubits[0] ) #qubits[0].index
                qubit1 = q_register.index( qubits[1] ) #qubits[1].index


                two_qubit_gate = {'type': 'cz', 'qubits': [qubit0, qubit1]}



                # creating an instance of the wrapper class qgd_Circuit_Wrapper
                Layer = qgd_Circuit_Wrapper( register_size )

                # retrive the corresponding single qubit gates and create layer from them
                if len(single_qubit_gates[qubit0])>0:
                    gate0 = single_qubit_gates[qubit0][0]
                    single_qubit_gates[qubit0].pop(0)
                    
                    if gate0['type'] == 'u3':
                        Layer.add_U3( qubit0, True, True, True )
                        params = gate0['params']
                        params.reverse()
                        for param in params:
                            optimized_parameters = optimized_parameters + [float(param)]
                        optimized_parameters[-1] = optimized_parameters[-1]/2                            
                        

                if len(single_qubit_gates[qubit1])>0:
                    gate1 = single_qubit_gates[qubit1][0]
                    single_qubit_gates[qubit1].pop(0)
                    
                    if gate1['type'] == 'u3':
                        Layer.add_U3( qubit1, True, True, True )
                        params = gate1['params']
                        params.reverse()                      
                        for param in params:
                            optimized_parameters = optimized_parameters + [float(param)]
                        optimized_parameters[-1] = optimized_parameters[-1]/2

                

                
                Layer.add_RX( qubit0 )    
                Layer.add_adaptive( qubit0, qubit1 )
                Layer.add_RZ( qubit1 )
                Layer.add_RX( qubit0 )
                optimized_parameters = optimized_parameters + [np.pi/4, np.pi/2, -np.pi/2, -np.pi/4]
                #optimized_parameters = optimized_parameters + [np.pi]

                Circuit_ret.add_Circuit( Layer )
   
       

        # add remaining single qubit gates
        # creating an instance of the wrapper class qgd_Circuit
        Layer = qgd_Circuit( register_size )

        for qubit in range(register_size):

            gates = single_qubit_gates[qubit]
            for gate in gates:

                if gate['type'] == 'u3':
                    Layer.add_U3( qubit, True, True, True )
                    params = gate['params']
                    params.reverse()
                    for param in params:
                        optimized_parameters = optimized_parameters + [float(param)]
                    optimized_parameters[-1] = optimized_parameters[-1]/2                        

        Circuit_ret.add_Circuit( Layer )

        optimized_parameters = np.asarray(optimized_parameters, dtype=np.float64)

        # setting gate structure and optimized initial parameters
        self.set_Gate_Structure(Circuit_ret)

        self.set_Optimized_Parameters( np.flip(optimized_parameters,0) )
        #self.set_Optimized_Parameters( optimized_parameters )



    def set_Gate_Structure( self, Gate_structure ):  

        if not isinstance(Gate_structure, qgd_Circuit) :
            raise Exception("Input parameter Gate_structure should be a an instance of Circuit")
                    
                    
        return super().set_Gate_Structure( Gate_structure )
        
        
                  

    def set_Gate_Structure_From_Binary( self, filename ):  

        return super().set_Gate_Structure_From_Binary( filename )



    def add_Layer_To_Imported_Gate_Structure( self ):  

        return super().add_Layer_To_Imported_Gate_Structure()




    def set_Randomized_Radius( self, radius ):  

        return super().set_Randomized_Radius(radius)


    def add_Gate_Structure_From_Binary( self, filename ):  

        return super().add_Gate_Structure_From_Binary( filename )
        

    def set_Unitary_From_Binary( self, filename ):  

        return super().set_Unitary_From_Binary( filename )
 

    def set_Unitary( self, Umtx_arr ):  

        return super().set_Unitary( Umtx_arr )



    def set_Optimization_Tolerance( self, tolerance ):  

        return super().set_Optimization_Tolerance( tolerance )
        

    def get_Unitary( self ):

        return super().get_Unitary()
        

    def export_Unitary( self, filename ):

        return super().export_Unitary(filename)



    def get_Parameter_Num( self ):

        return super().get_Parameter_Num()


    def get_Global_Phase( self ):
     
        return super().get_Global_Phase()


    def set_Global_Phase( self, new_global_phase ):
     
        return super().set_Global_Phase(new_global_phase)

    def get_Project_Name( self ):
     
        return super().get_Project_Name()


    def set_Project_Name( self, project_name_new ):
     
        return super().set_Project_Name(project_name_new)

    def apply_Global_Phase_Factor( self ):
     
        return super().apply_Global_Phase_Factor()


    def add_Adaptive_Layers( self ):  

        return super().add_Adaptive_Layers()


    def add_Finalyzing_Layer_To_Gate_Structure( self ):  

        return super().add_Finalyzing_Layer_To_Gate_Structure()



    def apply_Imported_Gate_Structure( self ):  

        return super().apply_Imported_Gate_Structure()

    def set_Optimizer( self, optimizer="BFGS" ):

        # Set the optimizer
        super().set_Optimizer(optimizer)  



    def get_Matrix( self, parameters = None ):

  
        if parameters is None:
            print( "get_Matrix: arary of input parameters is None")
            return None

        return super().get_Matrix( parameters )
        

    def set_Cost_Function_Variant( self, costfnc=0 ):

        # Set the optimizer
        super().set_Cost_Function_Variant(costfnc=costfnc)  



    def set_Iteration_Threshold_of_Randomization( self, threshold=2500 ):

        # Set the threshold
        super().set_Iteration_Threshold_of_Randomization(threshold)  




    def set_Trace_Offset( self, trace_offset=0 ):

        # Set the trace offset
        super().set_Trace_Offset(trace_offset=trace_offset)  



    def get_Trace_Offset( self ):

        # Set the optimizer
        return super().get_Trace_Offset()  



    def get_Optimized_Parameters(self):
    
        return super().get_Optimized_Parameters()



    def set_Optimized_Parameters(self, new_params):
        
        super().set_Optimized_Parameters(new_params)



    def Optimization_Problem( self, parameters=None ):

        if parameters is None:
            print( "Optimization_Problem: array of input parameters is None")
            return None

        # evaluate the cost function and gradients
        cost_function = super().Optimization_Problem(parameters)  


        return cost_function



    def Optimization_Problem_Grad( self, parameters=None ):

        if parameters is None:
            print( "Optimization_Problem: array of input parameters is None")
            return None

        # evaluate the cost function and gradients
        grad = super().Optimization_Problem_Grad(parameters)  

        grad = grad.reshape( (-1,))

        return grad



    def Optimization_Problem_Combined( self, parameters=None ):

        if parameters is None:
            print( "Optimization_Problem_Combined: array of input parameters is None")
            return None

        # evaluate the cost function and gradients
        cost_function, grad = super().Optimization_Problem_Combined(parameters)  

        grad = grad.reshape( (-1,))

        return cost_function, grad


    def get_Num_of_Iters(self):
    
        return super().get_Num_of_Iters()
    

    def set_Max_Iterations(self, max_iters):
        
        super().set_Max_Iterations(max_iters)
    

    def set_Cost_Function_Variant(self, cost_func):
    
        super().set_Cost_Function_Variant(cost_func)



    def get_Decomposition_Error(self):
    
        return super().get_Decomposition_Error()




    def get_Second_Renyi_Entropy(self, parameters=None, input_state=None, qubit_list=None ):

        qbit_num = self.get_Qbit_Num()

        qubit_list_validated = list()
        if isinstance(qubit_list, list) or isinstance(qubit_list, tuple):
            for item in qubit_list:
                if isinstance(item, int):
                    qubit_list_validated.append(item)
                    qubit_list_validated = list(set(qubit_list_validated))
                else:
                    print("Elements of qbit_list should be integers")
                    return
        elif qubit_list == None:
            qubit_list_validated = [ x for x in range(qbit_num) ]

        else:
            print("Elements of qbit_list should be integers")
            return
        

        if parameters is None:
            print( "get_Second_Renyi_entropy: array of input parameters is None")
            return None


        if input_state is None:
            matrix_size = 1 << qbit_num
            input_state = np.zeros( (matrix_size,1) )
            input_state[0] = 1

        # evaluate the entropy
        entropy = super().get_Second_Renyi_Entropy( parameters, input_state, qubit_list_validated)  


        return entropy



    def get_Qbit_Num(self):
    
        return super().get_Qbit_Num()