跳转至

TemplateOptimization

QuICT.qcda.optimization.template_optimization.TemplateOptimization 

TemplateOptimization(template_max_width=None, template_max_size=2, template_max_depth=None, template_typelist=None, template_list=None, qubit_fixing_num=1, prune_step=3, prune_survivor_num=1, backend=None)

Bases: object

Template optimization algorithm.

References

[1] Exact and practical pattern matching for quantum circuit optimization. https://arxiv.org/abs/1909.05270

Examples:

>>> from QuICT.qcda.optimization.template_optimization import TemplateOptimization
>>> # suppose the typelist is nct
>>> typelist = [GateType.x, GateType.cx, GateType.ccx]
>>> TO = TemplateOptimization(
>>>     template_max_size=None,
>>>     template_typelist=typelist
>>> )
>>> # optimize the circuit until no further gates can be reduced
>>> cur_circ = circuit
>>> while True:
>>>     circuit_opt = TO.execute(circuit)
>>>     if circuit_opt.size() == cur_circ.size():
>>>         break
>>>     cur_circ = circuit_opt

Execute template optimization algorithm.

Specify template_max_width/template_max_size/template_max_depth/template_typelist if you want to use templates in CircuitLib limiting size/width/depth/gate types. No limit if set to None. By default all templates of size 2 in CircuitLib are used.

Specify template_list if you want to use customized templates. Setting template_list will invalidate template_max_width/template_max_size/template_max_depth/template_typelist.

qubit_fixing_num, prune_step, prune_survivor_num are 3 heuristic parameters used in template matching. Their default values are recommended values. 1. qubit_fixing_num the number of additional qubits explored when enumerating the qubit mapping (default value is 1). 2. prune_step, prune_survivor_num are parameters for backward matching. Backward match will prune the search tree when depth=k * prune_step (k = 1, 2, ...) and at most prune_survivor_num maximal matching scenarios will survive (default values are 3, 1).

Parameters:

  • template_max_width (int, default: None ) –

    Limit on number of qubits of templates used.

  • template_max_size (int, default: 2 ) –

    Limit on number of gates of templates used.

  • template_max_depth (int, default: None ) –

    Limit on depth of templates used.

  • template_typelist (Iterable[GateType], default: None ) –

    Limit on gate types of templates used.

  • template_list (List[Circuit], default: None ) –

    List of templates used.

  • qubit_fixing_num (int, default: 1 ) –

    heuristic parameter for qubit exploring

  • prune_step (int, default: 3 ) –

    heuristic parameter for backward match

  • prune_survivor_num (int, default: 1 ) –

    heuristic parameter for backward match

  • backend (VirtualQuantumMachine, default: None ) –

    the target machine, by default is None.

Source code in QuICT/qcda/optimization/template_optimization/template_optimization.py 
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
def __init__(
        self,
        template_max_width=None,
        template_max_size=2,
        template_max_depth=None,
        template_typelist=None,
        template_list=None,
        qubit_fixing_num=1,
        prune_step=3,
        prune_survivor_num=1,
        backend=None,
):
    """
    Execute template optimization algorithm.

    Specify `template_max_width/template_max_size/template_max_depth/template_typelist`
    if you want to use templates in CircuitLib limiting size/width/depth/gate types.
    No limit if set to None. By default all templates of size 2 in CircuitLib are used.

    Specify `template_list` if you want to use customized templates.
    Setting `template_list` will invalidate
    `template_max_width/template_max_size/template_max_depth/template_typelist`.

    `qubit_fixing_num`, `prune_step`, `prune_survivor_num` are 3 heuristic parameters
    used in template matching. Their default values are recommended values.
        1. `qubit_fixing_num` the number of additional qubits explored when enumerating
            the qubit mapping (default value is 1).
        2. `prune_step`, `prune_survivor_num` are parameters for backward matching.
            Backward match will prune the search tree when depth=k * `prune_step`
            (k = 1, 2, ...) and at most `prune_survivor_num` maximal matching scenarios
            will survive (default values are 3, 1).

    Args:
        template_max_width (int): Limit on number of qubits of templates used.
        template_max_size (int): Limit on number of gates of templates used.
        template_max_depth (int): Limit on depth of templates used.
        template_typelist (Iterable[GateType]): Limit on gate types of templates used.
        template_list (List[Circuit]): List of templates used.
        qubit_fixing_num (int): heuristic parameter for qubit exploring
        prune_step (int): heuristic parameter for backward match
        prune_survivor_num (int): heuristic parameter for backward match
        backend (VirtualQuantumMachine): the target machine, by default is None.
    """

    if template_list is None:
        template_list = CircuitLib().get_template_circuit(
            template_max_width,
            template_max_size,
            template_max_depth,
            template_typelist
        )

    self.template_list = template_list
    self.heuristics_qubits_param = [qubit_fixing_num]
    self.heuristics_backward_param = [prune_step, prune_survivor_num]
    if backend is None:
        self.cost_measure = CircuitCost(method='static')
    else:
        self.cost_measure = CircuitCost(method='fidelity_simple', backend=backend)

execute 

execute(circuit) -> Circuit

Execute template optimization algorithm.

Parameters:

  • circuit (Circuit) –

    the circuit to be optimized

Returns:

  • Circuit ( Circuit ) –

    the optimized circuit

Source code in QuICT/qcda/optimization/template_optimization/template_optimization.py 
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
@OutputAligner()
def execute(self, circuit) -> Circuit:
    """
    Execute template optimization algorithm.

    Args:
        circuit (Circuit): the circuit to be optimized

    Returns:
        Circuit: the optimized circuit
    """

    circ_dag = MatchingDAGCircuit(circuit)

    for template in self.template_list:
        template_dag = MatchingDAGCircuit(template)

        if template_dag.width > circ_dag.width:
            continue

        matches = TemplateMatching.execute(
            circ_dag,
            template_dag,
            self.heuristics_qubits_param,
            self.heuristics_backward_param
        )

        circ_dag = TemplateSubstitution.execute(circ_dag, template_dag, matches, self.cost_measure)

    return circ_dag.get_circuit()